X68030 Compact Troubleshooting

[BlueBMW]

Alright, I need some help from you really smart guys!  I have two X68030 compacts that I'm trying to get running..

System A:
All capacitors on mainboard + sub boards + power supply have been replaced.
System powers on, shows 68030 bios screen.  Asks for a disk and flashes the light on Floppy 0, but will not load anything.  Any valid disk it will just eject.  It will seek an external hard drive, but I havent made a bootable hard drive yet to test that end of things.

System B:
Nothing has been touched on this system.  It powers on and loads from floppies ok but has no sound (i suspect capacitors)

Right now I'm working on system A....  I have swapped every component short of the mainboard from the other system to this one.  Nothing has fixed the no boot problem.  I have tried to clear SRAM but still no change.  Any suggestions on something else to try?

[kamiboy]

If you swapped in the working floppy drives from the other system and still got no boot then the problem should be with the main board.

A couple of things to try. First try and swap the ribbon cables connecting the parts to the motherboard from the good machine to the bad one. Maybe there are broken traces in them.

If that does not help then you might need to take the main board out and go over it looking for traces of corrosion.

Look for any exposed solder that looks dull, if you find any you are likely to find all the dull ones are concentrated around several hot spots. That is where capacitor fluid has leaked.

It might be a good idea to go over there solder points with a hot soldering iron and reflow them. That might reconnect broken joints.

The rest you know I think. Might be a good idea to inspect your recap job as well, just in case. Also test that fuse on the external drive daughter board to see whether it is good or not. I doubt that is the culprit, I think your machine would refuse to even start if that on was broken.

[lydux]

After what Kamiboy said (checking your recap job, especially around the FDD data separator : SED9420)
Here is a deeper inspection guide :

System A:
Assuming cable and floppy drives are corrects, it will be hard to identify without some logic probing and software debugging...
What about booting from the second floppy drive ?

The disk ejection and leds blink notify at least :
- A media detection
- Working IO Controller (PEDEC)
- Working FDD buffers

Media ejection signal raise on wrong data read, so the problem comes from the sub-board for sure, around the FDC (IC28 : UPD72065) or data separator (IC9 : SED9420).

Here is the sub-board schematics from your amazing x68030 service manual :
http://nfggames.com/x68000/Schematics/CZ-147%20-%20X68030%20Service%20Manual/CZ-147-087-088.TIF

You also have a screen, meaning the data bus (IODB0 ==> IODB7 signals) on sub-board probably doesn't suffer from broken traces.

So, I would suggest you to check first for traces continuity on these FDC pinouts :
- Between FDC and data separator (WINDOW, RDATA, MFM, WCLK)
- FDC control and interrupts lines to IO controller (CS, RD, WR, RESET, A0, INT)
- FDC DMA control lines to mainboard DMAC (DRQ, TC, DACK)


System B:
Easier one ! But need more infos :

Which sound part does not play ? (ADPCM, FM or both)
Which output does not produce sound ? (Main speaker, non-amplified output jack, headphone, all)

Remember that the PSU provides a +12v and -12v rails for powering sound amplifiers ! Be sure of these values.

[BlueBMW]

Wow, thanks for the great info.  Ill dig into it more tonight after work.

System B I havent done any work on, Im mostly using it to swap parts for now.  Ill recap it and see what happens with the sound after that.

As far as system A, both drives just eject disks upon initial read.  I have swapped ribbon cables, that rear connection sub board etc.  The only thing that remains is the mainboard itself.  Ill go over it really good tonight and check those xonnections you mentioned.

[BlueBMW]

Alright so far I've checked the following:

- Between FDC and data separator (WINDOW, RDATA, MFM, WCLK)
All tested good between IC28 and IC9

- FDC control and interrupts lines to IO controller (CS, RD, WR, RESET, A0, INT)
CS:
IC28:38 to IC29:98 and IC30:1 Good

RD:
IC28:36 to IC29:31 Good

WR:
IC28:37 to IC29:30 Good

RESET:
IC28:35 to IC32:21,  IC29:23 and IC30:33 Good but to IC27:7 did not pass continuity test, schematic shows IC99 in between?

