LC575 computer killed LC520 motherboard! 😭

[JDW]

Folks, I thought we could use the LC520 motherboard inside an LC575 computer body, but apparently not. My LC520 board is now dead!​

Some of you may recall my LC520 recapping video, and my 84V VGA Mod video. As you could see, I had a working LC520 board even prior to the recap, and it worked fine after recap, and I could use it on my Color Classic with the 84V Mod installed, along side the stock CC motherboard and LC575 motherboard.

In the past, my LC575 computer worked fine with the LC575 motherboard it came with, and you can see it powered in my Mystic Upgrade video. I purchased that LC575 on the cheap here in Japan for the lone purpose of extracting its motherboard, especially because the plastics were cracked, chipped, and ready to fall apart. Later, I desoldered one of the audio amplifier chips from the LC575's Analog Board in order to solder it into my Color Classic's Analog board for the Stereo Mod.

I kept the LC575 body in storage since the stereo mod. But this past week, I decided to get rid of the LC575 body, but before that I wanted to use it to check some things like the content of the spinner drive it came with. Not wanting to risk harming my precious overclocked LC575 motherboard, I decided to use my LC520 motherboard in the 575 body, because I had the understanding it was OK to do that. Sadly, it seems to have killed my LC520 motherboard.

I did the following this week, in the order listed below:​

1. I properly removed the Analog Board in order to reinstall the two Audio Amplifier Chips. (I had two spare chips taken from an LC520 Analog Board that I later junked. And yes, I soldered them into the LC575 Analog Board correctly. I cleaned off the flux with alcohol and gave it 24 hours to fully dry before testing.)
2. I reinstalled the Analog Board properly.
3. I installed my LC520 motherboard properly, and yes, the bottom metal prongs make ground contact with the metal frame of the computer.
4. I connected my ADB IIgs Keyboard and ADB Mouse to the LC520 motherboard inside the 575 body.
5. I connected the power cable, then flipped the back side switch to ON. I could hear that the CRT circuit was powered with the normal "boom" sound. It's a bit louder and bolder than the Color Classic's sound, but perfectly normal for the 575 due to its larger CRT.
6. I then pressed the power key. Nothing. No bong. No CRT display. No LED lit on front. No HDD spinning. No case back fan spun. No sign of life at all !!!!
7. I measured at the LC520 motherboards connectors to see what voltages I could read:
   (a) Meter GND to SCSI pin-7 (GND) and Meter POS to SCSI pin 25 (Term. PWR). No voltage when I flipped ON the PWR SW, and nothing when I pressed the PWR key.!
   (b) Meter GND to top left pin of ADB connector, and Meter POS to top right pin of ADB. At PWR ON, voltage slowly rose to 106mV, but it should be 5V!
8. I removed the case back from the LC575, flipped the PWR SW ON, and used my thermal meter. Various components get hot, which is what I would expect.
9. I powered OFF, removed the LC520 motherboard, then put the LC520 board into my Color Classic (with 84V mod). I flipped ON PWR, heard the CRT power-up boom sound, then pressed the PWR key on my IIgs keyboard. No bong. 100% black CRT. My heart sank! But I do hear the case fan and see the machine is powered.
10. I removed the LC520 motherboard and installed the LC575 motherboard into the Color Classic. It bonged and booted up perfectly.
11. I am not seeing any visible signs of burn marks on my LC520 motherboard. No smoke, nor any bad smells on it either.

QUESTIONS:​

1. Is it really WRONG to use an LC520 motherboard inside an LC575 body? Or have some of you done that successfully? If you have successfully done it, then it means something else inside the 575 computer body must have killed my LC520 motherboard.
2. Any ideas what might be dead on my LC520 motherboard? Identifying that would potentially lead to a fix.

Thanks.

 

[Volvo242GT]

Maybe something happened to the power supply after the initial testing. I don't think there's enough of a difference between the 520 and the 575 to make it impossible to use a 520 board in a 550 or a 575. With respect to the drive, pull it, then connect it externally to either your CC, or another Mac. That way, you don't have to worry about the 575 chassis anymore. I currently use my G4 DA with an Adaptec 29160N that has a 50 pin cable connected to it, run out one of the empty slot openings, along with a hard drive power cable. As long as I don't hot swap drives, it works fine for hard drive testing, or installing software that has been downloaded from the 'net onto one of my older Macs' hard drives.

