This masterclass brilliantly bridges the gap between theoretical circuit design and the practical thermal constraints of high-precision sensing. It is an essential deep dive for any engineer who wants to understand the sophisticated trade-offs behind laboratory-grade instrumentation.
Install our extension to search inside any video instantly.
Talking About Joulescope Schematic: How Does It Measure Voltage and Current So Precisely?Added:
In this video we are going to talk about precise current measurement from very low currents up to 10 amps and uh we are going to go into details. We are going to talk a little bit about schematic. We are going to talk also about what is important and how to design this kind of circuits. So uh everyone who is uh who needs this kind of measurements or who are who is designing these kind of uh devices then uh it can be very interesting uh to watch this video. Is this what we are going to talk about Matt?
>> For sure. Yeah. looking forward to talking about dual scopes, how we make great current measurements, um great voltage measurements, and uh can actually help people design awesome products that are lower battery power, um or longer lasting. And uh we have a whole bunch of things to talk about with analog signal chain, with processing, um the TI parts that we use that help enable the precision that we're now delivering with our third generation product, the JS320.
>> And uh we are going to talk also to Josh. Josh is from Texas Instruments and u Josh what you are going to help us with I guess with uh you do you know the Texas Instrument components? Uh >> yes. Yep. I'm familiar with the Texas Instruments portfolio. I'm Josh Brown a product marketing engineer um with Texas Instruments spec specifically on the precision ADC team and Texas Instruments has a very broad portfolio for test and measurement applications.
>> Okay. So um before we completely start I would like to say uh why Matt is the right person to talk about this topic uh because you know all over the internet you can find all the kind of videos you never know who is talking about specific topics topics sometimes these people maybe they don't really know what they are talking about so Matt why you are the right person >> well I've been uh working on this in one way or another for almost 10 years Now um so I'm the creator of dual scope um the first version of the dual scope the JS 110 we launched >> and you can you can use you can show your page I think there is a page with the dual scope correct >> oh yes I can uh let's see here we go so we start out with the JS 110 u which we launched on Kickstarter in 2019 uh our second generation product the JS220 we launched in uh 2022 and uh we're coming out with the JS3 320 which is launching uh just in in June here. The evolution from here is going from an instrument that was great to an instrument that is even more accurate and precise to now the JS320 that's going to even be more accurate. Um and we're building on this over time improving and learning things as we go. And uh right now the the JS320 is really built on a whole bunch of TI parts that enable the accuracy and precision that uh we need to in order to deliver the JS320.
>> What can we do with this?
>> And I like how you Sorry. Go ahead, Robert.
>> What can we do with this device? Can you can you be more specific?
>> Sure. Yeah. With the JS320, it's it measures simultaneously voltage and current. Um that's really what it comes down to. uh then it can multiply that together to compute power and then integrate over time to compute charge and energy. And with that information, if you're designing a product, you can optimize for longer battery life or lower energy consumption. So you can, you know, plug in your your normal setup, whatever you would normally be developing, put a dual scope in the middle on the current side, and it it allows you to instantly see your voltage and current consumption. you make changes in real time and you can instantly see that change and how much you've affected your your battery life.
>> And I think what I would what I would like to highlight uh because maybe some people if they don't know or Julo maybe they can be like oh I can use DVM or something. I think the what is one of the really important things to mention you can do measurements very quickly. So uh actually we can see even like instruction level of current consumption for example of a microcontroller.
Correct.
>> Yeah for sure. And one things that that's different with the dual scope compared to like a DMM is it's super fast auto ranging. So typically with the DMM it will auto range and take the fast ones are about 3 milliseconds. In that time if you go from micro amps all the way up to an amp your device is going to brown out. Um, so the way that you have to do it with a DMM is you set your device up in different modes, measure each of those modes independently. So it's a a manual process. Whereas with dual scope, it's high dynamic range. It measures from nano amps all the way up to 10 amps seamlessly auto ranging. Um, super fast so that your device always keeps working. No brown out. 20 molts burden voltage. So it's super low burden voltage, meaning that your device barely knows that dual scope's there. um whereas a multimeter you know some of them are 6 volts you are or even more at the at the higher current ranges.
>> Again I would like to highlight you can even see like when for example your modem is connecting to network when your Wi-Fi is sending data you can see all these current changes that's how precise it is correct >> definitely and what you see behind me which we'll be looking at in a little bit is a demo where we can show just two different setups. one that is much more controlled and one is much more like a real product. Um, it's using one of the TI eval kits for the Bluetooth low energy products.
>> Mhm. Okay, let's go back to the schematic because I think many hardies and engineers are >> I do briefly want to touch on the I am happy that Matt brings up the ADCs on this part because how the ADCs help to enable each generation of the dual scope that he's designing. I do want to point out that a lot of the ADCs that he was showing, the ADS 7056, the 9 uh the 9226 and ADS127 L14 have been multiple generations that we've been targeting towards test measurement applications, specifically integrating features like with the L14, a lot more digital features, which I assume we'll touch on later in this video. Um, that the L14 enables much higher flexibility with the features inside the ADC chip itself. And even in the future, we're looking into integrating anti-aliasing filters inside the ADC chip to reduce the overall burden on external anti-aliasing filters in signal chains.
>> I think we already started this uh topic to make it like look more complicated because maybe some people they think I only need ADC uh and I can measure whatever I want. But it's not so simple.
Correct.
>> Of course. Yes. you always need some sort of front end either to drive the ADC or to do some sort of filtering to feed into the ADC chip itself. Now, of course, in the future, we're always looking to make it easier on the customers whether we integrate inside the ADC chip or we improve the out the external amplifiers to make the overall signal clearer for the customers.
>> Okay, now we can go back to the schematic because I think the schematic is exactly what is before the ADC, correct?
Exactly >> correct. Yeah. So this is the analog front end. So typically with uh any system that is not just directly into an ADC which most analog you know acquisition systems have something in front. Um this is what is in front for jewel scope to convert the input signal into what we drive the ADC with. Um so it starts out with a shunt resistor over here. uh dual scope is a shunt ammeter which is just a fancy way of saying that it measures voltage over a known resistance and then multiplies that together to compute uh current or it actually divides. So V V equals IR. So current is equal voltage divided by resistance. Um and in order to compute that voltage we need to amplify it up.
So dual scopes have a full scale range at their input of 20 molts. That's the maximum burden voltage. uh and in order to get from that to what the ADC wants, we have to do some gain. Uh with dual scope, it's super important that we measure accurately. Uh ideally, this resistor is a fixed known value that never varies. Everything in the signal path is a constant gain with no no any no real world problems. It's all ideal.
And then you go into an ADC that's ideal. Um unfortunately, real world problems come into the pie. Um, so engineering is all about trade-offs.
With the signal chain we have here, we've selected for precision where we need it. And as you can see, we have a number of different parts. We have an OPA 387, OPA 387, OPA 388, and an OPA 323. Um, we've actually selected these parts very carefully. It's not an accident that they just have different numbers. They're all selected to meet the performance needs of that stage of the signal chain. So when we start out at the very beginning here, we have the needs of low-level voltage measurements.
Those have to be accurate over time.
Temperature um input bias current is a huge thing because we're trying to measure down to nano amps. So this front end op amps input bias current is super important to us. Um because any of that is actually real measurable current from your target device. Um so it has to be super low. uh and then as we go through each of these gain stages the demands of the offset performance you know gets less and less but still is important to us. Uh we've selected parts here the OPA 387 and 388 that are zero drift op amps and what that means they have extra circuitry just built into them that helps them maintain their performance over time and temperature. Um so we don't have to worry about that. Um but it makes it so that you know we as dual scope TI have to worry about it a lot.
Um and it makes us uh just be able to design a signal chain that works over all the time and temperature requirements that we have. Um the OPA 388 here doesn't have quite as good of an offset performance as a 387, but it's a little bit lower noise, higher bandwidth. So at that stage of the signal chain, it's uh it's it's the right selection for us.
>> This was actually one of my questions. I wanted to know like why you used four of them. So basically if I understand right each of them has very special features which you need uh for each this uh step but you can't have one with everything.
That's that's the biggest problem >> right it makes it it makes it so that we can select for the right features at the right place. The the thing about these precision opamps is that they don't have you know a one megahertz you know or gigahertz bandwidth um gain bandwidth product. So what we have to do is we have to choose our gain stages so that they maximize both the bandwidth that we need and also noise and performance dynamic response. So, with the dual scope, we're aiming for about 200 kilhertz of analog bandwidth, which is pretty high from a precision perspective, pretty low from a like high-end op um oscilloscope perspective, but it it is uh enough of a challenge to get that precision, that offset performance to get nano amps. Um it's really just not done uh not possible without these type of parts. And you know our competitors you know that are out there uh also struggle to get high bandwidth uh current measurements just because it's so difficult to make these dynamic range measurements.
