XVI Compact FM audio

[RobIvy64]

OK, so after a complete capacitor overhaul, the system works great except for one problem that still remains: FM audio is horribly distorted.

PCM samples sound normal (explosions, etc), but the FM synthesizer is quiet and very distorted, almost like radio station static.

Before the cap replacement this problem existed, however all audio was quiet.

I used new Nichicon tantalum and radial electrolytic capacitors.

Any ideas?

[RobIvy64]

Could it be a bad op-amp?

[lydux]

Hi !

Yes, if you have changed all caps, it could be a bad amplifier on op-amp, or the dual channel DAC (YM3012).
Could also be a power regulation problem on amplifier part. You could check this with a voltmeter on op-amp (pin 4 and 11).
They should be +12v and -12v (or +5v and -5v, I don't remember).

Good luck !

[RobIvy64]

Hey, thank you for the reply!

Is there a separate op-amp for ADPCM and FM?

EDIT: Looking at a few X68000 compacts for sale on some JP sites, they list the units as having capacitors and op-amps replaced. I just ordered some MC3403 OP-Amps and will report back once I replace them.

[lydux]

From what I can see here : http://4.bp.blogspot.com/-dwHOx7Cyxxo/TaBjIl-_Y2I/AAAAAAAABKc/x39yKxP6G5Q/s1600/rz3.jpg

There is 2 op-amp (ic39 and ic7). The one you have to check is located near to the left of YM3012 (ic39).

[RobIvy64]

Awesome. The chips are en-route and will be here in a few days.

Apparently this is a common issue for the '030 and XVI compacts. Thanks for your help.

[RobIvy64]

Alright consider me officially stumped.

I have replaced the following:

2 x MC3403 op-amps
YM3012 DAC
YM2151 FM synth

No change in the issue. 

[RobIvy64]

Here is the first stage of "Dive On!" as heard on my X68000

Normal music:
drive on x68000 x68k part 1

Distorted Music:
X68000 audio issue

[eidis]

 Hi!

This reminded me of Tiido's fantastic Crystal Clear audio mod which he made to improve the sound quality of SEGA Genesis/Mega Drive. X68000 is not exactly genesis, but they are a little bit similar. He is very friendly so you could try asking for his advice directly. Check out this thread for the schematic:

http://nfggames.com/forum2/index.php?topic=3719.0

I will try to make wild guesses and my knowledge in electronics is at a very basic level so please bare with me if I will say something stupid:

1) Could it be that the cause of distorted sound is faulty low-pass or high-pass filter ? If I recall correctly, in Tiido's schematic those were realized with tantalum capacitors. Are you willing to try replacing tantalum capacitors as well ?

2) Could there be a broken trace or short circuit somewhere ?

Here are some other things to consider:
http://www.mediacollege.com/audio/howto/distortion.html
http://en.wikipedia.org/wiki/Linear_filter
http://en.wikipedia.org/wiki/Low-pass_filter
http://en.wikipedia.org/wiki/High-pass_filter

Eidis

[eidis]

 Just found a very good article on repairing arcade PCB boards. Check it out, it might give a few valuable hints:

http://www.brasington.org/arcade/tech/repair.shtml

The main thing which caught my interest was:

Other things to check for while you are inspecting the board include burnt edge connectors, burnt components, broken components such as capacitors and resistors, and of course missing components. Note it is very common for boards to have unpopulated chips, so don't assume any empty socket means a missing component. Check the schematic. Also, TTL designs require that there be a capacitor (usually a disc capacitor) every so often (sometimes as often as each chip), and they need to be nearby each chip. Anyways, it is common to see one or two of these capacitors broken. I usually don't worry about this right now, and usually don't replace them. However, for broken disc capacitors, I usually just remove the entire capacitor and mark the area to remind me that I removed it. This prevents a possible short-circuit which can occur if the broken disc capacitor happens to have the metal plates inside it touching each other.

So I am guessing that the distorted sound could be due to broken disc capacitor.

