Mike's PCjr Page

Hey all! I'm new to the forum. I just recently discovered one of these in my parents' attic, and I was having a blast with it going through the included BASIC manual and exploring some of the many floppy disks I found with it in the box. And then it stopped working

Symptoms: It was working fine, then seemingly froze trying to run a program off one of the disks. I reset it, and suddenly all it would do is emit a short beep/chirp, and the screen would flash. Seems kind of like a boot loop? The floppy drive fans continue to run. One curious thing to note is that the program that eventually froze it was giving me an error along the lines of "not enough memory to run," which I thought was odd because clearly it was used on this computer based on the floppy disks I found. So maybe an onboard RAM issue?

Here's a short video of the behavior: https://www.youtube.com/watch?v=3mPhrhAEJ_E

I found some documents explaining the POST codes on here when I was lurking last week, but none of them seem to match this behavior.

What I've tried: All the basics. I disconnected all sidecars, the floppy drive, the ram expansion card, and all peripherals. I tried two different power cords because there were two identical looking ones in the box, and I even purchased some second hand Power supplies from ebay thinking that might be the issue, but same behavior with those.

At this point I'm just looking for some guidance on where to look next. Any help would be greatly appreciated!

Thanks,
Sean

I've never heard/seen that kind of failure before, and I'm trying to figure out where the regular "chirping" is coming from. I don't think the PCjr BIOS has a failure mode where the PC speaker makes a regular noise, but I could be wrong. The fact that the chirping sounds like it's coming from the side of the power board means it's possible your power supply board inside the PCjr has failed. When you take off the top cover, it's the board all the way to the left, with the scary-looking large electronics components on it. Performing repair on that, or replacing it, would be my suggestion.

If that doesn't help, then I think your initial guess of bad RAM may be correct. Luckily, sourcing new 4164 DRAM chips is still relatively easy and cheap; you'd have to desolder them, put in sockets, then put in new ones. There are 4164/41256 DRAM homebrew testers you can get on ebay for $40 or so that can help confirm/deny bad RAM.

Hi Sean,

Sorry, I don't know what's up but I believe the main speaker is in the back corner. The chirping sounds a little like when the wireless keyboard isn't properly communicating. But that may be louder and I don't think I have ever heard it that early in the boot.

Steve

Thanks for the suggestions Trixter! I did purchase and try a couple replacement power supplies to test out the power supply theory. They were labeled as "new old supply" so allegedly were sitting in a warehouse for a few decades, so there's a chance those were also no good haha. Seems unlikely that they would all fail in the same way, but who knows!

Well I might have a new hobby on my hands, I was already looking up soldering guides and supplies after reading your message earlier. I'll investigate the RAM theory further when I can get the proper tools!


Thanks for the tip, Steve! The IR receiver was actually the one thing I didn't try removing. It would be very convenient if that turned out to be the culprit! I'll give it another shot tonight.

I think you have a short on the 5 V power rail that is causing the power supply to cycle in and out of over-current protection, effectively resetting the machine repeatedly. The power supply:

1. Ramps up the output current for a “soft start”.
2. Detects over-current and shuts down the power.
3. Waits a short time.
4. Goes to 1.

The beeper speaker (in the back-left corner) normally chirps when the machine is reset. You can hear it after hitting Ctrl-Alt-Del, for example. The fan runs because it’s connected to the 12 V rail, which is independent.

If you check the 5 V rail with a multimeter in volts mode, you may be able to see the voltage jumping up and down.

Because you’ve disconnected everything else, the short must be on the motherboard. On machines of this era, tantalum capacitors are the prime suspects for shorted power rails. The only tantalums in the PCjr are on the diskette drive’s board, however, and you’ve eliminated them as possibilities by disconnecting the diskette drive.