>> Mhm. What about the last one? Last one looks different.
>> Last one is different. Yeah. So the first three here are all gain stages with some RC filtering. Uh so five, five and four and a half. The last one is actually an intentional filter that we're using as our anti-aliasing filter going into the ADC. The ADC we're running at 16 mehz is a delta sigma and this last stage provides a maximally flat filter, the second order bessel filter that drives that ADC input. Uh in the previous generation, the JS uh 220, we used a six order bessel filter. So, one of the things like Josh was saying is the newer filter I the new ADCs have filtering built into them that make our life uh a little bit easier on the analog side.
>> Okay. Josh, would you like to add something about these components like you are very proud of them. So, why?
>> Yep. So, I think Matt covered it greatly there is the it's all about trade-offs.
So when we design our opamp specifically, we have to focus on the trade-offs between either off offering low power, offering high precision, low noise. So we try and target a very broad portfolio so that we offer the multiple types of op amps to all types of customers.
>> So depending on what your application requires, we have an op amp that can meet that need.
>> So we try and design to meet all of the needs while still designing good opamps.
So if someone is uh designing this kind of uh circuit and they don't know which one to choose uh what would be the steps or how they can get advice or find what kind of opamp they really need?
>> Sure. So on ti.com we have tons of collateral that help to decide which op amp you specifically need. You can go on ti.com. We have a lot of different apps, collateral um specifically application notes, application briefs that describe um how to choose an op amp, how to pair an op amp with an ADC, how to design the entire signal chain. Um I think I'd also pose this question to Matt because one of the things I admire about Matt is he was able to design the entire signal chain with very little help from TI using a lot of the resources we had on TI.com. So I would like to ask Matt how you were able to pick out these opamps and design the signal chain using the collateral on ti.com.
>> Lots of time and thought and well it's been also almost 10 years now of iterating. So uh yeah so basically finding parts is is a challenge as an engineer right? So if you're designing hardware just knowing what's out there u if you don't know what's out there you don't even know what you can do with it.
So you can start by having a concept but that very very quickly has to go from a concept to something that can be implemented. Um so with the way that I've approached this you I pick really the key performance criteria for where I need in that stage. So let's take a look at the the first op amp here. So input bias current is incredibly important.
offset stability is incredibly important because I'm trying to measure down to nano amps which ends up being 0.1 microvolts um on the input. So it's super small amount. So that offset stability is really important. The noise performance is important. I would like even better noise performance but there's a trade-off you have to make. I would ideally like zero no one wants noise. You want zero noise but >> ideally yes >> ideally but that's not possible right?
So you have to trade off. So what I've done throughout all of dual scope's life is trade off noise. So more noise for better offset performance. So that means that your temporal resolution is a little bit less than it could be if I went to a higher bandwidth part. But it means that the accuracy of measuring current, measuring power, energy, charge, you know, um, all ends up being much more accurate over time than it would be if I had a lower noise part.
>> So, you have to kind of prioritize what you want so that, you know, in this whole vast array of available op amps and parts, you know, from like TI's portfolio, you can then narrow down and look for the parameters that matter to you. This happens to be the best performing input bias um current or sorry input offset uh device that they offer that TI offers.
>> So did you try different different chips and measured them or how did you come up with the final solution?
>> Yeah, so again this has been several iterations. The JS 110 used a totally different chip. Um that chip had some problems. It was not a TI chip actually.
Um, this the JS220 I went to TI chips um and they their performance I actually used an OPA 388 at the input stage. The OPA 387 wasn't available when I designed that. It was, you know, several years ago. Um, now the 387 is out and it fits nicely into this, you know, gap that I would have liked to have had for the JS220. Um, but now I have the part. So, you know, TI is improving, continuing to improve parts to address needs like the precision measurement that Jules Scope makes. So, yeah, just figuring out what's available, keeping an eye on what's coming out that's new. Um, is it just part of the challenge of being an engineer and and developing great products? I'm curious now because uh you are really stressing how important the component selection is and very often what happen if uh a device is kind of simple and easy to replicate then the clones are starting usually popping up.
So uh this is what the the clones make cheaper and bad because they don't really understand what they are doing.
They they use all the kind of components or is this what what do you think about like component selection and and can you use just any components any resistors any capacitors any opamps and it will work always the same or >> well let me tell you a different story so this is people use dual scopes to measure current consumption for products one big thing that is a very common problem when you go kind of off-brand with um MLCC caps is that some of them have huge leakage currents which is really surprising huge you know meaning micro amps which when you're developing a battery powered product that could blow your entire sleep budget so in some ways the capa the capacitor is going to work it's a valid MLCC cap it it does what it's supposed to do it just happens to have higher leakage current um capacitors normally have zero should have zero leakage current if they're ideal um so Yeah, selecting components is and and selecting components from a reliable suppliers, especially when you're producing, you know, a test instrument like Juul Scope, um, is super important because I I even though I go through and test every single one of these in the factory to make sure all of them are working, if I have, you know, 5% yield fallout or, you know, 10 20, that would totally kill the product. Um, so we we're able to select quality components that help us just push these through our factory um, and also uh, maintain a level of consistency and quality and reliability that all of our customers need.
>> So it's not only about all AMS, >> it's it's about all the components, >> all devices. Yep. And I do want to build off of what Matt said, especially if when you see these clones pop up, a lot of the time they'll cut corners to save on cost. So a lot of the devices need to be specked over temperature. And on in our data sheets, we try and spec over a wide temperature range, especially with these clones that I've personally seen is they'll cut those corners and the device will drift spectacularly over temperature and it'll completely ruin the final product of whatever the customer is designing. So I think uh to Matt's point, it's very important to get devices from a very reliable supplier that actually specs the device and follows what they say in the data sheet.
>> Okay, we can go to next schematic. I see there is on the next slide there is different schematic. What do we have here?
>> Yeah, this is the same analog front end except instead of for current now it's for our voltage channel. Um so the voltage we're just measuring a differential voltage um over the two terminals on the dual scope and you can put in any voltage you want up to plus and - 15 volts.
>> The way that this works is it's really an instrumentation amplifier. So an instrumentation amplifier consists kind of of two stages. You have the input stage and then you have the differential stage.
>> Um so the input stage is this uh 2197 OPA 2197 part which is just doing a buffer. Um really it has an RC filter for EMI reasons. Um and then it has a a a filter an analog filter in here for additional attenuation over high frequencies. Um and then that goes into our differential amp here. Uh which also divides down the voltage in this case.
So with the with the the current side we're taking 20 molts and multiplying that up to about two some 2.25 volts. In this case, we're taking plus or minus 15 volts and dividing it down into that same range. So, we're going the other way. Um, but we also have a a a way that we have additional amplification here to have our two-volt range um that I didn't show on here. But uh the basics of this is that it really is a customuilt in instrumentation amp that is designed so that it has a high voltage relatively high voltage 15 plus - 15 volts here but then uh operates nicely within the rest of our instrument at the 2 and a half volt range.
>> Mhm. Uh and from here it goes to the ADC also from the first one it goes to the ADC. Correct.
>> Correct. Yeah. This also I because I didn't duplicate it here. It has the same second order bessel filter that we saw in the current range. I just kept it out because we I wanted to make sure it was spread out enough to fit on the slide.
>> Mhm. Okay. Then let's go to the ADC. And I think this is the Josh part. Correct.
>> Sure. I can start by talking a little bit about it, but then yeah, he's he's going to have a lot to that he can offer because that he's done a lot of work to uh to design this part and make sure that it's, you know, it ends up being a great match for Jules Scope um and I'm sure a lot of other test instruments uh out there with the way that the the Jew scope uses this part. We have uh this comes in right here where my mouse cursor is from either the current channel or the voltage channels. Um and then we just have a simple uh it's a kickback filter here. So all ADCs um well maybe not all but almost all ADCs will have a sampling cap on the input and that sampling cap uh will pull almost instantaneous current from whatever is upstream. So this filter here is helped um to minimize that effect so that we get accurate measurements. Uh and the ADC gets the right voltage, the op amp can drive it.
uh and that goes all into here uh in the op amp or the uh ADC stage. And from the ADC stage, we now are using with this a delta sigma filter, whereas on our previous generations, we used a SAR filter. Um this just means it's for us, it's easier to drive. Um it it doesn't have quite the same requirements for a sharp filter that we would need on the analog side. we can have it be a much more gentle filter and still get the same, you know, quality response out of that ADC.
>> Okay. So, can can we just use any ADC or are there differences between them? Are some of them better, some of them are >> for sure. Josh, you want to take that?