This site might give some valuable hints as well:

http://www.mikesarcade.com/arcade/repairs/index.html

Best of luck and keep us posted!

Eidis

[RobIvy64]

Thanks for the tips! I haven't had a chance to work on this for a few days, but hopefully this weekend. I will keep this thread updated.

[RobIvy64]

Here are the voltage readings i'm getting on the MC3403 OP-AMPs (at idle).

Pinout


MC3403 for sound effects (working)
Pin #      Reading (V)
-----------------------
1 ..........3.78
2 ..........2.28
3 ..........2.54
4 ..........5.06
5 ..........2.55
6 ..........2.55
7 ..........2.65
8 ..........2.42
9 ..........2.42
10 ........2.03
11 ........0.00 (GND)
12 ........2.54
13 ........2.55
14 ........2.64

MC3403 for music (YM2151/YM3012, NOT working)
Pin #      Reading (V)
-----------------------
1 ..........0.08
2 ..........0.08
3 ..........0.00
4 ..........5.06
5 ..........0.06
6 ..........0.56
7 ..........0.56
8 ..........0.14
9 ..........0.14
10 ........0.10
11 ........0.00 (GND)
12 ........0.03
13 ........0.55
14 ........0.55

As you can see, we have a huge difference in voltage levels here. Seeing as how my input voltage is 5.06, i'm thinking this op-amp might be bad. However, it is brand new. Any ideas? :?

[eidis]

Please post clear high-res pictures of your motherboard. That way we might get a better idea of what is wrong with it. Meanwhile I found some links which might help.

http://www.bcae1.com/repairbasicsforbcae1/repairbasics.htm
http://www.bcae1.com/opamp.htm

The symptoms look like they could be linked to broken trace, short circuit, bad tantalum capacitor or bad transistor.

Just found another site on arcade PCB repair:
http://forum.arcadeotaku.com/viewforum.php?f=46

Best of luck.
Eidis

Update: Sadly this system is now being sold on eBay. Guess we'll never know what was wrong with it unless the new owner will come to this forum.

[BlueBMW]

He sold it to me semi-locally.  So the diagnosis will continue!

[eidis]

 Great news ! So we might find out after all what was wrong with it. Might I suggest inspecting tantalum capacitors and transistors ? I read that it is possible for corrosion to get under solder mask and eat up the traces so continuity testing could help too. In any case, please post your findings.

Best of luck !
Eidis

[lydux]

Great news ! So we might find out after all what was wrong with it. Might I suggest inspecting tantalum capacitors and transistors ? I read that it is possible for corrosion to get under solder mask and eat up the traces so continuity testing could help too. In any case, please post your findings.

Best of luck !
Eidis

Yes, that's possible ! I just have this case with an ACE that stt1 sold to me recently : because of battery leak on IO board, a trace from IOSC chip to main board has broken under the green silk mask and cause an SRAM failure at boot.

@BlueBMW : I don't know if you're going to build a schematics of the sound amplification part while searching for the problem. But if that so, it might be interesting to us since compact models seems to have many problems in audio section. As I haven't time to do this part (more concentrate on logic core), could you please post it somewhere ?

Good luck !

[BlueBMW]

It sounds like mapping out the whole sound circuit would be useful.  Like you guys said, I suspect its a transistor or one of the tant caps.  It only takes one component coming in and out to mess everything up.  Updates will come as I dig into things!

[RobIvy64]

Beemer, this sound issue was around before the re-cap FYI.

[BlueBMW]

Just a few updates here....

Did some poking around with a sound probe and a logic probe.

It seems that the YM2151 is sending a digital signal to the YM3012 and it's getting there... whether that signal is any good or not I dont know.  There is no output from the YM3012 chip though.  If I probe the data line between the two chips with my sound probe I mostly get static but I can also hear the music... very strange...

[lydux]

Hi !

