A 74-year old battery keeps this thermometer accurate | Thermocouple Potentiometer and Weston Cell

[00:00:00]Inside this box is something very special. So, let's open it and find out what it is.

[00:00:06]Pretty interesting box. We got a couple of exposed screws, a nice sturdy leather handle, couple sets of hinges, straight out of Ferris Bueller's state.

[00:00:21]There's also some kind of knob on the side.

[00:00:26]Leeds & Northrup Company, Philadelphia.

[00:00:29]Now, that's a name to recognize back when all scientific equipment had a fancy wooden box.

[00:00:35]I think this is what they used to bring Frankenstein back from the dead.

[00:00:46]And that's interesting. There's no wires on the bottom. The contacts are made in the corners through those screws.

[00:00:55]Well, not that one.

[00:01:01]Inside that box is another box.

[00:01:08]Standard cell.

[00:01:10]You got to be very careful with this one.

[00:01:17]There it is.

[00:01:19]Never been happier to see an intact glass before. Now, if you've ever taken a chemistry class, you might recognize this unusual H-shaped glass enclosure as some kind of battery.

[00:01:32]We got some kind of electrolyte in there and two metals at either terminal.

[00:01:37]In this case, the metals are mercury on the plus anode side, and on the other side, more mercury and cadmium metal.

[00:01:48]That's why I'm holding this with both hands.

[00:01:51]The electrolyte is cadmium sulfate.

[00:01:54]There's also a a bit of mercury sulfate in there, too. But, as fun as it is to make a battery out of the most toxic materials, this actually served a very important purpose. For about 80 years, until around 1990, this battery was the definition of voltage. The chemistry inside of this cell produces a voltage of exactly 1.018 and some other digits of voltage. That means you can be anywhere in the world.

[00:02:24]All you have to do is a bit of glass blowing and get the materials, and you too can know the exact same voltage.

[00:02:32]But, luckily, it's contained inside of this glass envelope. And the only exit points are two platinum wires, which make a very good glass-to-metal bond.

[00:02:42]Uh, that means there's one thing that I'm forgetting.

[00:02:47]We wouldn't want anyone thinking there's any hazardous lead in there.

[00:02:51]The catch is, if you want to use this as a voltage reference, well, it's a primary cell. Once it's dead, it can't be recharged. The other catch is, if you want to use this as a voltage reference, you have to make sure to draw a very, very, very little current from it, or else it throws off the reading.

[00:03:08]If you were to measure it using a meter like this, I mean, it seems pretty good.

[00:03:12]You get 10 to 20,000 ohms per volt.

[00:03:15]That'd be like attaching a 10K resistor across there. But, no, to this battery, that's basically a dead short. And you certainly wouldn't want to measure it with one of these. I mean, even if this is the worst multimeter ever made, it's probably still got better input impedance than that analog one. They say you shouldn't use a vacuum tube voltmeter, either. And well, if that's true, you probably should avoid touching both of these terminals, because just the resistance of your skin is higher than that. What should work for us, though, is a lab bench multimeter like this one.

[00:03:50]This is an HP, back when HP was cool and made a bunch of cool stuff. And when this was new, it had five and a half digits of accuracy, but more importantly, an input impedance in the gigaohms range, which means very, very, very little current uh being drawn from our Weston cell. Now, this is old and of questionable origin, so I got another old one of questionable origin, this time in Fluke.

[00:04:20]Uh also heavily discounted because of the uh broken screen cover here. Fact, let's fix that.

[00:04:41]There we go.

[00:04:44]Perfect fit.

[00:04:47]You can get the fit right by first starting with cardboard.

[00:04:51]Then once you get something that fits, you move on to a little piece and uh really get the dimensions in there.

[00:05:00]So, when you cut the final one, it's the correct size.

[00:05:08]It's times like this where I ask myself, why do I have a laser cutter but not any thin double-sided tape?

[00:05:15]This is the thinnest I found. I guess it'll work.

[00:05:19]There we go. Good as new. Now, both of these meters do have I triple E 488, which uh well, we'll use that for another project some other time. All right. Well, I guess we should see if these are any good.

[00:05:35]I'll just uh hook them together here.

[00:05:38]Then we'll just use this little battery as our voltage reference for now.

[00:05:43]So, if they both get the same voltage, they should read the same number.

[00:05:52]And you know what? That is pretty close.

[00:05:55]That's a good sign. It makes me think they might even be calibrated, seeing as they came from two different sources.

[00:06:02]I was kind of worried there. You might be getting into a situation where, you know, a man with one watch knows what time it is.

[00:06:09]But a man with four cuckoo clocks is never sure. But after letting them sit here for an hour, now they're off by 6 7 8 9 1 2.

[00:06:20]Still a half a digit there at the end.

[00:06:23]Not sure which one went up or which one went down, or if stacking them has anything to do with it. But I guess they're technically still in spec.

[00:06:32]Now, in truth, I'm not that concerned about messing up the cell, even though it is irreplaceable, seeing as how I don't know of anyone who would sell you one, and you definitely cannot send mercury through the mail. Well, there still are Weston cells out there in laboratories being well maintained as a precision voltage reference, this is not one of those.