>> Yeah, I think I could take that. So, within ADCs, you'll see multiple types of ADC architectures. Um, so the one Matt's using here is what we call a delta sigma or sigma delta ADC. These are usually uh very good for uh great noise performance and very wide bandwidth. We also have a SAR ADC which he's used in previous generations. Those are very good for low latency type applications. So this specific ADC that he's using here is uh ideal because it has a built-in uh wide bandwidth filter and low latency filter as well as multiple other filtering options that help to reduce the burden on external filtering that needs to be designed. So this uh ADC specifically um is a wide bandwidth up to 24 bits. Most SAR ADCs that you'll see will only cap out at around 18 to 20 bits. The market has started to move towards full 24-bit SARDCs, but they're usually very expensive. So if you need a very uh cost optimized and good performance ADC for an application like dual scope, a delta sigma is definitely the way to go.
Mhm. What should be people careful about when they are selecting ADC?
Is there something special like I don't know reference voltage for example or I don't know.
>> Yep.
So, uh, as Matt's been pushing a lot with the op amps as well, it's all about trade-offs. So, with the ADC, you have to one be cognizant of the power consumption. Mhm.
>> So as the ADC starts to get higher and higher performance, it's usually going to consume more and more power. Uh depending on the architecture you choose, whether delta sigma or SAR, you have to choose between whether you want very wide bandwidth, you want very low latency, you want good noise performance, you want high speed. There are multiple different things you have to trade off between. So it's really important to identify what your specific application needs. And on tar.com we have a lot of uh calculators and application notes and briefs that help explain which ADC to choose depending on whatever application you're designing for.
>> And can we talk a little bit about the reference? Um like what are the requirements?
>> The reference. Yep. So the reference is also very important for the ADC. You usually want a very low drift reference for precision test and measurement applications because that is what the ADC base its conversion off of. So if you have a very low drift reference, you can ensure that the ADC conversion signal or the ADC code that it outputs is very precise to what the actual signal it's measuring is. So this specific reference um that Matt chose is a very precise low drift reference.
Usually references will drift over temperature and over time. So within TI, we try and uh specifically uh focus on creating very low drift references uh over temperature and over time. This very accurate readings component.
>> Correct. Yes, both the ADC and the reference are both Texas instruments.
>> Ah, cool. Uh, and >> and TI does Oh, TI does make more precise references.
>> Um, but they also cost more. So, Dual Scope, we have we have some price targets, too. So, we can't just go out and pick always the the the the mo the most uh accurate part that's out there because there is a cost trade-off, too, right? Um, so this uh this ref 66 or 6225 ends up being a good performance uh part for the price that it's it's coming in at. Uh, and that's really a huge part of driving this ADC. Um, that really is the core of, you know, converting from the analog world to the digital world, which is what we end up needing in order to see it on a screen behind me.
I just remembered some of the uh op amps sometimes they need like um not like zero and 5 volts but they may need like minus five and five or this. Can we talk a little bit about this because I I forgot what do you use to power up your uh op amps is that >> yeah so this is an op this is this ends up being 2 and a half volts uh for the analog system actually 2.6 volts. Um, I bumped it up just a little to get some headroom on, uh, the op amp side. Uh, but the reference is, uh, two and a half volt reference, which is what we're actually using to go into the the full scale range of the ADC. When you're designing a test instrument, you need to have overhead. um that allows not only just the natural variations of the parts, um but also if you if you say that you're going to be measuring, you know, a 2 amp range, let's say, and you measure 1.9, people get grumpy. Um so you want to make sure that you have enough margin in order to to measure what you're committing to, right? It's a properly designed test instrument. Uh so yeah, the full front end is 2.6 volts plus and minus by, you know, bipolar supply. One big change between the JS 110 and JS220 to the JS320 is that this op amp allows us to remain on zero biased inputs. So it operates from that full scale plus and minus 2 and a half volts as you can see here two you know 2.6 in minus 2.6 in and our signal just comes right >> in curious how how you do this. Yeah.
>> Yeah. We don't need to do a level shift.
So, one of the biggest contributors for air in the JS110 and JS220 was that level shift. So, even if you buy accurate matched resistors, it's still going to be an air as you shift up because that's going to drift with temperature. Um, with the the JS320, we don't have that issue anymore by using this ADC.
>> Interesting. But then you need to be you need to have way actually to make this kind of voltages. Correct. Minus 2.6.
>> Correct. And we use other TI >> other TI parts. We use uh there's actually a trick you can do to really easily generate minus voltages using a buck converter. Um so we use a buck converter with an LDO to generate the plus 2.6 6 volts and then we use a buck converter to generate minus 3.3 volts and then another precision LDO on the minus rail to generate the 2.6 volts.
>> Mhm. What about the digital interfaces on AD? Uh is it like important to consider which one we would like to use?
>> Yes.
>> Yeah, it's it's somewhat so from a product perspective as long as it's reasonable. Um, I'm using an FPGA in the in order to communicate with this. So, I I don't have the constraints of a microcontroller. Um, so if you're designing an ADC to be paired with a microcontroller, there's a lot more things you have to do in order to make sure that the software doesn't have to be crazy. The hardware can support it, but this part is kind of meant to be used with some custom logic. Um, it can be used in other modes, but the way that I want to use it is with an FPGA. Um, in which case I don't really care as long as the data comes in in a sensible format that doesn't add noise to the whole system. Uh, so Josh, you had something to say too.
>> Yeah, so that's a great great way of putting it, Matt. The the interface really depends on the application you're using. Um, so there's spy, IC, there's LVDS, there's multiple different types of interfaces. Uh using an FPGA is great because it allows you that custom customizability, but uh if you're using a microprocessor, a microcontroller, you usually have to be very cognizant of the interface that the ADC is using um in order to be able to read out the data efficiently and not miss any bits from the ADC readout.
>> Mhm. I'm curious because on this one I see there are there are GPIO2, GPIO3. So >> Mhm. Is there something special inside you? You you need to talk to it through I square C do some setup or how does it work? How how can you control these GPIOs? Why why do we have them?
>> So uh this one specifically uses a type of spy parallel type interface.
>> Um the GPIOs can um be written uh or can you can write to the GPIOs from a microprocessor or an FPGA. In Matt's case, I don't know if you're utilizing the GPIOs in this device, um, but this allows you to, um, essentially send commands from the ADC chip itself without having to directly interface with the microprocessor every single time.
>> Mhm. And how do you control settings of these GPIOs?
>> So, with with the way that I'm using it, and again, this chip has different operating modes. They actually have two different interfaces essentially.
There's the control interface and then there's a data interface.
>> Okay.
>> Yeah. So, at least at least the way that I'm using it, I have this control interface that I use to configure it.
And the way that I the JS320 does it is it just does this once at the very beginning and then it's in a freeun mode. It's just providing it's sampling and providing samples as they arrive.
The FPGA is then just getting in a data stream that you in some ways is very much like a video stream. you have an F-sync a clock so you know where your start of frame is or you can look at this as you know um uh not I2 I2S you know it's the same similar type sync simple simple frame and then it just toggles as you go to the next sample and the bits just come in in the way that I've configured it in a known order um so if you're only sampling from one channel which you can do or or if the eight channel version they're all going to be a little bit different in how you configure that with the way that I've selected to run this, it's only running in one mode all the time forever. Um, once it gets started at the beginning of the JS320 turning on and and streaming, it just keeps doing that for, you know, until it loses power. Uh, and that samples just keep arriving. The FPGA just keeps processing them and sending them over USB. So, there's very little, well, no configuration after that initial setup for the JS320.
>> Mhm. Okay. So if someone would like to design this kind of precise uh device, what what was like the hardest thing or one of the hardest things to figure out or or maybe doing wrong wrong wrong and then you finally were like, "Oh, now it works.
Now I'm happy."
>> Um well, there's a lot that went wrong as far as learning. Let's see. Do I have that slide here? I I don't. I think that's in our another one. But yeah, I did a ton of prototypes um going through trying to figure out what was the right trade-off. You know, when I was doing the JS 110, you know, I've I've been an engineer for many years. Um so I I I I knew a lot of uh approaches, but getting the right balance between what was good enough and what was possible was challenging. So, you can obviously spend $10,000 and buy something that can be really good or build something that can be really good. Figuring out how to build something that was competitively priced so that every engineer could afford definitely took a lot more of the time. So, there's that that balance of what works or what can work but not cost too much I think was a huge part on the hardware side and that's continued to evolve as parts change. you know, the JS110 to JS220 to JS320 parts that I'm using now on the JS320 weren't even available back then. So, that's all that's all been a continuous challenge.
Then there's uh another one that uh is kind of outside of this which is how do you then build this and know that it works in manufacturing? Um that's a whole challenge on its own right that is in more in a lot of ways more complicated than building the product.