Datas between ym2151 and ym3012 should be in a serial fashion. SD, SAM1 and SAM2 pins in phase with CLOCK1, each sampling starting on falling edge of SAM1 for one channel and SAM2 for the other, and should be 16 bits wide. SD is the datas signal.

Can you post a picture showing these 4 sniffed pins ? I can tell you if there are correct.

Also, I just remember that there is a RESET pin on YM3012 ! (pin 7 = ICL#). It should stay to a 1 logic on normal operation.

So, if there is a correct input data, no stucked reset input, and no output at all, maybe that's a sound bias current issue ? (Pins MP, BC, RB = 13, 14 and 15)
Or analog amplification ? (ToBUFF, COM = 12 and 11)
From here, we need a schematics... I'll try to draw it from an ACE dead I/O board. But should be slighty different from a compact.

Do you have an oscilloscope ? It will help us.  ( personnally, I don't have one -___- )

Good luck !

[BlueBMW]

Hi !

Datas between ym2151 and ym3012 should be in a serial fashion. SD, SAM1 and SAM2 pins in phase with CLOCK1, each sampling starting on falling edge of SAM1 for one channel and SAM2 for the other, and should be 16 bits wide. SD is the datas signal.

Can you post a picture showing these 4 sniffed pins ? I can tell you if there are correct.

Also, I just remember that there is a RESET pin on YM3012 ! (pin 7 = ICL#). It should stay to a 1 logic on normal operation.

So, if there is a correct input data, no stucked reset input, and no output at all, maybe that's a sound bias current issue ? (Pins MP, BC, RB = 13, 14 and 15)
Or analog amplification ? (ToBUFF, COM = 12 and 11)
From here, we need a schematics... I'll try to draw it from an ACE dead I/O board. But should be slighty different from a compact.

Do you have an oscilloscope ? It will help us.  ( personnally, I don't have one -___- )

Good luck !

An o-scope is about the only tool I have left to buy   If I cant figure this out with just a logic probe I guess I'll have to spring for the scope finally.

I was checking pin 22 of the YM2151 and pin 4 of the YM3012... I was getting a good pulse signal there, and when probed for sound I got mostly noise, but faint FM audio also.  I checked pins 9 and 10 on the YM3012 and got no sound.   I seem to remember also probing pins 5 and 6 and got pulse on 6 but nothing on 5.

Pardon my ignorance, I'm no expert at electronics, just a hobbyist.  I've only recently got into all this logic and data stuff   Up until now it was mostly replacing bad capacitors to fix things lol.

[lydux]

Your logic probe will only be usefull the left part of ym3012 (pins 1 to 7). These are digitals signals. The right part is analog out. So, we need an oscilloscope here.

I was checking pin 22 of the YM2151 and pin 4 of the YM3012... I was getting a good pulse signal there, and when probed for sound I got mostly noise, but faint FM audio also.

Ok, datas is coming, do you have a clock pulse with this ? (pin 2).
You will not be able to get sound here. As I said, it's all floating-point digital signals that's will be converted by the DAC.

I seem to remember also probing pins 5 and 6 and got pulse on 6 but nothing on 5.

You're probably facing a single channel sound then. That's ok.

I wonder if configuring the ym2151 in order to output a continous, static linear sound on both channel will probably help us...
I'll try to give you a simple program for this.

Also, please check for ICL# (pin 7) ! If stuck to a logic 0, it'll keep initializing both channels, so no analog output !

Pardon my ignorance, I'm no expert at electronics, just a hobbyist.  I've only recently got into all this logic and data stuff   Up until now it was mostly replacing bad capacitors to fix things lol.

Don't blame yourself ! I'm not here to judge you. I just want to help and share knownledges with this growing community. 
I'm myself an hobbyist. Like you, I've learnt by testing, trying to understand, (and sometimes destructing -_- )
And we will all learn from this sound problem ! 

[eidis]

 I really hope that BlueBMW will be able to repair his X68000. This will make a great Wiki article

Keep the scene alive !
Eidis

[BlueBMW]

Ok, I fired it up again today...  I probed the left 8 pins of the YM3012 chip.  Here are my results...