 

[David Cook]

Any ideas what might be dead on my LC520 motherboard? Identifying that would potentially lead to a fix.

It sounds like you have short between +5V power and ground. Do you have any tantalum caps?

 

[JDW]

It sounds like you have short between +5V power and ground. Do you have any tantalum caps?

The LC520 motherboard was recapped by me using these Polymer Tantalum capacitors.

 

[JDW]

When I tested my LC520 motherboard inside the LC575 body, I had my external BlueSCSI v1 attached, but its LEDs never lit up.





I tested that BlueSCSI just now with my LC 575 motherboard inside my CC Mystic and the blueSCSI lights normally and it’s drives mount on the Desktop just fine, so it undamaged.

 

[David Cook]

With the motherboard removed, do you measure a short between +5V and gnd on the motherboard?

 

[JDW]

I put my multimeter into continuity check (beeper) mode and put the probes across each polymer tantalum capacitor. No continuous beep, which means they are not shorted. Then on the solder side, I checked the factory solid tantalums. No long beeps on any of those capacitors either.

Separately from that testing, here's a photo I shot in the past...

A single gold-finger (pad) connects tp +12V, but there are two pads for GND and another two for +5V. I used my meter in beeper mode to confirm the two GND pads are connected (as they should be), so are the two +5V pads, and GND & +5 are not connected, neither is +12V connected to GND or +5v.

And so I'm rather stuck trying to figure out what happened and what's wrong with this LC520 board.

And no, there are no noticeable burn marks on the stock solid tantalum capacitors on the solder side of the board. But if any of those had failed, they would have failed shorted, and the failure mode is to burn for those caps, and yet, there are no burn marks on them.

 

[Fizzbinn]

I have a Performa 575 and have successfully used both 520 and 550 logic boards in it with no problem, boards that I normally use(d) in my Color Classic. I don't think it's a fundamental incompatibility but I see how it could be some issue with your particular 575 chassis "killing" logic boards. :-/ I'd be hesitant to try any other logic board in it too.

In reading your testing summary could it also be possible that your 520 board died between the last time you last used it in your Color Classic and when you tested it in the 575? I've definitely encountered boards that work after a recap and clean and then won't work when pulled out later. IPA loosened leaked cap goo resettling after storage? I know in at least one case of mine I got the board working again after an ultrasonic cleaning. I also think I had an LC III board where it stopped working and the fix was desoldering, cleaning under and resoldering the CUDA and/or sound chip. There could also be accidental board damage from handling, it doesn't take much to damage a surface trace in handing. My 550 board was dead even after recap when I first got it until I repaired a knick on trace I initially did not see. Sorry, its likely this won't be much help.

In any case I feel your pain, this kind of thing is absolutely no fun.

 

[JDW]

I sadly didn’t re-test the LC520 motherboard inside my CC Mystic immediately prior to testing that motherboard in the 575 chassis, but I did test the 520 board many, many times after my recap inside the CC Mystic and it worked perfectly every since time. The last test of it inside the CC was a couple months ago. It's a very clean board and that was true even prior to the recap. That makes it all the more unsettling for me.

 

[JDW]

Whatever killed my LC520 motherboard didn't harm the Apple-branded 320MB spinning platter drive (Quantum LPS340S, which originally came with the LC575 and was connected at the time I tested the LC520 board). I put that HDD inside a PowerMac 9600 this evening and it booted the machine right away into System 7.5.5. That has some positive meaning since SCSI drives get 5V and 12V rails, and if one argues a power supply spike in the LC575 killed my LC520 board, why then was the HDD not affected?

I measured the RESISTANCE between the GND & +5V rails at the power connector on the LC520 board. It measured 1.18kΩ. Of course, I never measured that before so I don't know what is the "normal" resistance. Maybe one of your LC520 motherboard owners could check that, if you have the time and desire.

I then soldered a wire to the GND rail and another wire to the 5V rail so I could feed power to it via my bench top power supply, which has a good crowbar (short-circuit) protection feature, but it never engaged.



With the LC520 board connected, it draws about 1 Amp, which I think is about what it probably should be...



(I did not connect +12V to the motherboard — only +5v and GND.)

I covered the shiny metal parts with black electrical tape so I could take a better thermal scan...

I left it powered about 20 seconds and shot this with my thermal camera, showing hotspots that make sense...