A0:
IC28:39 to IC29:87, IC30:4 and IC27:9 Good

Int:
IC28:3 to IC29:34 Good

- FDC DMA control lines to mainboard DMAC (DRQ, TC, DACK)
I am having trouble locating these lines to test with.  Can you tell me which IC / pin numbers to which IC/ pin numbers?

[caius]

- FDC DMA control lines to mainboard DMAC (DRQ, TC, DACK)
I am having trouble locating these lines to test with.  Can you tell me which IC / pin numbers to which IC/ pin numbers?

These signals change their names in  FDREQ, DONE, FDACK and they go to the mainboard so look at other schematics.

[lydux]

Ok, don't bother about the RESET signals between IC27 and IC99. This is the reset handling for SCSI controller which is out of subject actually.

So, FDC <==> IO control bus wiring seems good.

About DMA lines : they are buffered and like Caius said signals names differs from FDC pinouts names. These signals are wired directly to the DMAC (IC92 = HD63450) and the system bus arbiter (IC59 = SAKI) on the mainboard, so test for continuity with both boards connected to each other.

Here signals list to check for :
- DRQ (FDREQ) :
   * IC28:49 --> IC99:3
   * IC99:4 --> IC92:4 (DMAC on mainboard)
- TC (DONE) :
   * IC28:51 --> IC99:12
   * IC99:13 --> IC105:3
   * IC105:2 --> IC92:24 (DMAC on mainboard)
- DACK (FDACK) :
   * IC28:50 --> IC105:1
   * IC28:50 --> IC59:41 (SAKI on mainboard)


I forgot also some others signals to check for between the data separator and the FDC (synchronization part) :
* IC28:3 --> IC33:13
* IC28:13 --> IC33:112
* IC33:11 --> IC105:12
* IC28:49 --> IC105:13
* IC105:11 --> IC9:5
* Clock line, IC28:4 --> IC9:3

References :
Bus masters (CPU and DMAC) : http://nfggames.com/x68000/Schematics/CZ-147%20-%20X68030%20Service%20Manual/CZ-147-075-076.TIF
System bus arbitration : http://nfggames.com/x68000/Schematics/CZ-147%20-%20X68030%20Service%20Manual/CZ-147-079-080.TIF


Sorry, there are many signals to test, but this is one very complicated part...
Next will be floppies to FDC wiring.

[caius]

System B:
Nothing has been touched on this system.  It powers on and loads from floppies ok but has no sound (i suspect capacitors)

Yes, you can change capacitors but check  also the op-amps chips (MC3403 or TL074 or LM324 compatible)

[BlueBMW]

System B:
Nothing has been touched on this system.  It powers on and loads from floppies ok but has no sound (i suspect capacitors)

Yes, you can change capacitors but check  also the p-amps chips (MC3403 or TL074 or LM324 compatible)

I will dig into this machine once I get this first one done.  I'm pretty sure system B is repairable without much trouble.  Sound problems are usually not a big deal to track down.  This booting problem on the first one is awful tricky though!

Doing some more testing tonight.  I did real quick try and boot from an external hard drive, still no go.  It still just goes straight to the floppy drives and flashes the lights and spits disks back out at me.  Checking traces now.

Oh and a pic of the total ghetto setup for testing:

[BlueBMW]

Here signals list to check for :
- DRQ (FDREQ) :
   * IC28:49 --> IC99:3
   * IC99:4 --> IC92:4 (DMAC on mainboard)
- TC (DONE) :
   * IC28:51 --> IC99:12
   * IC99:13 --> IC105:3
   * IC105:2 --> IC92:24 (DMAC on mainboard)
- DACK (FDACK) :
   * IC28:50 --> IC105:1
   * IC28:50 --> IC59:41 (SAKI on mainboard)


I forgot also some others signals to check for between the data separator and the FDC (synchronization part) :
* IC28:3 --> IC33:13
* IC28:13 --> IC33:112
* IC33:11 --> IC105:12
* IC28:49 --> IC105:13
* IC105:11 --> IC9:5
* Clock line, IC28:4 --> IC9:3

I checked all those traces and all have good continuity.  The hunt continues!