Test was done running "Dive On" basically the only game I've got hehe

Pin # :  Results
1 : High
2 : nothing
3 : Low
4: High Pulse
5: nothing
6: Low Pulse
7: High
8: Low

You mentioned pin 2 requiring a clock signal?

I checked Pin 23 of the YM2151 and got a pulse signal there....  but there's no signal on pin 2 of the YM3012... checked continuity between those two pins... open circuit...  jumped it, so now I'm getting a pulse signal on pin 2 of the YM3012... but still no FM sound as of yet.

Any thought's?


Also, I should mention...  the way I shown to "probe" for sound was to basically make a homemade probe with wire and a 1uf cap.  I drew this up to describe how it was made... originally I used to locate missing sound on a Turbo Duo.  It worked great for that setup since I was basically able to follow the sound all the way down the traces / components from its source chip until I found where it stopped.  On that Turbo Duo it ended up being an amp chip that wasnt getting power.  The sound went in but didnt come out.

If anyone thinks this is a "bad idea" please let me know  

Don't blame yourself ! I'm not here to judge you. I just want to help and share knownledges with this growing community. 
I'm myself an hobbyist. Like you, I've learnt by testing, trying to understand, (and sometimes destructing -_- )
And we will all learn from this sound problem ! 

I was just giving a heads up in case I use incorrect terminology or something   Working on these things is an excellent way to learn about electronics etc.  We're going to figure this one out though.... even if we rebuild the whole sound circuit on an external board hehe.

[BlueBMW]

Ok, I didnt get far probing etc, so now I'm working on mapping out the whole sound circuit as best I can.  To start with I removed all the caps and some of the chips that had been replaced.  I wanted to inspect for broken traces or corroded vias.  I have found several vias that look like they could use some work.

So for now, here are some pics of the compact's board before I removed the caps:















And the main sound area after I cleaned all the caps off and removed all the gunk etc:





And a list of the caps I removed and their values... (note, these are the values of the replacements, I dont know what the originals were)

c01 sub: .47uf 50v
c01: 1uf 50v
c02: 4.7uf 25v
c03: 470uf 16v
c04: 470uf 16v
c05: 470uf 16v
c06: 100uf 6v
c07: 47uf 6v
c08: .47uf 50v
c09: .47uf 50v
c10: .47uf 50v
c11: 4.7uf 25v
c15: 47uf 6v
c16: 470uf 16v
c18: 100uf 6v
c19: 100uf 6v
c20: 100uf 6v
c21: 100uf 6v
c22: 100uf 6v
c23: 47uf 6v
c24: 47uf 6v
c25: 1uf 50v
c26: 100uf 6v
c27: 10uf 16v
c28: 100uf 6v
c29: 4.7uf 25v
c30: 470uf 16v
c31: 4.7uf 25v
c33: .47uf 50v
c34: 100uf 16v
c35: 100uf 16v
c36: 100uf 6v
c37: 100uf 6v
c39: 4.7uf 25v
c40: 4.7uf 25v
c41: 4.7uf 25v
c42: 100uf 6v
c43: 22uf 16v?
c44: 100uf 16v
C45: 100uf 16v
C46: 100uf 6v
C47: 100uf 6v
c48: 1uf 50v
c49: 100uf 6v
c90: 47uf 16v
c348: 1uf 160v
c351: 1uf 50v

[lydux]

Hey wait ! What the hell is that ! All smd electrolytic caps were changed by tantalum ?!

There is nothing wrong with that in theory... But I know it can cause distorsions depending on frequency and voltage ! As an advice, you should go back to electrolytics.


