When troubleshooting vintage computer systems, systematic diagnostic procedures are essential: first verify power and basic connections, then check for NVRAM sanity failures which indicate memory integrity issues, and finally test hardware components by swapping drives and checking termination. The AT&T 3B2 computer demonstrates how a 'bucket of parts' machine with multiple variations (non-Theta RW power supply, different floppy drive cage, different card cage design) can be diagnosed through methodical testing, including ROM analysis showing 99.97% consistency across machines with only checksum differences, and identifying potential main board faults when disk detection fails despite working drives.
Deep Dive
Prerequisite Knowledge
- No data available.
Where to go next
- No data available.
Deep Dive
What Can We Do With A Broken 1980s Computer?
Added:Hello everybody. Welcome back to Retro Computing Ranch. On the bench today, we have the oddball 3B2. Now, this machine is kind of a bucket of parts.
It's got a lot of issues with it, and it looks strange. So, we're going to dig in further and see what we can find out about this machine today.
So, let's go ahead and take the top cover off.
And as you can see, this is our oddball 3B2.
Now, the things that we had discussed before are the non-Theta RW power supply, the different floppy drive and floppy drive cage, and the card cage is of a different design.
Now, this unit was also missing a RAM module. It only had one 1-meg module in it. I borrowed a 1-meg module from another 3B2 to put in here just for testing purposes.
So, uh we need to go ahead and get a hard drive installed here.
Now, one of the things that I noticed looking at this machine is that it is missing a lot of screws.
That makes me feel like this machine has probably been apart. Now, I don't know if that means that it was being worked on, or if that means that >> [sighs] >> like I said, it's a bucket of parts.
It's basically spare parts that somebody just stuck together, and maybe it's never worked right. Who knows?
But, we're going to find out.
So, we've got our hard drive installed.
Uh I missed the power connector.
So, let me dive back in and get that.
Now, this drive is the drive that had the operating system on it, and during the process of trying to recover it, it became so damaged that it was completely unreadable.
I've since taken this drive and attached it as a slave on one of the other 3B2s and I got it to format. So, I feel like this drive is good.
So, it's a a good candidate for us to use. I'm going to disconnect the floppy drive.
While I have a box of floppies, I do not have the 3B2 maintenance diskette uh on physical floppy, but I do have it in my GoTek.
All right.
I think that is going to get us set up for the most part.
So, let me finish setting up the rest of the connections to the machine.
We're going to start the machine and see what happens.
Okay, so we've got everything cabled up.
We're ready to go.
We're going to go ahead and hit the power switch. Now, I've got you set up to where you see the front of the machine. I can't. Hopefully, everything is okay with the front of the machine and then we've got the view of our terminal application here. So, in fact, before we do anything else, let me check to make sure I am on the correct serial port. I am. I'm going to hit the power switch.
Our hard drive is spinning up.
We've got our NVRAM sanity failure.
So, now we have to wait through this process and of course, as you know from previous episodes, when you've had an NVRAM sanity failure, it takes an extra long time for the machine to come around.
So, while we're waiting for that, I will tell you I have pulled the EPROMs out of this machine, dumped them, and uh did some reverse engineering, some software analysis on them, compared them to the ROMs that Seth Morabito has on his site, and some of the ROMs from my other 3B2s.
And what we figured out is that basically they are 99.97% consistent with each other.
The only difference being that the ROMs between the different machines, the last three bytes are different. So, we believe that that is most likely going to be checksum and version, but everything else is consistent throughout. So, it looks like all the subroutines are the same, and that basically almost everything is the same.
All right.
So, we've got a disk sanity failure, and we've got our system failure prompt. Let me type MCP, enter.
Our only option is the FD5.
So, it did not like that drive, but I know that that drive works.
Let's try a couple of tricks. Uh let's play with termination, and let's play with uh the cabling, and we want to see if we can figure out uh what we need to do to get this to work, or if we possibly have a problem with the main board. That may be the reason that this was a bucket of parts machine.
There may be a fault with the system board. So, let me grab another drive and some other tools, and we'll get started on that.
All right. So, one of the big things that we run into on the 3B2, especially with the hard drive cabling, is that everything is so tight in here.
So, what I want to do is first disconnect power, disconnect our console cable.
We're going to pull the system RAM.
And once that is out, we're going to lead into one of the things that makes me think think that this is a uh a big bin of spare parts.
The card cage was not bolted down anywhere.
Memory retaining bracket was missing.
So, it was uh very easy to get to the system board on this machine. And instead of having cap head screws, it has flat head screws.
So, let's go ahead and uh we'll just pull those out real quick.
I need to get uh more lighting going on in here.
That's one of the things I miss about the studio is having a ton of lighting.
I'm going to disconnect our LEDs and our battery.
Our battery is flat anyway, so it doesn't even need to go back in.
Now, that one does not want to come out.
We'll just get that at the end.
So, we'll have these screws out momentarily. There's not even the correct number of screws here.
Well, I didn't have that one out all the way apparently.
Wow, that screw is uh extra long apparently.
Let's pull our serial connection here.
Okay, and the only thing left is the main board power connection.
We pull that.
And now we have all of our ribbon cables.
>> [snorts] >> So, looks like the one closest to the CPU and MMU is going to be the floppy connector.
This one that's next to it goes to the rear panel connection, so this will be the drive one data connector.
This will be the drive zero connector.
Maybe that's the data connector.
I need to go back and look at what's what on the uh the 3B2.
Well, on MFM really.
Oh.
Yeah, we've got uh I don't know if you can see it. We've got gooey feet.
Gooey degraded feet some seem syndrome, the gooey degraded feet syndrome that we had on the uh the other 3B2 that we took apart and cleaned.
All right.
Let's set that aside.
And of course I dropped the screw.
Fortunately, I swept earlier, so it shouldn't be hard to find that.
Let's pull the hard drive out.
We'll disconnect from power.
Disconnect from the back panel connector.
Now that is sitting to the side.
Uh we're going to pull this and set it out of the way.
We will uh leave that where it's at.
Okay.