Um, I've probably spent more time on the manufacturing side, you know, of how do you test, build, know that it works as I have on actually designing the product.
>> Mhm. We will talk about this because we are preparing another video where you will explain this. Mhm.
>> I'm just now I have like two questions like um one how can you when you were designing this how did you confirm the accuracy and also if what kind of test you did like did you do all the temperature test to be sure like it's going to work or but you needed some kind of like more precise device than what you are designing or >> Yeah. And as as the instruments have gotten better that's changed too. what I used to design and test the JS 110 with was not accurate enough to do JS220.
What was accurate enough to do the JS220 is now not accurate enough to do the JS320. So it's not like I have one answer. It just keeps changing as things go. But you can see, you know, there's test equipment up here. Um, so I rely on other people's test equipments. It's kind of how the test instrument, you know, test industry works. Everyone has to rely on everyone else's equipment in order to verify things. Um, so as my needs have gotten better, I'm actually at a point now where the offthe-shelf equipment is not good enough to give an a high enough test uncertainty ratio. So test uncertainty ratios with test instrument when you when you measure one instrument, you typically want a test uncertainty ratio of at least four. That means the the instrument you're measuring it with is four times more accurate than the one you're doing. um when you start talking about where dual scopes are now with current, even the really expensive instruments are struggling to give that test uncertainty ratio over that whole range. So with the JS320, we have a different solution that um we're using to to make sure that it works. But yes, it's a full challenge to measure that whole current range.
Measuring a much more normal current range, milliamps up to an amp is very well known. when you start getting down to micro amps and especially especially nano amps the equipment gets more and more specialized you know you you know pico a meters electrometers but that those instruments only measure a certain range they don't measure too high and they also don't like capacitance um so load capacitance is a huge problem with pico meters typically because uh of their architecture so when you think about what dual scopes are intended to measure which are target devices that are real world electrical products with bypass pass capacitors, pico meters can't don't even work. Um, which where dual scopes do. So, it's hard for us to find the right instruments to pair to measure what dual scope is measuring.
Yes.
>> It's a real problem.
>> So, now people need to use uh jeweloscope to actually reference when they are designing their own products.
>> For sure. Yeah. And one of the things that we we do have an accredited laboratory um an ISO 1725 laboratory that tests jewel scopes and we had to work with them to bring new equipment onto their scope in order to test jewel scopes. So yes, it's it's a it's a problem.
>> Okay. Okay. And about the temperature range, I'm curious because you mentioned like temperature can change accuracy completely. So what kind of test do you have to do to confirm it's still measuring correctly? Well, fortunately, we don't have to go too crazy. We're uh we're spec as an indoor instrument, so 10 to 40 degrees C as our operating temperature. And if you dual scopes will continue to work outside of those range.
It's just the specs we don't guarantee.
Um so the 10 to 40 is relatively easy to qualify. I actually have a little mini fridge with a PO electric heater cooler that I control to do those temperature tests because it's a it's a really easy range to be able to test. I don't need a huge thermal chamber in order to do that. Whereas if you're going full industrial range or extended industrial, you you need a real temperature chamber.
>> Mhm. Okay.
>> Industrial oven.
>> What else do you have in these slides?
because I would like to see also the demo but I would like to see what did we what we missed in the slides. You can start from the top. So what is >> cool I think you've met us now. Um but >> it's awesome.
>> Uh but yeah just going through we've touched on I think a lot of this. Um we're going to see the demo here in just a moment. Um and I'll show this uh so you get it real in in live uh view. Um we have you know we talked about dual scopes and people are using them. We have tons of customers all over the world um that are using dual scopes for a whole bunch of different things but mostly for developing products with longer battery life or better energy consumption.
>> So it's it's both hardware and software engineers >> hardware software we have some people using it for more on the science side as well um for some fundamental research but mostly hardware software test um end of line manufacturing. So, as you're developing battery powered products that have a 10 year life cycle, like you there are new products that coming out that have 10 or 20 year battery life, like uh energy meters, water meters, and they don't want to roll a truck to upgrade that. It's really important to know that when you produce one of those units, it's actually meeting your spec on the manufacturing line. So, people use dual scopes on the manufacturing line to make sure each and every one of the units coming off works.
So, we kind of skipped over the functional architecture, but you know, now that we've talked a lot about the details, you can step back and before the demo kind of show what dual scopes are doing. They're really simple. They measure voltage on one side, current on the other, as we talked about, collect it through an ADC and send that back over to a host computer. And uh we actually have the ability to display that using our dual scope UI. But uh you can also script all of this with Python, NodeJS, C, however you'd like.
>> So usually people who are designing like test setup or something, they can use the scripts.
>> For sure. Yeah. So initially what most engineers will do is they pull it up and just use either the multimeter view or the oscilloscope view that you see here.
Um but then as you start developing more hardware in the loop tests you even automating things with continuous integration. Um and then certainly manufacturing you want all of those to be scripted and uh we have a lot of people just using our API to do that and that's actually gotten easier to do with AI recently. You know you can just tell cloud code to take a look at the docs and do what give it what you want to do.
You know what you need to do in order to exercise your equipment. And uh it's for a lot of things it's pretty easy to get something up and running now.
>> Oh yeah. Here is the block diagram.
>> Mhm. So we've kind of gone through this.
This is where the uh the current came in. This is the the filters and opamps on the first page. This was the ADC we talked about. I've referred to the FPGA.
That's this big block here. Um and that does a lot of the digital processing uh applying calibration. Uh so all dual scopes are calibrated in the factory and uh part of that digital processing is to apply that there. We also have downsampling on the FPGA. Um we compute statistics um that can then be shared and much lower data rate uh if people want that over the host computer.
>> Oh it's and it's isolated. I think that's what you also mentioned when we had our previous call.
>> Yeah. So it's electrically isolated which is really important when you start talking about nano amps. So when your your device is hooked up to anything else, so let's say you have a target device, you hook it up to a JTAG or SWD probe, you hook it up to a UART, you hook it up to dual scope, each of those things is a potential different path for the current to essentially escape from your device or be consumed out those ports. A lot of times you don't have to worry about this. It's milliamps. But when you start talking about, you know, measuring precision and measuring battery lives that go down to sleep currents of micro amps, each little device that you attach up has the potential to totally blow your entire measurement or power budget.
>> So, dual scopes being electrically isolated means that no current flows from the sensor side here to the USB side, which means that the current that flows in is the same as the current goes out. So when you measure a nano amp with jewel scope, you know, you you know you can trust that. Um whereas if it's not electrically isolated, there's ways of um of being clever and trying to do that same type of isolation, but it gets much harder to get that nano amp level.
>> Mhm. Okay. This this was important to mention.
>> Next uh next are the three uh versions.
>> Mhm.
>> Which we took a look at.
>> Um we took a look at the ADC page. Here we have a this is a little bit more detail on that second order bessel filter. It's just showing the frequency response and the time domain response.
Um so all analog filters you can view either from frequency domain or analog domain or a time domain. And um when you when you look at them for different reasons it's important to look at different domains. So on the frequency side this is what's giving us our frequency response for the whole system.
uh it's limited a little bit by some of the RC filters that we have out front.
It's limited by the sync 4 filter inside the ADC intentionally. So um and this gives us our designed response. The other thing that's really important for dual scopes is they have to respond quickly. Um so we talked about the difference between SAR and delta Sigma.
The delta sigma is a little slower in the in the time response, but uh we figured out ways to manage that with the JS320 whereas the SAR was more responsive over time. um the the the system as we've designed it with the JS320 is actually more accurate despite that extra latency. So again more trade-offs.
>> Mhm. So this is for example important when uh we are trying to measure on like microcontroller instruction level for example when something is changing very quickly. Yeah >> correct. Yeah. So the the latency doesn't matter as much when we stay within a single current range. we can compensate for that, you know, relative to everything else. We know what that delay is. So, it doesn't really matter what it is. The delay going over USB is way way longer than any of this, right?
So, there's no realworld response that we have to have with JS 320 or any dual scope. Um, as long as we know what that alignment is. The biggest challenge that we have with delays is switching current ranges. So, we want to switch current ranges super fast and then get on to that new current range. Um, and with the JS320, we now have longer latency there, which means we can't switch quite as quickly as we could before, but that's actually not important. Um, because we are doing something that we we introduced in the JS220, which is measuring higher current ranges as we're upranging. So, we can actually measure through those current range changes. So, it's a way of cheating, if you will, from an engineering perspective. It's a way of actually measuring measuring what we care about even though we're switch in the middle of switching how we're measuring that. So jewel scopes are able to measure through current range changes very cleverly. Um and that's one of the things that really make dual scopes uh dual scopes.
>> Mhm. I I just realized you mentioned it is up to 10 amps. That's a lot actually.