[00:06:55]Those kind use a saturated solution of cadmium sulfate, but this one is a mass-produced unsaturated cell.

[00:07:05]Apparently, they do that and it improves the temperature range of this, making it more stable over wider temperature range.

[00:07:12]But it also makes it more consumable.

[00:07:15]Supposedly, these only have a shelf life of maybe a couple decades, whereas the other ones can last several. Now, I have no idea how old this one is. They were making these, I think, into the '70s when they were replaced with semiconductors.

[00:07:30]But judging by the look of that box, it could be, you know, '40s, '50s. This is definitely 50 years old. Might be 70 years old. I found a date at the bottom.

[00:07:42]Is that August 10th, 1952?

[00:07:46]So, this thing really is like What's that? 74 years old?

[00:07:50]That is kind of interesting that there's a some mercury in the bottom there under all of the under the water.

[00:08:01]So, it looks like uh It looks like there's air above that, but no, it's water. Mercury is just really heavy.

[00:08:11]I don't think it's supposed to do that.

[00:08:13]Uh this is supposed to be, you know, an amalgam.

[00:08:18]So, there wouldn't be liquid puddles of mercury in there.

[00:08:22]But apart from that, you know, I've seen pictures of ones online that are all corroded and brown.

[00:08:28]Those, I guess, you have to take straight to the hazardous waste disposal.

[00:08:33]But for today, well, let's see if there's any voltage in here.

[00:08:39]We use this meter first just cuz it shows up better on camera.

[00:08:46]Let's see, manual ranging at uh 2 V.

[00:08:52]It locks in our input impedance.

[00:08:55]And now, well, that's not bad. 1.017 instead of 1.018. Based on what I've read, once it gets below 1.018, the cell is effectively dead.

[00:09:13]But right here, we're still getting a voltage of, you know, one part in a thousand. For a 70-year-old battery, that's pretty dang good. We might as well check on the HP while we're here.

[00:09:24]Yep, about the same, 1.0174 or something. All right, so we know we've got some kind of a voltage reference here, but uh what was that other thing that it came out of? When you look at the back of this thing, and it does not look like it was cheap to make.

[00:09:41]It has all sorts of uh seemingly precision resistors here, and a bunch of big Bakelite components, and uh a meter movement of some kind.

[00:09:53]Now, luckily, this is from the time where they printed basically the whole manual inside of the case. So, let's see. Directions for multiple range potentiometer indicator. It doesn't really narrow it down that much. This dial does reveal the answer.

[00:10:10]Degrees Fahrenheit. This is a thermometer. Here's how it works. We have a galvanometer, which is basically the same thing as an ammeter, except the units are arbitrary. That doesn't sound that useful, but you can actually use it to compare two voltages. If they're the same, then no current will flow between them, and the galvanometer reading will be zero. We can come up with different voltages by using a potentiometer set up like this. Then, let's start off with a reference voltage, and hook it to the potentiometer like this, along with the galvanometer. Then, we balance the potentiometer wipers so that the galvanometer reading is zero. Now, the voltage on that wiper is going to be the same as our reference voltage. Now, we replace that with an unknown voltage that we want to measure, and balance the potentiometer again, and now say it moved twice as far as it did when it was at the reference voltage. Well, now we know that unknown voltage is twice what the reference voltage was. The same thing is true if our unknown voltage is half of what the reference voltage was.

[00:11:04]And now, we can write an arbitrary scale on our potentiometer. And there you go, you've just made a tool that can compare voltages very precisely, using only one reference voltage and the fact that you have a linear potentiometer. Pretty much everything else in this setup, like the scale on the galvanometer, or even the exact resistance of the potentiometer, can be arbitrary. There is one problem, though. If you want to use this scale later, you have to make sure the same current is flowing through the potentiometer. If our main battery voltage over here changes, that's no longer going to be true.

[00:11:37]But what you can do is put a rheostat in series with it and use that to limit the current. And then you can calibrate the whole thing by marking a special location, say the voltage of our standard cell, on the potentiometer.

[00:11:50]Then adjust the rheostat until the reading matches and now the potentiometer is calibrated again. The actual voltage we're measuring comes from one of these, a thermal couple.

[00:12:01]This is a commercial one that came from some random multimeter. And there's nothing special going on here. That little bead at the end isn't some kind of precision temperature gauge or anything.

[00:12:12]It's just two different kinds of wires that have been welded together at the end.

[00:12:17]And when you heat up that junction, a voltage is produced.

[00:12:22]Now, here's one I made out of just some random wires. We got a copper wire on this side and some kind of might be nickel chromium or stainless steel on that side.

[00:12:35]But if I heat it up, sure enough, a voltage is produced and the amount of voltage is proportional to the temperature.

[00:12:47]There's a chart you can reference depending on what kind of materials you've got. Now, for our instrument here, we've got to use the kind of thermal couple that this scale was written for and unfortunately the instructions aren't super clear on that giving a list of different types of metals.

[00:13:05]But someone had the foresight to write iron constantan over here in the corner.