All right. Let's get this uh kind of stuck back together.
So we'll do the uh power connection first.
And uh we'll reconnect our console connectors.
We'll get lined back up generally.
And we're only really going to drop in just a handful of screws. That's really all they had in the first place.
So what I'm trying to do here is uh I want to be able to have more slack on these cables. So instead of running them under the motherboard, I'm going to see if we can get the drives back over here or on top of the card cage or something where they uh they're in a better position for us.
So we just need to get our screws tightened back down.
And of course, everything is a flat uh not necessarily a fan.
But, that's uh what was commonly used back then.
Okay.
So, we are going to definitely need our floppy connected.
I think we could even fire this up without the card cage.
Although, you know what?
We need the card cage to set the drive on cuz I don't think we're going to have enough Yeah, we're not going to have enough slack in our power cable to be able to do what we want to do here.
See, we just need to get this properly lined up here.
Okay.
Perfect. That's in place.
Let's uh drop in our RAM.
Okay, perfect.
So, let's get our disk over here. Now, one of the things that we're going to have to do is since we're no longer going to be connected to the back panel, we need to install a terminator pack.
So, we will make sure that we drop that in aligned correctly.
Well, there's notches at both ends.
So, which end is the correct end?
Uh let me take a look at some of the other drives I have here.
Okay.
Notches towards the power connector on these.
And then also drive ID. So, I've already changed it on this one.
But all the way that way is drive zero.
The next one over is drive one.
So, we'll get this set up.
Now, we have to make sure that we get this connected correctly.
Everything is hard to see here.
Okay, we've got that one.
And uh let's get this one.
Okay.
And then that's going nowhere, but that's fine because we're not using onboard termination.
Let's get our uh console cable connected back up.
And we'll grab our power cable.
Okay. So, we are now set up.
Uh let me go ahead and clear our scroll back buffer here.
And I think power. We need power.
Let's get this over here and plug it in.
That's great.
Okay, so we've got our working 30 MB drive connected. We've got our GoTek connected. We have things where we can access them a little bit easier.
So let's make sure that we have everything set correctly. I'm going to hit power.
All right.
We got our firmware error.
Our NVRAM sanity failure as expected.
All right, so we just have to wait our way through this process as usual.
Now, if we get the disk sanity failure warning again, I have another drive sitting off to the side over here that's ready to go.
So we'll hook it up to the system just to double-check and make sure that everything is okay in the sense that we're using good drives.
And if this continues to get disk sanity failures or not detect the disk at all, then we're going to have to figure out where we're going to go from there.
We should get there eventually.
So, we got our disk sanity failure.
If we do MCP, we are not detecting any disks.
Okay.
Let's shut down.
Let me grab another disk that we have over here that I know should be working.
Okay.
So, what we need to do is play the swap and around game here.
Now, I don't necessarily want to format this drive because this drive is the drive that we just recovered all that data off of.
So, I need to swap the resistor packs here. So, give me just a moment to do that.
And I will check yeah, everything's good on that resistor pack.
All right. Resistor pack is installed.
Uh the drive ID is set correctly.
So, we're going to connect here and make sure, of course, that we're correct, but these cables are keyed, so uh there's not a whole lot to get wrong. Although, I check every time just in case the key has broken or fallen out or something along those lines.
All right.
Let's fire up.
Okay, there's our NVRAM sanity failure.
So we just as usual have to wait our way through.
The drive sounds good.
I just love the look of those beautiful socketed uh purple ceramic and gold ICs that uh Western Electric used in this machine.
So we're going to get there in time.
You know, it's really unfortunate if this main board is bad because this is the uh the four-socket main board.
So the four sockets that it has, it has uh a socket for the CPU, of course, the MMU, the MAU, and the clock timer assembly. So we got disk sanity failure.
And yeah, we're getting the same thing here.
So I feel like we probably have a faulty disk controller IC or something is wrong in the glue logic.
And either way that's quite a bit of an issue.
But I have a crazy idea, a really crazy idea.
So I've mentioned it several times that I feel like this machine is kind of a box of spare parts.
And so, if it truly is a box of spare parts, because we already know we have a fault in the main board.
And we know that we have at least one IC fault because it's not detecting any hard disks.
And our only option is an MFM hard drive.
What I am thinking, I'll show you here.
I'm going to dig deep on this one.
The idea in my head, and it's one that I've had in my head for a while. I just didn't think it would be possible to start so soon on it.
Is uh Oh, I just broke a wire.
Well, that's not going to be an issue once I explain my idea.
Let me get these screws out real quick.
So, let me disconnect the last couple of things here.
And uh drop these screws.
What we know on this board is that >> [clears throat] >> our CPU and our MMU are working. Because if they weren't working, we wouldn't get console output.
So, we're getting that far.
We know that our oscillator is working.
With these three ICs that are all socketed, we can take them and we can build a home brew.
Now, I've never built a home brew and uh I don't know if doing a 32-bit home brew as your first home brew is a good choice, but uh nothing ventured, nothing gained. So, I'm going to get started working on it.
We're going to go through the process together.
Fortunately, I have the microprocessor manual that basically gives you all the information about the the CPU, the MMU, the clock chip, and then everything else we can figure out as we go um how to handle D-tech, how to do various other interfacing tasks, how to make it talk to RAM. I don't think I want to try to reuse this RAM. I think we're going to go the SRAM route, just use some static RAM with it.
I think that'll be perfectly fine.
So, let's get started on it.
Not today, not this episode, but uh soon. We'll start cycling through on this. We'll do some regular 3B2 videos, but we're also going to do some Western Electric 32100 or Bellmac 32 home brew videos. That will give me a chance to talk about the Bellmac 32 architecture because there is some truly interesting stuff there we can cover as well. So, thanks for tuning in for watching till the end.
I hope you uh had a good time and learned something. I know that uh I've had a good time because I'm looking forward to doing this home brew. So, make sure to subscribe so that you get notifications when I make new videos and we'll see you next time.
Related Videos