Here are some things to try:

• Examine the motherboard closely for any metallic debris that might have fallen in and caused a short. Look at components that stick up, making sure they haven’t been bent in a way that would cause their leads to touch.
• Look for any charring or things that smell burned. (Unlikely in this case, I think, because of the over-current protection.)
• Leave the machine running (though repeatedly resetting) for about 5 or 10 minutes, then start feeling the tops of the chips with your fingers. Shorted chips often run noticeably hotter than the others, though the over-current protection may make the difference less noticeable in this case. Prime suspects are the RAM chips (in a 2×4 array in the front-left corner), the character ROM (the fat chip immediately behind the IR receiver), and the BIOS and BASIC ROMs (the two fat chips behind the left cartridge slot), though any chip could be at fault.
• If the ROMs and CPU are in sockets on your motherboard (as they usually are), you can try gently prying them out of their sockets. The machine won’t boot without them, obviously, but shouldn’t reset repeatedly if one of them is the culprit.

One curious thing to note is that the program that eventually froze it was giving me an error along the lines of “not enough memory to run,” which I thought was odd because clearly it was used on this computer based on the floppy disks I found.

This might have been a coincidence. I see three sidecars in your video and bet that at least one of them is extra RAM. Most software won’t be able to use that RAM if you haven’t booted with something like JrConfig in your config.sys. Furthermore, a spontaneous RAM failure would cause your machine to lock up, corrupt the screen, behave erratically, or something like that, not just give you a message about not having enough memory.

I did purchase and try a couple replacement power supplies to test out the power supply theory. They were labeled as “new old supply” so allegedly were sitting in a warehouse for a few decades, so there’s a chance those were also no good haha. Seems unlikely that they would all fail in the same way, but who knows!

New-old-stock PSUs from a warehouse…hmm, sounds as if they came from Computer Reset. CR had (and still has) boatloads of NOS long-board PSUs, and of the 100+ we used in PCjr kits, I’m aware of only 2 that failed. They seem pretty solid.

Another thought: If you can’t “finger” the culprit due to its heat, you can use a multimeter in ohms mode instead.

• Take out the power card.
• Confirm the short by measuring resistance at the power connector between 5 V and ground. The 5 V rail uses the four pins in the middle of the connector, and ground uses the two pins on each side of the center four. Normally the resistance should be in the neighborhood of 400 ohms. When there’s a short, the resistance will be much lower—close to zero.
• Measure the resistance between 5 V and ground on the various chips (and electrolytic capacitors) around the board. Though all the readings will be low, the resistance will be lowest at the faulty component. All the chips with part numbers starting with 74 (e.g., SN74LS00N) have easy-to-find power pins in the corners:
  

  Code: Select all

           5V
            | | | | | | |
            _____________
           [             |
  notch --> |  74LS00    |
           [_____________|
            | | | | | | | 
                       GND
  

  For the others, you’ll have to consult the schematics in the PCjr Technical Reference, the SAMS Computerfacts for the PCjr, or look up the datasheet for the chip online.

Adrian Black made a video on this technique recently. He was looking for shorted tantalums, but the idea can be applied to other parts as well.

Thank you for the in depth suggestions! I'm going to try all of this.

DoctorOctal wrote: ↑Tue Mar 29, 2022 3:35 am [*]Look for any charring or things that smell burned. (Unlikely in this case, I think, because of the over-current protection.)

I did take another look at the motherboard during lunch today, for the most part everything looks fine from what I can tell. However I did notice one resistor (I think) that looks perhaps a little burnt or corroded. I can't quite tell, it almost looks like it's intentionally colored, but I couldn't find any close-up images of the motherboard online to check against. I'll have to figure out how to test it.

I'm learning a lot with this project already!

Those look like diodes to me

Kablamabam wrote: ↑Tue Mar 29, 2022 2:58 pm Those look like diodes to me

Right, that's totally what I meant! Hahaha I have no idea what I'm doing, but I'll get there

itsseanl wrote: ↑Tue Mar 29, 2022 3:16 pm

Kablamabam wrote: ↑Tue Mar 29, 2022 2:58 pm Those look like diodes to me

Right, that's totally what I meant! Hahaha I have no idea what I'm doing, but I'll get there

Ha, I remember in college having to ask one of my professors what the "chicklet looking thing" was...he laughed and told me it was a capacitor .

But yeah, I'd go with checking for a short on the M/B as suggested by DoctorOctal if you've got a multimeter.