>> It is. Yeah. And we're not sustained 10 amps. We're sustained three amps. Mhm.
>> Um 10 amps is a lot to sustain especially when you start talking about you know uh well our burden voltage is 20 molts up to two amps. So that burden voltage increases up to 10 amps. So that ends up being a fair amount of power that you're dissipating through a little teeny shunt resistor. Um it's not a big beefy shunt resistor. So we're limited to about uh 50 milliseconds going at 10 amps. And we have a soft fuse that helps prevent if someone is is going to be putting in, you know, 20 amps or something for a second, it'll shut that off.
>> Uh, but what this allows you to do is measure inrush currents.
>> Yes.
>> So, what happens a lot and it'll be in the demo here is when you have a product that you connect up, it'll have capacitance that it almost instantly presents and that capacitance has to charge up. Um, if you don't have the ability to measure up to, you know, 10 amps or somewhere in that range, you actually chop off that inrush current.
Um, one thing that we don't really advertise as much, but dual scopes are pretty much the best pre-ompliance USB inrush tester that's out out there on the market. Um, just because it's we integrate with the USB bet test. So you can just plug it in, it just works and you have an instantaneous test report saying that you you passed or you're presenting 20 micro uh microfarads to the USB port and the USB forum says that's bad. Um jewel scopes are able to measure that inrush really easily.
>> You know why I ask about this? Because I know uh uh many people they may have problems for example with GSM modms because they can drive very high peaks and then suddenly their boards will reset or they will lose signal or disconnect >> and they can't figure out why. So I think this is exactly why maybe they should use this kind of device.
>> For sure. Yeah. And the thing with a lot of those products is that you people tend not to think about your overall burden resistance. So your battery has an internal resistance. Your little teeny 28 gauge cables that you decide to choose rather than something a little more beefy have resistance. So when you're designing a product that has those huge spikes like cell modems, satellite um Lauraan even they have they can consume a fair amount of current especially when those radios are turning on. You have to size everything appropriately so that that voltage drop is small. You know dual scopes are designed to to make sure that they don't contribute to that problem but it doesn't help you uh without doing more measurements for solving that problem if you already introduced it to a product.
>> Yeah. But it's good to see that oh there are I I designed for one amp but actually there are ps and maybe I should add more capacitors.
Yeah, I mean a big part of engineering, especially of electrical engineering is, you know, making the invisible visible.
You know, I can't see electrons. I don't think, as far as I know, no one else can see electrons. Um, so we have to rely on test instruments to do it. And if you don't have the right test instruments, you're just guessing. Um, and that just consumes time, consumes effort. Whereas, if you can just get the right test equipment affordably enough, um, it makes things so much easier.
>> Okay, next. What do we have next?
So, we talked about the different types of op amps that we have here in order to achieve, you know, the the huge range that dual scopes are able to do. Um, and just choosing the right products uh for the right >> part of that signal chain that dove right into the signal chain.
>> One thing we didn't talk about too much, I just skipped over the voltage here, is the uh the zero drift opamps, also called chopper opamps. Uh they are pretty unique and interesting in that they are designed so that they switch their polarity at a frequency so that they can cancel out any offset error that they have because there's never no way to design perfect silicon that has no temperature correlation at all. Um if someone develops that that'd be really cool. I'll I'll be looking out for that.
But you know what one thing about it's a joke. Uh everything measures temperature. Some sensors measure other things. Um so getting rid of that temperature effect and aging effect which is also related in a way is very challenging. and chopper amp app amps.
If I go to the next slide here, they do this switching and then essentially decide if you will they integrate to figure out what that difference is and maintain that so that they self- cancel out the offset error that you would normally see accumulated in other op amps.
>> I think Josh Josh is exactly he knows exactly what we are talking about.
>> Yes. So, actually on the previous slide, you did cite that white paper there which explains it in a lot of detail.
>> Um, I would definitely recommend checking that out because these chopper opamps are a great resource because it removes an error source that you would have to worry about in your error calculation. So, these opamps being able to switch the polarity and remove that uh error over temperature, error over time, makes it uh it simplifies the overall error calculation and makes the overall signal chain easier to design.
So when >> so this technical white paper there is a good resource to check out >> when people have to start worrying about the things like this.
>> So theoretically yeah Matt I was going to say theoretically as he was saying everything measures temperature as electrons run through circuits they're going to heat up. They're going to generate heat. It's going to cause some sort of drift in the measurement itself.
So, uh, as soon as you plug in, uh, something, it's going to cause heat.
It's going to cause some drift over temperature. If you intend for something to run for a long time, you're likely going to face drift over time as well, because, uh, as hard as we try to design things to last forever, they're going to degrade slowly over time. So, essentially, if you want something to last long, if you want something to be able to power on, you're going to have to worry about drift over temperature and time.
>> Mhm. It uh, and for what kind of precision? like because maybe if someone is measuring I don't know two volts do is it still going to influence around the two volts or so uh at a very high level like two volts you probably won't see too much of course you may want to account for some sort of offset um but it won't affect the measurement too much but when you start getting down to the nano amps as Matt was saying with dual scopes that's when it really starts to affect it >> it's it's also with a lot of things like op amps and ADCs it's percentage of your full scale. So, and all the TI parts have great specs on this. So, if you're you're measuring two volts or two and a half volts like reference and you want to measure a signal that's at one volt, it's going to be pretty accurate and pretty accurate is left as an exercise to the reader. Um, now if you try to measure something that is, you know, 0001 volts and your full scale range is 2 and a half volts, now you have a dynamic range problem. And that's just fundamental of engineering in any discipline almost. Um but when you start talking about percentage of what you are measuring as full scale things get harder. So if you're measuring one in 10 human eyeball we can do that. Um one in a 100 almost you know we're getting there but now you're talking about moderate engineering. One in a thousand you're now talking engineering.
>> Um it's not something that just happens.
That's a 10- bit ADC. 10 bit precision of actual effective number of bits. When you talk 1 in 10,000, you're now at that 16 bit ADC. This is, you know, precision engineering. And uh with dual scopes, we're going to be pushing, you know, well, over the full range, we're effectively 36 bits, but we we cheat in a way by switching shunt resistors. The actual range is about 16 bit, but and that that ends up being pretty serious engineering. And that is a good point you brought up Matt is you brought up the effective number of bits. So a lot of ADCs will list the total resolution that they can measure but due to error sources you'll actually get less resolution output from the ADC itself because when you start to get in uh factor in the gain offset the uh temperature offset the reference drift.
Sometimes ADCs have references inside of them. The actual effective number of bits you'll get will be lower than what the ADC itself claims it can output. So that's another thing to be cognizant of when you're reading the ADC data sheet and choosing an ADC >> and sometimes by a lot. So it's a, you know, some of the competitors out there in the micros microcontroller space, they'll say they have 16- bit ADCs, but you only get nine effective bits.
>> That's kind of common for a lot of microcontrollers that are >> I never really noticed this. Very important.
>> Yes. So it'll still output 16 bits of data, but a lot of it will just be and then you'll get actual nine bits.
Sometimes. Yes.
>> I didn't know. I didn't know. Okay, we can go to the next one.
>> Everyone says 16 bits. We're We're good.
We're good.
>> Exactly what I see in data sheet. Yes.
>> Uh-huh. Yeah. The devil's in the details, right? Um Yeah. So, uh isolation. We talked a little about isolation and how that makes uh great uh measurements for jewel scopes and keeps things from finding other paths. Um and some other points that we kind of brought up here is how do we end up designing a full product um that actually works well you know figuring out how to ma manage and maintain noise throughout the circuit. Um so a lot of people will just throw down fite beads in analog designs. Uh fite beads are resistors and inductors. Um and the problem is that inductors when you have a capacitance out there which is your bypass capacitance form resonance circuits. So in a lot of ways you can make things a lot worse by adding in ferites. Uh unless you know a lot about your signal characteristics and where your peaks are and where your noise is and the full uh power supply rejection of your entire every component in your design. So a a lot of times if you can manage the energy consumption you know so reduce currents and choose lower current parts then you can actually just put in uh like small value resistors like in the case of one place with the dual scope before the LDO we put in a 1 ohm resistor and that's it. Um the that energy is going to be dropped in the LDO anyways but by putting in a 1 ohm resistor that forms an RC filter that ends up attenuating with a break point of 16 kHz. So when we're talking about looking at our frequency of interest, you know, for some other things that go on the the system at 500 kHz, 1 meghertz, that ends up dropping at least 20 dB um with just a simple RC filter with no possibility of resonance.
>> So uh there are a lot of circuits that I've seen over my days that I've I've uh recommended that you very carefully consider or remove your fite beads.