TOP 15 Data compression Interview Questions and Answers 2019 Part-2 | Data compression | Wisdom jobs
wisdomjobs
281 views•2019-06-28

CTS 158: 802.11w Management Frame Protection
ClearToSend
4K views•2019-02-04

NDSS 2019 Send Hardest Problems My Way: Probabilistic Path Prioritization for Hybrid Fuzzing
NDSSSymposium
496 views•2019-04-02

How realistic is Cities: Skylines?
CityBeautiful
159K views•2019-02-14

GUIs & TUIs: Choosing a User Interface for Your Python Project | Real Python Podcast
realpython
2K views•2025-04-04

The OSI Model - Explained by Example
hnasr
225K views•2019-05-12

Cloud Computing - Introduction
elithecomputerguy
98K views•2019-10-07

From Traveler's Dilemma to Dynamic Routing | Demystifying Networking
IITBombayJuly
5K views•2019-08-04
Trending

Gremlin Arrives… While Dorothy May Takes Another Step Forward
The-moons
10K views•2026-07-23

Playstation NO DISC/NO BUY Fight Is Over...
DavidJaffeGames
4K views•2026-07-23

Americans Confused in Australia for 17 Minutes Straight
IWrocker
17K views•2026-07-23

FURIOUS Raskin CORNERS DOJ over Trump DARK PAST!!!!
MeidasTouch
237K views•2026-07-23