Well, I'm currently drawing a schematics for audio part base on my I/O board. Actually, first stage is done : OPM=>DAC, bias, dac amplification and channels preamplification. First thing I notice is the power level used by opamp : they definitively use the 12v and -12v power rails (pins 4&11).
Could you confirm me that pins 11 of opamps are tight to GND ? If that so, twin towers models audio schematics is not applicable to compact... And I will disassemble mine (I don't really want, but if it's needed, I will do... just for giving you correct capacitors values).
And if not, then you got a PSU failure on -12v rail ! Which will cause distorsions.

[RobIvy64]

I don't see a problem with the tantalums.

It should be noted that this audio issue existed before the capacitor swap, which prompted me to recap the motherboard originally.

I could tell that some of the thru-hole caps had been replaced prior to my ownership, but the SMT aluminum caps had not.

[BlueBMW]

It is definitely a possibility that the power supply is to blame.  This power supply is just about shot.  I debated recapping it but then I test fit the PSU I used in the Expert and it fits perfectly in the compact.  Im going to replace it.

I started tracing the audio circuits back.  I ran into a bit of a dead end on the AO port side of it...  but Im making progress tracing back from the headphones port.  Before I reinstall caps I want to map out as much of the sound circuit as we can.  I also want to look into replacing all those transistors.  Ive got replacement DIP transistors but I need to order the SMT ones.

[lydux]

Here is a quick and dirty drawing of fm audio pre-amplification part from an ACE I/O board : Model K5710DE

Its' not finished yet : some signals to/from OPM are missing, adpcm mixer is not here and this drawing stop to the output connector located on the back (We will refer this connector as AO). Please, note also that transistors are DTC114 (NPN with built-in resistors), my symbol representation is wrong.
Components values may also differ from a Compact, as well as references names. But I'm pretty sure implementation is the same. So any value/names equivalent from Compact are welcomed.

Anyway, the important part is fully here !

Now, testing while fixing sound issues should be done using this AO connector, as it seems to be the closest to the YM3012.

If the problem does not come from the PSU (correct +12v and -12v rails) or opamps or broken traces, it should be located somewhere around here : DAC amplification (the part at the right of YM3012), and channels preamplifications. Relevant components : all ceramic precisions caps (C41, C28 and C29) ! These are rated at the lowest possible tolerance (B = ± 0.1 pF), this is really important.

Also, Q4 an Q5 may be removed for testing purposes (AND ONLY FOR TESTING !) with the use of AO connector. If the sound is correct then, these transistors may be dead, or the diode D7 could be shorted.


A word about the design : it seems to be a typical implementation of YM2151 and YM3012. I would like to try some improvment by changing both opamps with something some dedicated to audio amplification (like a TL074, or NJM2059). Changing C41, C28 and C29 by polystyrene type capacitors could also be a good idea.


Hope this help !

[BlueBMW]

Fantastic info there!  I've got my replacement power supply in, just waiting on the 7804 ICs I ordered to come in so I can do the swap.  I did find a few broken traces under the YM2151 chip, so I'm going to look over the board real good to make sure there arent any others that are bad.

[eidis]

 Thank you Lydux for the contribution. Now we have two files with schematics

BlueBMW,

Please check these schematics if you are willing to try resurrecting the original PSU:

http://nfggames.com/X68000/Misc/X68000_PSU_Schematics.pdf

Keep the scene alive !
Eidis

[BlueBMW]

SOLVED!

Thanks to the circuit diagram, I was able to check each and every trace of the sound circuit for the FM audio starting at the YM2151 chip.  I followed it all the way up to the Op amp (IC26) in that schematic (IC39 in the compact)  and found that both the left and right FM outputs from that op amp were cut.  Apparently there were some test points on the output traces and they had corroded away thereby breaking the trace.  I bridged them, recapped the unit again, powered it up and the FM sound has returned in all its glory!  The crackly popping nastiness of the other sound is also gone!

Here's a picture of said failure point.  Its the two spots circled in red to the left.  On the right are 4 traces that also seem to have suffered the same fate.  I'm not sure what their function is,  they terminate to vias at one end (but dont go anywhere from there.... maybe ground plane in the PCB?) and the other ends look like they go to some transistors.



