No sound on right sound channel (ADPCM fixed)

[Duo_R]

Charlie,

I jumped pins 4 and 5 on the IC504 and still did not hear ADPCM. I haven't replaced the tiny components you mentioned, can those go bad? I did reflash the solder on them but that didn't fix anything. All the large caps on this board was replaced, but all those tiny parts are still stock. Do you know the values of all those parts? Would it be bad to just temporarily jump them?

[Charlie]

Well, if you're not hearing it AT ALL, it has to be within that group of components mentioned.  That brings up a new bunch of questions, like:
1. Is ADPCM even being generated?  That is, are you SURE you are operating in a mode whereby ADPCM should even be there?  Can you check against a good unit?
2. Given the 5202 is the analog source (and you've replaced it - are you sure it, and any other new stuff, is actually any good?), you could check to see if it actually has any power on it, and also check to see if the digital data is actually coming into it. 
3. A quick trick, but I don't know if this will work:  many times, if an amplifier or ampliflying circuit is strong enough, you can inject a signal of your own using just the tip of a screwdriver (carefully) at the source, and hear a hum or some kind of noise at the output.  If you have a working unit to compare to, this makes the test easy, as you can compare the loudness of the two; it may help to use headphones.  The idea is to determine if the "screwdriver sound" is equally loud, or even close, in both units.  However, this trick only works if the amplifier can amplify the signal strongly enough; some amp circuits do only a minimum of amplification.

To answer your question, the "tiny components" are much less likely to go bad, but of course they are not 100% guaranteed.  It is possible two other things are the issue:
1. Within that entire amplifier circuit chain of components, there is enough leakage/corrosion on the board to kill the amplifing function,

or

2. The smaller (value-wise) caps, which typically don't actually leak, can short or open.  Either will kill the signal.

After-thought edit:  Are you sure the amp chips are getting power?

I would NOT recommend you jump them, at least not yet.  I will get more data on the specific parts and/or part-values.

Charlie

[Duo_R]

I am not fully understanding the screwdriver test. I am positive that it is not outputting adpcm, I have a good unit for cross compare.
Which pins should I test for power? I am stumped and not sure what to do next. Maybe swap IC504 next then small parts?

[Charlie]

U502 should have 5V on pins 1,3,18.
U503 should have power on pin 8.
>Edit: U503 should have 4 volts (I think - check against good unit) on  pin 8, and 2.5 volts on pin 3 and pin 5<

I assume you can find the Ground connection for yourself, maybe on the power plug?

Clarify this statement:
"I am positive that it is not outputting adpcm".  Do you mean you have determined positively that U502 itself is not GENERATING audio, or do you mean the final output of the whole amplifier chain has no audio?

Assuming you mean the whole amplifier chain has no audio (that is, you feel U502 is ok):
To do the screwdrive quick test, plug in the headphones.  With the unit on, carefully touch the screwdriver (actually, any reasonably stiff piece of metal that you can handle safely) to U502 pin 10.  Make sure that you are touching the metal of the screwdriver; that way, your body becomes part of the circuit, which is what we want.  Perform this test on the good unit first.  Also, of course, the volume should be full up. You should hear something in the headphones, either static, a hum or buzz, or at least a kind-of-click when you actually make contact.  We need to get this to work first, so do whatever is normal to enable ADPCM audio (it obviously won't work anyway if you have the good unit in a configuration where the ADPCM audio is turned off!)
Once you have the good unit working, and have a kind of "feel" for how the circuit responds/sounds, repeat the test on the bad unit.  If you hear the same sound, the amplifier chain is good, and U502 is not generating any sound.

If you don't hear any sound, move to U503 pin 6; if no sound, move to U503 pin 2.  Remember to check on the good unit first, so as to "train your ear" for what the correct sound is.

Finally, if you still don't hear any sound, move to pin6 of U506 (this is right channel audio only).  THIS MUST WORK, as U506 is part of the generally amplifier chain, and you have indicated that other, non-ADPCM sound works.  If at this point, you don't hear any sound, then either you are performing the test incorrectly, or the other non-ADCPM sound actually are NOT working.

Assuming you meant that U502 is not generating at all(or, of course, you discover that via the screwdriver test):
check for the voltages as listed above.  The digital data comes into U502 from the CPU on pins 4,5,6,7,14,15,16.

Let me know where you do, and do not, hear the static/hum/buzz/click.

Good luck,
Charlie

[cosmos99]

Thanks Charlie for the info about ADPCM sound , i did change c653 on my japanese DUO and i get loud ADPCM sound now,it was weak before,as redbook audio ,i did change the 6 usual caps with success.

[Duo_R]

ok so update here - I swapped out IC503 as previously mentioned with no success. This morning I did a trace and discovered that pin 4 of the 4558 was not grounded. I had a bad trace and I proceeded to jump pin 4 to ground. I grounded to the closest ground (about an inch away) with a small wire. Booted up the Duo and to my surprise the ADPCM now works! that is great news, a step forward after a long time of no progress.