21.7°C was about what the ambient room temperature was at the time.
51.2°C is hotter than the CPU by a little bit. This chip...

But maybe that's normal. Not sure because I never thermal checked that motherboard before. But 51.2°C really isn't hot. Nothing abnormal there, I don't think.

The solder side shows these warm spots, and by no means hot at all...

Still not sure what to think at this point.

 

[David Cook]

I agree with @Fizzbinn that it is more likely that something simple went wrong with the board in storage (or when being jostled during insertion) as opposed to the 575 frying the board.

With power off, but a battery installed, check pins 3 and 4 of the CUDA with an oscilloscope to make sure the crystal is running.

Also check pin 24 to see that it goes low when the keyboard power button is pressed.

I'd like to direct you to which CUDA pin triggers power up, but unfortunately I don't know. With +5V applied to the board, maybe pin 26 goes low?

Might be worth pulling the FPU and memory stick -- just to remove variables. I'll bet they both work fine -- which would further suggest something simple on your motherboard rather than 575 damage.

You could pull, clean, and reseat the ROMs. Bit of a longshot there.

 

[JDW]

With power off, but a battery installed, check pins 3 and 4 of the CUDA with an oscilloscope to make sure the crystal is running.
Also check pin 24 to see that it goes low when the keyboard power button is pressed.
…which CUDA pin triggers power up… With +5V applied to the board, maybe pin 26 goes low?

I appreciate your excellent advice. My 520 board and scope are at the office so I’ll check tomorrow, but for now the CUDA appears to be the 28-pin IC shown top-left in the photo below.

 

[BigBadBiologist]

I believe David is correct about looking at the EGRET/CUDA. In my experience from boards of this general era, no power = bad CUDA. Check to make sure the 5v standby voltage and 32KHz clock are getting there. If they are good, I’d wager dead CUDA.

 

[JDW]

I believe David is correct about looking at the EGRET/CUDA. In my experience from boards of this general era, no power = bad CUDA.

You've done a lot of board repair on your YT channel, so I really appreciate your kindness in making time to offer me feedback.

If the CUDA is bad though, finding a replacement chip may prove difficult because it could only be sourced by removal from another logic board, correct?

 

[BigBadBiologist]

DosFox came up with a way to program replacements, but finding the blank chips is difficult and the programmer is special. I don’t know about the 520, but the 575 CUDA can be replaced with ones from 630/5200/6200 type machines. Sometimes you can find battery-bombed boards from those pretty cheap (e.g. I got a nasty 5200 board at VCFMW for $1 that I used to save a LC475).

 

[David Cook]

If the CUDA is bad though

It might just be a bad joint (your solder joints seem good). I've had capacitor gunk get under a leg and eventually disconnect when exposed to mechanical stress or temperature changes.

 

[JDW]

With power off, but a battery installed, check pins 3 and 4 of the CUDA with an oscilloscope to make sure the crystal is running.

View attachment 26784

View attachment 26785

@David Cook
CUDA chip PIN-3 scope measurement (LC520 board with PRAM battery only)...



CUDA chip PIN-4 scope measurement (LC520 board with PRAM battery only)...



CUDA chip PIN-24 scope measurement (LC520 board with PRAM battery only, IIgs Keyboard Attached)...

Keyboard Power Key NOT PRESSED:


Keyboard Power Key PRESSED:


You said you don't know "which CUDA pin triggers power up," so I connected +5V and GND to the LC520 board via the two wires I soldered at the main power connector, and I left the PRAM battery installed, and I connected my scope's GND to main power GND, then I probed all 28 pins of the CUDA chip while pressing and reading the IIgs keyboard's Power Key. Only Pin-24 that I mentioned above changed from HI to LO. No other pins changed state.

After that, I removed the VRAM & RAM SIMM, then inserted the board back into the LC575 body, connected everything including keyboard, flipped ON Power, heard the CRT circuit charging "boom" sound, waited a few seconds, then pressed the IIgs Keyboard's Power Key. Like before, nothing. I then removed the motherboard, installed the PRAM battery, reinstalled and re-tested, but still nothing.

I then checked the F2 fuse on the motherboard, but it tested fine (I measured continuity across the fuse):



Hmmm....