[00:13:11]We know that today is a J-type thermal couple like this one here. And yes, this seemingly fancy thermal couple is just two wires that are welded together in the end inside of this fancy case here.

[00:13:25]But the two different wires make it out to this end. We've got iron which sticks to this magnet here. And another material called constantan, which is an alloy of nickel and copper.

[00:13:38]Constantan is so named because it's a material that changes very little in resistance over wide range of temperatures. Which is actually something we might want to do because a thermocouple like this, where the whole thing's at room temperature, produces no voltage. And if you don't want your scale to start at zero for room temperature, you have to compensate for it. For some kinds of setups, it makes sense to use a second thermocouple in an ice bath at 32° F. A thermocouple in that condition will actually produce negative voltage. But it seems like this model is one with an automatic reference junction compensation. Which probably means that some of the resistors in here are temperature sensitive. First, we need a power supply. The standard cell is back inside of that, but it's only voltage reference. This was originally powered by a number six dry cell, which would have went inside of this door.

[00:14:35]Although it's obviously gone.

[00:14:38]Luckily, they did think of everything.

[00:14:40]If you don't have the internal battery, you can connect an external one to these binding posts here. And a number six dry cell is a 1.5 V cell just like this AA from earlier. So, this is going to work for us.

[00:14:57]Next thing to do is to make sure this galvanometer is set at zero. So, we remove the clamp and then turn this knob so that the needle rests at zero.

[00:15:11]Next, we need to adjust the potentiometer current using the rheostat. So, for that, you take off the knob and put it in the socket on the left.

[00:15:22]Then you use the standard cell switch and adjust this so that the galvanometer is at zero.

[00:15:31]So, we're using the standard cell as a reference here to set the current from our double A power source over there.

[00:15:40]That looks pretty good.

[00:15:41]Now, I can attach the thermal couple and I didn't look up to see which terminal is the positive, so I'm going to guess that it's red. Oh, actually it says up here.

[00:15:54]The iron terminal is positive.

[00:15:59]So, was that That was the red one.

[00:16:04]So, that goes on positive.

[00:16:06]All right, now we're finally ready to make some measurements. So, we choose either the red or black scale on the dial.

[00:16:15]I'm going to choose uh black cuz that's I think the lower of the two.

[00:16:20]Then we push this button and we balance the dial where the galvanometer is zero.

[00:16:29]Okay, and that should be about room temperature, which is what's that? 68° F about? Now, let's check that against the room temperature. We got uh this guy that just hangs out on the wall and it is oh, almost exactly 20° C, which is 68° F.

[00:16:48]Which is 68° F.

[00:16:50]So, this thing is right on accurate.

[00:16:54]Now, let's see. If this thing is actually working, then if I hold the thermal couple, the temperature should go up.

[00:17:01]So, you can push in this button and turn it and that will keep it turned on.

[00:17:07]And then, I'll just hold the end of this.

[00:17:14]And [snorts] sure enough, our galvanometer is drifting.

[00:17:17]But then I turn the potentiometer knob up and it goes back to zero.

[00:17:24]So, that is working. It says it's about 80° F. And since we're here, we might as well check the accuracy. Let's try some boiling water. And while we're waiting for that, we can check some ice water, too.

[00:17:37]All right, it's been in there like a minute or so.

[00:17:41]And let's just make sure this is as close to zero as we can get it.

[00:17:53]And yeah, that right there is about 32°.

[00:17:58]Amazing.

[00:18:00]And I'm just now hearing our water start to boil.

[00:18:04]So, let's uh pop this in here.

[00:18:11]Instantly starts going up.

[00:18:20]All right, and that there is some boiling water.

[00:18:23]So, let's go back to our potentiometer here.

[00:18:28]Just go off scale just to make sure we're not cheating.

[00:18:32]Very carefully lining up that zero.

[00:18:41]What do we get?

[00:18:44]Wow.

[00:18:45]That's pretty close to 212.

[00:18:51]So, there you go.

[00:18:53]74 years old still measures temperatures perfectly accurately.

[00:18:59]It's physics.

[00:19:01]Let's try for something a little more aggressive. I'll put it at the bottom of this alcohol flame.

[00:19:10]Something like that.

[00:19:13]It's going up. Might have to change to the red scale. Oh, that's too high.

[00:19:18]So, it's less than 600.

[00:19:21]Yeah, still going up.

[00:19:23]Let's see. I guess I should be able to switch over to the red scale now.

[00:19:30]It has 600 at the minimum.

[00:19:35]And yeah.

[00:19:36]And that seems to be about where it's ended up, somewhere around, you know, the 600s.

[00:19:43]But, the top part of the flame should be hotter.

[00:19:47]So, if I lift this up, then the temperature goes up, oh, quite quickly.

[00:19:57]1,100, oh, we're almost off the scale.

[00:20:02]That's 1,200° right there.

[00:20:06]I should probably take this off then.

[00:20:10]>> [snorts] >> All right. Well, that's going to do it for the thermal couple potentiometer.

[00:20:15]If you want to measure temperature with electricity, you don't need fancy active components or vacuum tubes. All you need is some batteries and a bunch of different kinds of metal.

[00:20:27]Pretty cool.