[kamiboy]

Your ghetto setup brings a smile to the old face. It look exactly like my system looked for a good month and a half, complete with random articles stuck underneath each part for balance and insulation.

That black DB15 to HD15 adapter looks sleek. Where did you pick that up? All I could find were these hideous grey looking ones, and I got one with the wrong gender on each side so now it is this foot long tumor sticking out of the ass of my Compact.

It really bothers me. So much that I tried to do my own custom alternative, but I found the image quality suffered a bit and had to switch back.

[BlueBMW]

Haha, yeah there really is no good way to run a compact out of its shell.

As far as the adapter... that one came with the XPC-4 I bought.  Pricey little bugger but worth it in my opinion

[BlueBMW]

Alright, some news here...

System A:  Still no idea what's wrong.  Everything seems to be checking out ok.  Any interest in trying to repair it Lydux?

System B:  I went ahead and removed all the capacitors, washed the board and replaced all the capacitors.  Powered it up, loaded a game, no sound at all.  After a bit I noticed a smell like hot electronics.  Looked around for smoke or heat, couldnt find anything so I started touching each component.  Both MC3403 amp chips were too hot to touch.  I removed them both and cleaned underneath them.  Almost all the pads lifted when I removed one of them.  I replaced both amps and repaired / jumped corroded traces / pads.  Powered the system on again,  the first amp stayed cool to touch but the other one (IC7) got burning hot again.  This time however I have FM audio only (no PCM it seems at the time) and it is quieter on the left channel.

Any ideas what to check that might explain why IC7 is burning up?

A pic of the IC that is not burning up (excuse the messy solder job):

[BlueBMW]

Looking at this schematic for the XVI compact (which should have a very similar audio circuit)




I'm wondering if there is something going on with those two transistors in that schematic.  If somehow they are shorted and feeding voltage back to IC7 somehow.

It looks like IC7 handles PCM amplification and additional FM amplification.  I could see that some FM sound may pass by IC7 through those two resistors and that is probably what I'm hearing coming from the system.  I'll check voltages on each pin of the op amp tomorrow.

[lydux]

It looks like IC7 handles PCM amplification and additional FM amplification.  I could see that some FM sound may pass by IC7 through those two resistors and that is probably what I'm hearing coming from the system.  I'll check voltages on each pin of the op amp tomorrow.

That's right. IC7 serves as ADPCM amplifier (both output and line-in) and FM+ADPCM mixer.

I'm wondering if there is something going on with those two transistors in that schematic.  If somehow they are shorted and feeding voltage back to IC7 somehow.

I also heavily suspect those 2 transistors. Q23 and Q24 should normally used as soft-shutting down ADPCM channel via the PPI.
You can remove them safelly for testing, see if adpcm sound get back.

But it does not really explain why IC7 amplifier get hot... So for now, just let's see without those 2 transistors.


Reference schematic :
http://nfggames.com/x68000/Schematics/CZ-147%20-%20X68030%20Service%20Manual/CZ-147-087-088.TIF
The whole audio part is located on bottom left. ADPCM is IC32 (MSM6258), and the pin to follow is 22 (DAOUT).


PS : As for your first system, I've never get into x68030, so yes I am. Please PM me.

[BlueBMW]

Did some more investigating...