Actually this is uh something what uh when I was designing boards like maybe 15 years ago or something like this then uh there were many recommendations to place ferite beads into power rails and this kind of stuff and uh sometime ago I've done a video uh if I find it I will post it somewhere here and it was very interesting because I did these simulations And exactly between the ferite bead and where the capacitor is and where the power pin is it it created a kind of resonance circuit and for some frequencies it actually did power actually was not delivered to the pin for some frequencies because because of the resonance.
So I think many people they don't realize this or they may think like it just goes through. Yeah, it goes through for some frequencies, but for some frequencies it may not go through completely.
>> Yeah, for a some designs people just think they're magic. You sprinkle them down and they get rid of noise. Um, and they they are if you compute everything, >> if you sprinkle correctly. Uh, okay.
When we started about this, uh, we still have a couple of minutes. So I would like to talk a little bit about maybe layout challenges and placement challenges because many people they uh ask about I have this digital circuit on my PCB, I have this analog circuit on my PCB, do I have to split grounds or what is your opinion on this when you were designing this kind of product the layout and component placement? So splitting grounds gets it just causes problems unless you know how to manage it. Um there's two things that happen. One, return currents have to go somewhere. So if you have to cross a signal over that split ground part, that's really bad.
You're going to just create an antenna and rate slot antenna that's going to radiate. Um and in addition, the return currents are going to go somewhere else and mess up your signal. Um, with the JS220, um, I actually did have a slot, um, for managing some of the I had some really noisy digital things on one side, and I did put a slot. Everything went around that slot, so there's nothing there, and that was uh, good for the JS220 because of how some of the layout things had to happen um, on that design.
With the JS320, it is a solid ground plane on the sensor side, solid ground plane on the host side. there is, you know, because it's isolated, there is a split there, and there's always been um but there's no split on the ground plane. I've managed with the JS320 to keep the sensitive analog stuff way from all the noisier digital stuff. Um, and that seems to be doing well at managing that noise level.
>> So, the placement of the analog and digital components, that's what is really important. keep the solving ground plane and people don't have to like worry that somehow somehow currents from the digital part will go to the analog part.
>> Yeah. And it's it's really placement is part of it. I mean if you can place things far apart that makes everything easier, but it's really where do the currents go? So, if you have a digital signal, even though the part's way over here, if there's a digital signal that goes right through your analog stuff, then you're introducing those digital edge rates into your analog. And it's really edge rates that matter more than anything for noise pickup. Um, so if you have super slow edges, it doesn't really matter what's going there. That the DVDT coupling doesn't happen as much. Whereas if you have super fast edge rates that you're not controlling, so you didn't put any, you know, series term resistors or whatever on your digital, then you'll end up radiating more noise between your sensitive, you know, or to your sensitive analog from your digital. Um, one thing that was on here, you know, these little resistors, these are on the ADC.
>> They don't you don't really need them, but if you don't put them, then your edge rates are going to be a little bit higher. Now, this part does, I think, pretty well, but I'm just abundantly cautious and they only cost, you know, a fraction of a penny. So, why not? Um, I don't want to take my risk right where my analog circuitry is at having fast, you know, untamed fast digital. Mhm.
>> Um so by thinking about edge rates and how do you manage that um terminations because anything that rings and reflects back and forth just creates even more edges that end up radiating into your uh you know your sensitive analog and also they can help you uh or hurt failing uh compliance uh testing. So for radiated emissions.
>> Mhm. Josh uh does Texas Instrument have some good documents about this because you always point out some documents. So >> yes, we do have uh some good documents on our website. Another thing I do want to bring up is the clocking signal. So when you write the the trace for the clocking signal going into the ADCs or any other analog component, it's very important that you uh make sure that the clocking signal itself reaches the components at around the same time.
Especially if you're trying to synchronize multiple of the components together so that the rising clock edge occurs at the same time. Now, some of our ADCs do have internal features that can allow them to synchronize the clock edges. But with the clocking signal itself, if you can try and make the trace lengths about the same or if the parts are far apart on the board, make sure that the clocking signal itself reaches the part at the same time. Um, this allows to save a lot of burden down the line where if you need to synchronize multiple components together or you want to try and read out all the data from different parts of the board at the same time, we have a lot of collateral covering that as well on ti.com.
>> Mhm. Okay. Uh, what about the uh OP arms? Do you have some uh I'm thinking oh I I've done one video with uh Arthur I think or maybe even two videos which are really good >> about op arms. Okay I I will put them somewhere here.
>> So that uh that white paper I think was actually written by RK. So >> yes he's he's our op amp guy. He's very great.
>> So if someone is uh someone would like to learn a little bit more about opamps so then watch these videos. They were super useful. Okay, I think this was uh the last slide. Then there was the demo.
>> Mhm. Yep.
>> Okay.
>> So, we can switch over. I can uh switch the screen so we can see the demo a little bit more closely.
>> Okay.
>> Um so, let's do share to this one.
Share.
Okay. Can you now see?
>> Yes.
>> You do see that one? Okay.
>> And we also have a down camera. So the top down view here um of of the setup.
So let's see here. Let's let me get my mouse so I can drive it >> and on the right screen.
>> You can explain what do you have on table first.
>> For sure. Okay. So what we have here is we have two different demos. We have a demo with the TI Bluetooth chip and we have a TI a demo with a bunch of other things here. We'll start with the one on the left. Mhm.
>> Um, we have a dual scope that is getting power over a USBC connector here and then that goes into this little switch box where we have a number of switches that we can press. But it also has one that's going out that's always driven and I'm touching this so I'm causing po electric effect. Um, but once we once we have this going, we have a little teeny Raspberry Pi Pico that is just driving a bunch of signals here so that we can see something at the super low end of the response. So this is looking at something that's peaking at about 6 micro amps. Um so shows that that dual scope can can do really well at the low end. The first thing we have uh that we can press a switch for is this LED. When I press that LED switch, we instantly jump from these micro amp ranges all the way up to here about 13 milliamps. So >> So we are measuring I'm sorry for interrupting. We are measuring the current of the Raspberry Pi. Correct.
Uh well, we were measuring actually there were a bunch of resistors here that it's twiddling. So the Raspberry Pi I have powered um over USB, but it's using MOSFETs here to control a bunch of different shunt resistors. Yeah. So that's how we're able to get that super low current.
>> Um yeah, and on the display here, we're seeing current. I didn't explain this well, >> voltage, and power. Mhm.
>> So the three curves that you typically care about, the integral of that of charge or of current is charge. The integral of of power is energy. Um, and we can use jewel scopes to just measure that. So we can take a look here, add dual markers in, and we'll we'll take this a little better look at this in just a moment, but you can look at the average current, the integrated current, which is charge, the average power, and the integrated power, which is energy.
um just by using dual markers and that's how you can uh make super easy measurements on products that you're building or any target device that you have >> and I think you can use these values to basically calculate for example how long your battery will last correct >> exactly yeah so there's there's we we have uh some requests to make that even easier for people >> the problem is that there's there everyone has a different opinion about how that needs to be done um so we we have not tackled that as directly as you know the industry would like and that's an outstanding problem that we hope to solve in the not too distant future. Um but right now just by knowing that that energy per event you can then add up your events per time and then estimate the total amount of energy in your battery. Um the thing about batteries and powering your devices it really depends upon how you're powering it. If you go battery direct um or battery through an LDO, you care about current because the current is maintained across that. Whereas if you use a DC toDC converter, you care about energy or or um you know power because that's maintained across the DC toDC converter with some loss. Um so depending upon even how you have your target device set up, you care about different things. Um so going back over here uh yeah we took a look at this the resist the uh LED here.
>> So every button is doing something different. Yeah. One button is switching the LED other button is doing something.
>> Correct.
>> Okay.
>> So this one is a little fan which you can't feel. Um I can feel here but it's a nice motor. Um so motors produce really interesting waveforms. A lot of people think oh I have a DC motor. It's just a constant current thing. That's all it is. Well, it actually is this inductive load curve as the the motor coils switch and as the motor rotates and you actually can see here how many times a second that this is alternating um and see a nice little curve waveform on that current.
>> So you can you can calculate the the speed of the motor.
>> You can uh so if you know enough about the motor I just bought this one so I have no idea. I could probably figure out, you know, how many how many times that it has to switch per revolution.
Um, but once you know that, and if you're building something based on a stepper motor, you definitely know that.
Um, you can estimate the speed just by using dual scopes.
>> Okay, >> it's kind of cool. Um, so people are using these uh like for robotics type applications where this is now super important uh for all the new AI robotic stuff. Um, and dual scopes for not the main drive motors, but for all the more sensitive motors all work really well.
Uh, then the last one here.
>> No, I was just checking like 200 milliamps for the motor average.