 

[JDW]

I'm not testing the LC520 board inside my CC Mystic again. I removed the RAM and VRAM SIMMs, as well as the socketed FPU. I installed the PRAM battery after that and tested. Same as before. I flip the power switch, hear the CRT circuitry power up, then I press the Power Key on the keyboard, and the machine comes alive (case fan turns on and hard drive spins up), but there is no bong, CRT never lights up, no booting, and LED on front/left of the CC does not light. Same as before.

Of course, the LC575 body is different with the LC520 board in that there is no sign of life at all in that machine after pressing the Power Key (no case fan spinning, no hard drive spinning either). Oh, and when I tested the LC520 board tonight in the LC575 body, I removed the SCSI drive, but that made no difference.

I then took the FPU I removed from the LC520 board and put it into my stock CC motherboard, then tested. Booted just fine. I then ran TattleTech which confirmed it had a hardware FPU. After that, I ran Norton System Info's FPU benchmark, and finally I ran TechTool Pro 2.5.x and tested the FPU to confirm it's working 100% without error. In other words, whatever is wrong with my LC520 motherboard didn't harm the FPU. And had it be a voltage spike or over voltage that killed the motherboard, I must assume it should have also killed the FPU too, and yet the FPU is working. So that lessens the likelihood of "excessive voltage" being the cause of the LC520 MB's death.

After that, I removed the 32MB RAM SIMM from my LC575 motherboard and installed the 32MB RAM SIMM that I removed from my LC520 board. It booted up right away, and I used TechTool Pro 2.5.x's memory checks to confirm all is well. So once again we see that whatever killed my LC520 board didn't harm the RAM SIMM.

 

[JDW]

Last but not least, I tried my TechStep. I tried to enter TestMode but the TechStep responded saying ADB communications failed. Sadly, I had the same issue when I tested my fully working stock CC motherboard, as mentioned in my video here. Sadly, I simply cannot remember if I tried to test the LC520 board at that time, and I didn't seem to include that test attempt in my video either, so I have no idea if the LC520 board, even when working, has the same "TestMode" issues as the stock CC motherboard! Drat!

Well, after ADB communications failed, the TechStep asked me if I wanted to try SCSI communication, and I proceeded. That didn't work with my stock CC motherboard (as you saw in my video), but it works with the LC520 board!

Here are all the TechStep tests I ran on the LC520 board after enabling TstMd via SCSI:​

1. Power Supply: 4.98V (normal)
2. Power Up Voltage 5.01V (normal)
3. TstMd: Enabled via SCSI
4. CPU ID: Unknown / MISMATCH, but it also shows: ROM Version: $67C, Subversion: $25F1
5. Logic

• ROMck: ROM Checksum PASSED
• RAMsz: Soldered 4MB (correct because I have no RAM SIMM installed)
• Addr.: Address Bus Test PASSED
• Data: Data Bus Test PASSED
• RAM: Standard Test PASSED
• VIA: PASSED
• RBV: RBV Registered FAIL : CPU mismatch (CPU Expected:LC 520, CPU found: Unknown)
• VRMsz: Video RAM Size 512KB (SOLDERED)
• VRAM: Video RAM Test PASSED
• CLUT: Color Lookup Table PASSED
• Clock: Clock Test PASSED
• P/RAM: Non-Destructive Test PASSED, Destructive Test PASSED
• ADB: ADB Test PASSED
• SCC: Serial Comm. Test PASSED
• SCSI: SCSI Logic Test FAIL (Data Xfer)
• SWIM: PASSED
• Sound: PASSED

6. FPU: FAIL (I had the FPU chip removed, so that makes sense)
7. Built-in Video

• Turn PwSvr ON: "Screen should be solid black" (my comment: well, it is because it never turns on!)
• Turn PwSvr OFF: "In 10-30s, screen should turn gray" (my comment: I hear the circuit power, but the CRT remains all black.
• Video Patterns (test): nothing works because the CRT remains black

8. PDS Card ID: None Found (correct because I have no card installed)
9. More RAM Tests: I did the Extended Test and it PASSED (I didn't do the "Long" test because it takes 22 minutes)
10. ADB Status: PASSED
11. SCSI Functions:

• Termination Power: 4.76V (Normal is >4.15V)
• SCSI Term Check: PASSED
• SCSI Bus SCAN (Let's you type any SCSI ID, but the TechStep is ID6. I tried all others but the internal HDD was not found. Interesting.)
• Power OFF CPU: Command Worked. Case Fan and HDD powered OFF.
• Power ON CPU: Command Worked (without any bong). Case Fan and HDD powered up.