I checked the voltages on the various pins of IC7 and I found one pin not correct.  Pin 11 or Vee should be -12v if I'm reading the specs right.  I'm getting a +1.3V on that pin.  Not sure if that would be the cause of the high temp and failure, but I could see it causing the amp to not operate correctly.  I checked IC36 to compare and it has a solid -12v on pin 11.  Both pin 11s get -12v from VCC4* according to the schematic.  So somewhere I have a problem with the -12v supply to IC7.  I'm attempting to track it down now.

UPDATE:

I traced the VCC4' line from IC36:11 towards IC7:11... about halfway from IC36, the line goes in a via that appears to be an internal layer connection as it is not visible on the bottom of the board.  I traced the line back from IC7:11 and found it also to go to a via that seems to go to an internal trace.  I didnt have any continuity between these two vias.  I soldered in a jumper wire to reconnect this trace.  The result....  No more heat on IC7 and I get full loud PCM sound!  Only problem now, I seem to only be getting right channel FM.  I'm going to do some probing and see what I can find.  I wonder if the high heat destroyed part of IC7.

UPDATE 2:

Traced the left channel all the way from the AO port, through IC7 and IC36... turns out there was a bad via in the trace between IC36:12 and IC37:9 (the YM3102)  Jumped the trace and restored left FM sound!

So as of right now this x68030 is fully functional again!!!

I'll post some pics in a bit of all the beautiful (UGLY) jumper wires and solder bridges I had to do in order to fix it.


Pics:

Right after removal of the first Op Amp... the pads lifted so easy



After all the wire jumpers etc were finished:




And a quick youtube vid of the final results...  Excuse the messy work area / shop.

http://youtu.be/PI4USowBq9E

[eidis]

Good Job BlueBMW !!!

Just out of curiosity, how much watt does your soldering iron has and which method did you use to remove the op-amp ?

Eidis

[kamiboy]

A soldering iron is not the proper tool for removing multi legged SMD's.

You need a heat gun for that.

[caius]

Good Job BlueBMW !!!

Just out of curiosity, how much watt does your soldering iron has and which method did you use to remove the op-amp ?

Eidis

Hi Eidis.
This was the point.I think BlueBMW lifted the traces because he used a soldering iron in order to desolder the op-amp.He should have used an hot air station instead , it would have been a neat job.

[BlueBMW]

Hehe busted!!

Actually Im waiting on a part for my hot air station, so I had to resort to the iron for the amp replacement.  I got one of them done no problem, but the other one the pads lifted as soon as I put any heat to them.

For the record, I think my iron is 65 watts, I had it set on 400C for soldering and about 550C for desoldering with braid.

[eidis]

Hi Guys !

Just a few cents from my personal experience. I am afraid of using heat guns and hot air stations in fear of damaging the chips so I use a slow, tried and time proven method. Soldering iron can be used in cases to remove SMD's, it just takes more time. I use either 8W or 16W soldering iron, depending on which one is enough to melt the solder, to avoid de-laminating the traces and do not heat the legs for more than four seconds. Lukemorse inspired me with the wire under the chip trick. Basically you need to slip a stripped kynar wire under the chip and solder one end to some sturdy component nearby, then heat the legs one by one and gently pull the kynar wire so that the pin lightly lifts and detaches from the trace. Tried replacing RAM chips on a SNES and this method worked like a charm.

P.S. Yes, the SNES is still working

Can anyone share their experience with using hot air guns or hot air stations ?

Eidis

[BlueBMW]

I have never gotten to try my hot air rework station since it was missing some parts when I got it.  I ordered the parts, but they're taking a while.

Next x68030 question.....

Are there any known issues with MIDI on x68030 machines?  I've been trying to get MIDI to work on this system, but I cant seem to get it to work...  Games seem to recognize that a MIDI board is installed (Mahou loads up in MIDI mode for sound) but I am not getting any midi data out of the MIDI board.  I've tried both a Sacom 68M and 68M-2 with no results with either.  I also tried swapping the IO slot board just in case that was the issue.  The fact that games are seeing the boards tells me that its interfacing ok.  Any ideas guys?