>> Oh, yeah. Yeah. Well, this one we're going to see the inrush current from Happy Dancing Cactus over here. Good afternoon, Cactus.
Oh, he shorted out. Okay. Good afternoon, cactus.
There he goes. Okay. So, with the cactus here, when I press the button, remember I was talking about inrush currents. So, this spiked at uh about three and a half amps right here. Wow.
>> So, we're able to measure that inrush current. So, we went from nano amps over here all the way up to 3 and 1/2 amps.
So, just showing how dual scopes can seamlessly switch this. We're now looking at these dots are samples. Dual scopes sample at one mega sample per second. So we're looking at one microcond intervals. So we went from almost nothing right over here 100 nano amps all the way up to three and a half amps in microsconds.
>> Five mic.
>> But then we get >> Yeah, it's super fast. And then we get to see the dancing cactus which you notice there's a little of that that waveform like you saw from the motor before for the band >> coupled with some other things in here.
You know, he's doing the song and dance and lighting up. So you see the whole waveform.
So let's switch over now to the other demo that we have which is much more of a real product, right? So if you're if you're a firmware engineer or hardware engineer working on like a BLE device, this is much more what you're thinking about. Uh so let me switch to the other dual scope. So we have both dual scopes here. I'm going to switch over to the other one. We can have multiple simultaneous dual scopes with uh the UI and they can work together. Um but what we're seeing here is the device turning on periodically. I have this beaconing at one um beaconing cycles per second essentially. So it's it and you can configure this part to do whatever you want. Um I just did this for a demo um so that it's happening often enough for us to take a look at it.
>> Uh so what I can do here is zoom in on one of these beaconing events. Uh this is what happens when this product uh turns on. And this is uh let's see the part number here. This is I don't know if you can can you see that in the down cam?
>> I can see the board but not the >> CC 2340 R53 development kit. Um and I'm just powering it here from this little teeny uh desktop supply. It's a really cute miniware uh supply that comes in from USBC and it's just generating 3.3 volts that ends up powering the board.
>> Mhm.
>> And then we're just using dual scope to measure that. So in this place here, this is one of the beaconing events for the the TI BLE chip and it's consuming about 12 microcoolum of charge or 39.8 microjoules of energy. So let's say you had a battery target that where you wanted to last for a year on this. This would allow you then to say I can afford this many events per day um of this.
Likewise, we can move this over to the sleep current here. Um, let's go over here and make it a little bit bigger since there's some variation. Um, but it's consuming about 150 nano amps of sleep current, which is very impressive.
You know, BLE devices, you know, continue to to get lower and lower power. This is one that's really showcasing how low that sleep current can get. And then what we can do, um, is we can Oops. Let me go over here. Oh, I have to click on it in and let's uh I have a keyboard here. Let's Oh, hard to look both ways. Let's do it for one second here. Exactly. Oh, there we go.
Um, so I'm just going to spread this over because we know we're beaconing once per second. We can then look at this total thing. The beaconing plus all the sleep current um ends up being 42 microjles. And I think I said what 39 just a moment ago. So most of the energy right now is being consumed in that beaconing mode. Um so the trade-off the sleep is so good compared to that beaconing energy that at once per second most of that energy goes to the beaconing. So if you reduce that then your sleep current becomes more important. Um so let's say we switch to beaconing once per minute. Now your sleep current and your beaconing current may be get closer. At some point they'll end up being the same. And when people are designing products, designing both for active mode and sleep mode ends up being more and more important as they're trying to get longer and longer battery lives.
>> Mhm. And I would like to point out, can you zoom in again on this beaconing uh sequence?
>> Sure.
>> Like that's what I uh what I mean when I always mention like on instruction level. So basically we are almost see instructions how they are executing and what they are what part they are switching on and off and correct.
>> Yep. Yeah. So I don't know exactly what this chip is doing. Um I I'm not familiar with a lot of the details but you can see here there's an initial this looks like an RCDK. So it's turning on something internal to the chip. Um and then it's going and doing a bunch of different actions you know. So, some of these are going to be BLE transmits, you know, when it's pushing out uh some information over the radio. Um, some of it, you know, this flat part and this flat part are probably more on the software side, I guess, of it just managing and setting up for the next interval. Um, so this is using free art toss internally. Um, so it's it's doing everything that you need to do in order to have a microcontroller go to sleep, wake up, and be low power. Um, so there's a bunch of register transactions that are happening probably both here and here. But again, I'm just hypothesizing.
>> Yeah. But without this kind of measurement, I think it would be super hard for software engineer trying to optimize battery life.
>> Oh, for sure. And you don't necessarily need to know everything that's happening here because if you think you are >> doing something in a range, so you can toggle the GPIO. For example, if I if I was developing this and I cared about the initialization and finalization time, I could easily toggle GPIO here, which you can add to Juul Scope. So, Jew Scope has GPI inputs. And then you can see exactly when you start to enter that region and when you exit that region.
And as you optimize, you can then instantly see the changes that you've made, whether they've been effective or not. Mhm. So basically you can toggle GPIO or around the instruction which is interesting for you because you are doing something special and you would like to know current consumption exactly when this instruction is executed. Then you tidy it up to GPIO and basically this is used like some kind of flag or interrupt. Yeah, you can see it then.
>> Yeah, you can use the GPIOs however you want. So, we have six different or sorry, well, four inputs, six total um that you can connect up to your target device. Uh, and that's usually enough.
You know, by the time you get more than that, you're working on something else um entirely. So, having one or two GPIO is usually enough for most applications like this. Um, unless you try to want to go more of power profiling the exact firmware code. There there are some tools out there that try to do that.
Dual Scopes with some help do that. um but not on their own.
>> Mhm. So, and then basically you just add this GPIO into this graph and you will see. Okay.
>> Yeah. Yeah. So, I don't have the GPIO connected up but you know with the dual scope you just say um I'm on this one right now. Zero and then add in zero down here and go. It's a really boring thing because there's nothing connected to the GPIO port right here. Um but we provide a cable to make it really easy.
And those uh GPIs are designed to be low current. They are not isolated. So if you want to keep the isolation and have your good sleep current um sometimes it will work if you can if you don't connect anything else to your target device um but we do recommend some isolators that you can uh buy and design in and now you know TI uh we use the ISO 7762 uh inside the the dual scope um you can use that part or any of the other ones from their portfolio that are really really accurate and um they have a new part that is a isolated DC/DC converter converter as well. So if you start thinking about designing your own test instruments um to connect up which people do when they are developing products that are super low power and where energy really matters um you can actually use those parts off the shelf now with almost no effort. Um they are you know it's a lot easier than it was when I was designing the JS 110 you know eight years ago nine years ago.
>> Okay. Now I have a question. I think some time ago I made a video where this guy he he mentioned uh you can change the panels to make some measurements easier. Correct.
>> Right.
>> I would like to highlight this because uh because we didn't really talk about how exactly it is connected but I think you can have also like USB panel or >> Yeah. And I I don't have one within arms reach right now, but this front panel here is actually a circuit board. And it just connects in. I have a better example just sitting right here.
>> So you have Oh, it's the lighting is not set for there, but you can just disconnect this and you can there's just two screws to take apart the dual scope.
There's a YouTube video if you want to link to it of how you do this because I think it was the JS 110, but it hasn't changed very much. Um, we actually have two screws rather than four, which you know the JS 110 had, so it's even easier. Um, but we sell a number of these different front panels. Um, we also have uh drawings and a keypad base design so that you can easily customize it. We've had people contribute. Uh, there's one that is the USB all the things one. It has every single USB connector plus terminal blocks that a customer designed. Um, and it sells on Tindi. Uh, we have a few other ones as well that customers have contributed. Uh and you we've the bigger case that we've seen people do is actually just take out the circuit board and use this directly for factory automation. Um and just that's your terminal your connection to your your device under test and you just design this into your fixture.
>> Is there anything else what we would like to mention in this video?
>> Yeah, we have a a few new things uh to mention before we wrap up. One of the thing that I created was uh you know we talked about a lot of the parts that were on the dual scope the JS320 u but it's really kind of cool to see them all so where they are. So, we have the the connector that we were just talking about, which is where the sensor goes.
And then we have our first stage op amps that we saw in the schematic for the OPA 387, the OPA 388s, um the bessel filters, plus an extra one for regulating the GPIO voltage, the front end OPA 2197 for doing the voltage. Uh we have some comparators as well. Uh, we didn't talk as much about the analog switches, but the T-MX 1108, it helps make dual scopes switch between the current ranges. Um, another one that we use for the GPIO reference voltage, the reference. We talked about the ADC. Um, and then a bunch of power regulation parts. You know, this whole script was is comes out of Keycad and, uh, spent a little like one night with Claude trying to figure out how to make a nice little animation here that looked visually appealing and cool. So, uh, you'll be seeing this hopefully on, you know, media as well.