And here's a terrible quality video that I recorded right after I first heard the good sound!  The speakers were behind me so it barely picked up the music, but darnit it works!

X68000 Compact Lives!

[BlueBMW]

And a few pics of my recap:



And the area of trouble up close....

[lydux]

Glad to hear that !

It seems you may have found a common issue on Compact models.
If you look to the picture in post #3, you'll notice this one have the exact same 6 points corroded.

On my board, I don't have this transistors block where these 4 lines point to.
It's probably not related to the audio block, but video.

Good job ! Have fun with your x68k !

[BlueBMW]

Glad to hear that !

It seems you may have found a common issue on Compact models.
If you look to the picture in post #3, you'll notice this one have the exact same 6 points corroded.

On my board, I don't have this transistors block where these 4 lines point to.
It's probably not related to the audio block, but video.

Good job ! Have fun with your x68k !

I didnt try jumping those other 4 traces yet, but I cant tell if anything is not working correctly.  Video seems great and so does the audio.  I suppose more testing will tell for sure!

Thanks all for the assistance!

[eidis]

 Thanks Guys for all your effort !

I consolidated this thread into a Wiki page. Please feel free to suggest which part needs to be changed or how to improve it. Check it out here:
http://www.gamesx.com/wiki/doku.php?id=x68000:x68000_xvi_compact_audio_troubleshooting_and_repair

Keep the scene alive !
Eidis

[BlueBMW]

Thanks Guys for all your effort !

I consolidated this thread into a Wiki page. Please feel free to suggest which part needs to be changed or how to improve it. Check it out here:
http://www.gamesx.com/wiki/doku.php?id=x68000:x68000_xvi_compact_audio_troubleshooting_and_repair

Keep the scene alive !
Eidis

Looks great!  The only thing that might be worth mentioning is that the component numbers are different on the compact versus the ACE.  The general layout and component values seem to be identical though.

Component numbers in red are for the compact XVI board I have here.  I didnt get every single component, but most of the major ones along the sound circuit.  Some of the traces seem to go through some middle PCB layer that I cant physically see.






it also looks like RM9 on the ACE is split up into separate resistors on the compact...
RM9 pin -- Compact
2 -- R99
3 -- R100
4 -- R81
5 -- R83
6 -- R82

[eidis]

 Thank you BlueBMW !

Schematics have been mirrored:
http://nfggames.com/X68000/Misc/X68000%20XVI%20Compact%20audio%20pre-amplification%20schematics/

And the article has been updated with the improved schematics:
http://www.gamesx.com/wiki/doku.php?id=x68000:x68000_xvi_compact_audio_troubleshooting_and_repair

I have a question regarding C28 and C29. Are they the same on Compact as well ?

Keep the scene alive !
Eidis

[BlueBMW]

Thank you BlueBMW !

Schematics have been mirrored:
http://nfggames.com/X68000/Misc/X68000%20XVI%20Compact%20audio%20pre-amplification%20schematics/

And the article has been updated with the improved schematics:
http://www.gamesx.com/wiki/doku.php?id=x68000:x68000_xvi_compact_audio_troubleshooting_and_repair

I have a question regarding C28 and C29. Are they the same on Compact as well ?

Keep the scene alive !
Eidis

I wasnt able to locate those and confirm their number.  If I didn't re write a red component number it means I didnt locate / check that component.  C28 and 29 in that schematic were buried in a pack of resisors and caps and I couldnt' follow the traces very well.

[lydux]

Hi !

Thanks for this wiki page and schematics update.
I have some corrections and improvements to suggest :

Needed tools : Ok for both Elenco logic probes (the 560 can capture logic signal up to 20 Mhz which is also the x68k's system bus frequency, it could be used for many others x68000 problems). But for audio part, the logic probe will only be usefull for checking the 5 digital signals between both YMs (ICL#, Clock1, SD, SAM1 and SAM2). For analog part, while an oscilloscope is more recommanded, a simple voltmeter will give more hints than a logic probe.
We have also to check for power rails and broken traces. So basically, a correct multimeter is required.

If that does not help, you could have a bad amplifier on Op-Amps (MC3403) or the dual channel DAC (YM3012). It could also be a power regulation problem on amplifier part. You can check this with a voltmeter on op-amp (pin 4 and 11). They should be +12v and -12v. If there is a PSU failure on -12v rail, it will cause distortions.

@BlueBMW, can you confirm this while your x68k is opened please ? I'm 100% sure of this for twin tower models, that's probably the case also for desktop models, but not for Compact (it can be +5v/0v, +5v/-5v or +12v/-12v).
Also, just for precision, these measurements have to be done with volmeter black lead on any ground (for simplicity, let's say the shield of any connector). The red lead on the pin to test.

Compatible devices are TL074 and LM324.

There is multiple references for TL074 in smd package. The exact needed one for Compact is TL074ACD and is preferred over LM324.

There are 2 Op-Amps:

    IC39 - Responsible for FM synth (Top)
    IC7 - Responsible for DAC (Bottom)

That's wrong :
IC39 is for OPM DAC internal buffer amplification and first stage of stereo FM channels pre-amplification.
IC7 is for second stage of FM channels pre-amplification, ADPCM output amplification and channel mixing for internal speaker (we haven't checked so far).
There is also a third opamp we haven't covered because it wasn't involved in the problem : IC106 (NJM2073). It's a dual opamp, used for headphone amplification.

Finally, I've built a flow chart from what I've learnt all over this thread. It covers most possible cases. Except for component references and voltages, it's available for all X68000 models. (Following components references are from Compact models). Fill free to add it to the wiki page.

Start by isolate the faulty block by testing each output, following the chain :
Internal speaker ==> Headphone ==> Back connector (A0)

No sound at all, on any output :

• Using a voltmeter, x68k powered on, check for power supply on opamps. If one present wrong voltage, check your PSU.
• Change opamp IC7.

Only Internal speaker and headphone present issues :

• Change opamp IC7 and volume potentiometer.

Only headphone present issue :

• Change opamp IC106.

Missing FM sound, on any output :

• Following the schematic, check for continuity between YM2151 and YM3012. Any broken trace will cause the DAC to not receive its data. So no output. Fix this by soldering wire between involved pins.
• Using a voltmeter, x68k powered on, check for power supply on opamp. If one present wrong voltage, check your PSU.
• Using the logic probe, x68k powered on, test for the following pins :
  
  
   
  • pin 7 of YM3012  and pin 3 of YM2151 should be high, if not that's IOSC/PEDEC issue.
  • pin 24 of YM2151 should pulse, if not that's IOSC/PEDEC issue.
  • pins 2, 4, 5 and 6 of YM3012 should pulse, if not change YM2151.

All outputs present unexpected FM sound behavior (distorsions, low volume, saturation, ...) :

• Using a voltmeter, x68k powered on, check for power supply on opamp. If one present wrong voltage, check your PSU.
• (Compact models only) Using an ohmmeter, x68k powered off, check for resistivity between pin 13 of IC39 and + pin of C31. Do as well between pin 7 of IC39 and + pin of C28. The indicated value should be < 100 ohm. If more or open circuit, the 2 test points near to IC7 pins starting to get corroded (see picture), and need to be cleaned. Soldering a simple wire between involded pins will also do the job.
• Change opamp IC7 and IC39.
• Remove transistors Q# and Q# and test using AO connector. If sound is ok, change Q#, Q# and D#. (#=Need Compact reference names)
• If the issue always remains, it's time to change some capacitors. Start by changing precision capacitors. Then, full recap the whole audio part.
• The DAC is faulty, change YM3012.

Missing ADPCM sound on all output :

• Using a voltmeter, x68k powered on, check for power supply on opamp. If one present wrong voltage, check your PSU.
• Change opamp IC7.
• We need more investigations here...