>> So, all the black chips are Texas instrument chips.
>> Yep.
>> There are many.
>> Yes, >> they they make good chips that make dual scopes work well.
>> Okay. I would like to add I would like to add uh very often when I'm designing boards, actually I select Texas Instrument as uh as the manufacturer of the chip. This is not paid advertisement or anything. Uh I just do it because from my experience uh it's uh one of the most reliable selection. So I know when I select the chip from Texas Instrument now I will be able to buy it. Uh in past it has happened to me many many times if I use different kind of manufacturers or random chip manufacturers. What may happen? One day you can buy it, other day you may not buy it and then after one year the company may disappear completely. So that's why I really like to buy components from like good cheap manufacturers.
>> Yep. And the Jew scope products. Yeah. I mean I have the ability to select from any vendor that I want and uh you can see what I selected. But I understand this because then once you are designing this kind of product and you use many components from same company it's also maybe good for the company to make collaboration with you exactly what you are doing right now because they they know like your product can be used for good you know promotion of the chips >> and I think sometimes this doesn't happen >> we especially admired >> yeah Some companies they don't do this.
>> Sure.
>> Go ahead, Robert.
>> Yeah, some companies they don't do this.
I've done some designs with some chips like and I made them open source and I made them like quite complicated and these chip companies they never recognize me. So, so I'm very happy Texas Instruments is actually doing this like recognizing people who are using their chips and also making like taking special care of these uh design engineers. So, thank you for this.
>> Yep. And we especially admired how Matt was able to build his entire board without too much help from the Texas Instrument side. We learned about the dual scope board that he was building um after we started engaging with him once he was most of the way completed. And that was a great thing on our side um because we were able to realize that we had the collateral on t.com to help enable engineers to design their signal chains. And it was good to see uh Matt the passion he showed for dual scope as well as using the TI parts to help enable his solution.
>> This is also good to mention. So you didn't like sponsor made or or push him use? No.
>> No.
>> No. I mean I I've been I've been hoping that you know TI would notice for a little while but you know they they finally did. Um you know one of the big changes and we'll we can talk about this another time. Um, on the business side, they've they've helped to support us now with a backlog account. Um, which just makes a lot of our ordering get a lot easier. Um, so TI, you know, has been a great partner. Um, now that, you know, we have a little bit more closer relationship than we've had in the past until, you know, back in what is it, November, December. Um, I didn't have an inside salesperson at TI um, really doing anything. I just order on the TI web store or mouse or Digi Key. um and just do my thing. Uh now uh with the JS2 320 launch, you know, we're a little bit more closer uh on the collaboration side uh and they're helping to support me so I can be successful as a business with that backlog account so I can uh actually forecast parts and ordering over time which you know when you start getting to a certain size is super important.
>> Mhm. So your latest scope has not been released yet correct?
>> It is coming out very soon. Um, so the end of June, uh, we are launching it.
Uh, you can check in on the Juul Scope website for more details as we get a little closer. Um, I think, uh, by the time most people see this, we'll hopefully have a little bit of information out there. Uh, but it is coming pretty soon. We passed compliance testing and, you know, which is like for every engineer the last big hurdle of stress. Um, that was just last week. Uh, and we're moving towards uh, ramping up the production and launch.
>> Okay. So I think uh when the video will go out then uh it will be very close to the one so people should be able to uh buy it very soon.
>> Yeah. Okay. Uh what else? Uh there are there is this is the last page correct?
>> Yep. That was the last one that we just skipped over.
>> Okay.
>> Yep. And I do just want to touch on we discussed a lot of TI parts um that BAT has using in his dual scope but beyond even that we have a very broad portfolio that's targeted towards precision test measurement applications. So even if the part wasn't mentioned today we likely have something that can fit for the power side, the analog signal chain or even on the embedded processing side that can be used and to enable a wide variety of test and measurement applications.
Okay. So, here is the list.
>> Yep.
>> These are all the things we didn't talk about what Matt's using.
>> Yes. But we didn't talk about today.
Yes.
>> So, yeah, we could we could talk about each of these parts, but they end up a lot of them are, you know, much more common parts except for the precision analog supplies, but again, they they make great parts across the the lineup.
And I selected these not just because they were TI, but also they're ones that were cost effective and got the job done. Um, so they're they're good parts.
>> Okay, perfect. Uh, so I would like to remind everyone uh we will be doing another video with Matt uh and talking about the like uh designing process, manufacturing, uh even like marketing.
Am I correct? And so watch out there will be another video coming soon or maybe it's already there just uh people can watch it. And if there are any questions about uh jeweloscopes then leave them in the comments. And uh there is one more thing if they would like to contact you. Oh here is the contact mjuloscope.com.
>> Yeah.
>> And I'm also Matt Liberty on LinkedIn.
Um if you'd rather use LinkedIn um and and the website we also have a forum. So if you have any technical or detailed questions that you want to share publicly, uh forum.jscope.com.
>> Oh, you can show your company website so everyone knows uh you know where to buy it. How much does it cost? You know, that's the question.
>> Yes. So let's pull that up here and type in Oh, whoa. Went above jewelscope.com.
There we go. So this is uh before the JS320 launch. This is still the JS220.
Um so we'll be updating this website. So by the time you visit hopefully it'll be uh showing the JS320.
>> And there is I see shopping cart.
Correct.
>> Yeah. So you can uh you know go to the JS220 here. Decide if you want the factory one or the NIS traceable. If you don't know what you want, you want the factory one. Um this is for people doing more medical industrial that require that. Um add it to your cart and it's as simple as that. This is a Shopify web store. So, you've seen probably a bunch of them that you've ordered from.
>> Mhm. Now, before we completely finish, because I think everyone see this and they will be like, "It's $1,000. It's $1,000."
So, can you tell them a little bit more why it's worth $1,000?
>> Sure. So, dual scopes are trying to hit a sweet spot of price. there there are things out there that you know the semiconductor manufacturers supply that are more affordable um they're meant to just give you a sense of where things are whereas dual scope is a test instrument is meant to be accurate calibrated um it we get we are able to get ISO 1725 calibration from an accredited partner so even if you don't get that what it means is that someone other than us can say that dual scopes do what they say they do >> um which is important so that's you can trust trust it as an instrument, not just as it's somewhat close. Dual scopes are accurate, right? So, it's it's the product that when you go to doing engineering, um it gives you that level of engineering confidence without breaking the bank. The next step up from this is over a price of price order of 10. Um so, dual scopes are aimed to be affordable enough for any engineer that is doing work on these type of products.
And if you're launching a a product, your company is spending way way way more than this in order to develop that product. This is what allows you to design a product with long battery life, excellent energy consumption performance um without having uncertainty. It removes the uncertainty of product development. So you don't get surprised when you go to the final integration test and your battery dies in 3 days.
Mhm.
>> You can know with dual scopes up front that your product works all the way from like day one through the end through continuous integration, through factory test. Um, and it's priced in a way that every engineer that is working on this in a team can actually have a dual scope at their desk and be responsible.
>> Okay. Okay. Thank you so much, mate.
Thank you so much, Jos, for helping me with this video.
>> Thank you.
>> That's all for this video. I hope it was helpful. If you would like to learn more about electronics and board design, check out our online courses. You will find everything important there from basic board design up to advanced board design and highspeed PCP layout. We have courses in Alium, Cadence, Keycat, and also courses covering many different topics, for example, FPGA, EMC measurements, and so on. Visit our website at fedel.com. That's all for this video. Thank you very much for watching and don't forget to leave your comments. See you in the next video.
Bye.
Related Videos
U.S. Military Just Flexed The Most Dangerous Aircraft Ever Built The F-47
MaxAfterburnerusa
11K views•2026-05-29
Heating Staying On On The Hottest Day Of The Year
PlumbLikeTom
507 views•2026-05-29
발전 효율을 높이는 태양광 추적 시스템의 기술적 원리 #공학 #공정 #태양광 #알고리즘 #재생에너지
찐현장기술
2K views•2026-05-29
Wire To Wire Connection Trick | Strong And Secure Electrical Joint #shortvideo #wireworks
ElectricianTips-b1h
5K views•2026-06-02
Peterborough to Newark Northgate Driver's Eye View aboard an InterCity 225 - East Coast Main Line
TrainsTrainsTrains
822 views•2026-05-31
AI turbine design: hypersonic cooling leap #shorts #ai #hypersonic
bobbby_rn
671 views•2026-05-31
직관 및 곡관 배관 결합 고정 작업 #worker #process #fabrication #pipework #clamp
월드촌촌
2K views•2026-05-30
How Far Can A Tomahawk Missile Actually Travel?
WarCurious
13K views•2026-05-28