However, final issue remains, my sound volume is now about 20% of normal capacity. Somehow jumping this 4558 has caused my already low volume to decrease even more. Since this is now the issue, think I should start swapping out the 4558's on the top of the board? Here is a summary of the work done:

* EVERY cap on the TurboDuo except for the 3.3uf cap
* OKI5205
* OKIM41464-10 x2
* M51131
* I503 IC (4558 Op Amp)

[Duo_R]

ok I am almost there. I discovered a bad connection in a cap in the pre-amp circuit (did the pencil eraser test). After reflowing solder I got 100% sound volume in the left channel, and about 20% in the right channel. I am suspecting another bad connection. It is no surprise with how many times I have been tinkering with this board, just a matter of identifying the right cap now. I swapped out the 4558 near the audio circuit but didn't do anything different. Does anyone know which caps are a part of the the right sound channel here (post and pre):




Also this is the pinout of the 4558 I have been swapping. It is useful to know the pinouts for tracing, etc:

[Charlie]

Does this help?

http://www.pcenginefx.com/forums/index.php?topic=960.msg113601#msg113601

If it's not clear, C604 is Right, c610 is Left.


Charlie

[Duo_R]

That should, forgot that you posted that. Do you have an idea on how I can test the circuit (perhaps earlier on) to figure out when the right channel quits? Like could I jump this to a small speaker somehow? I would need to know the left and right channel circuits when doing this.

[Duo_R]

Also I could use a headset to test, just match up the + and - to the caps that are associated with the right sound channel correct?

[Charlie]

At this point, the L & R circuits going forward are pretty much identical.  So, in theory, you can just connect identical points between L and R and see what happens.  One common sense things to be aware of:  If there is a serious problem in one channel, jumping it "in parallel" to the good channel could very well blow the good channel.

There are a couple of ways to reduce the chances of an error:
1. Use a meter/scope to compare identical points on each channel.  If the DC voltage is very close (< .5v or so) , those two points are probably ok; if there is a more serious voltage difference, don't just connect via a jumper wire.  Of course, if it IS a serious difference, you have a hint of where the problem is already!

2. If you don't have a meter, or you do and it shows non-compatible voltages, get a non-polarized cap.  (Or, make one out of two polarized caps...connect them in series, either + to + OR - to -).  Use a low value cap (1uf, maybe 4.7uf), and make the jump via the two open connections on the cap pair.  This will protect the good channel from any bad voltages on the bad channel.

In either case, you want to try to determine whether the problem is further "ahead" in the circuit, or further "back" in the circuit.  Using the jumper technique, common sense will tell you which way to go.  If connecting C604 and C610 together gives good output on both channels, then obviously the "ahead" circuity is ok...the problem is that the audio is not getting to it. Keep working the jumper cap forward in the circuit, until you find the point at which something drastically changes. And, of course, vice-versa.

Also, is the problem in the headphone jack...maybe the internal switch?

Finally, remember that there is an anti-thump circuit, but only on the headphone out...and you can check audio on the L & R of the AV out.

If you have a question on this part of the circuit, especially in reference to L vs R, please ask...but try to be specific.
Charlie
 

[Duo_R]

I only have a ohmeter/multimeter, but if I am measuring voltages then that won't be a problem. So I will do voltage test and start at one end and move forward / back. I really like your idea of using the non-polarized cap, that makes perfect sense. So jumping the pairs live and see if I get sound or don't get sound. Great idea! I know where the pairs travel near the 4558 opamp circuit, but further behind there I am not so sure (like once it gets to I believe IC505). I know based on what you stated what IC is in the series, but not sure on the pins. But I will start checking that out.

Yes the right channel is not working via AV connector or the headphone jack. And it isn't 20%, it is 0%. I though it was 20% but realized it was my TV using the mono sound and putting sound in both speakers. Hooked up to my headphones / speakers the right channel is dead. 

Thanks a ton Charlie, your insight has been extremely helpful! Again where did you get such insight into the Duo circuits?

[Charlie]

Well, I don't have a lot of insight into the internals of a DUO, but I do know electronics.
Ok, rather than go back through all this stuff, let's just hit the high points:
1. Right channel NG, left channel ok?  I assume so, for the following (If I'm backwards, then, of course, reverse the L & R stuff)
2. Are we really talking ADPCM?  As comes from the U502?  To check this, have left channel working, then temporarily connect U503 pin6 to ground.  If the sound continues, it's NOT ADPCM from U502. 

Check this first.  Does the  sound continue, or disappear?

Charlie

[Charlie]

Next, if the sound disappears, then yes it is from U502.  This is good, as at this point it is monoaural, so if either one channel works, all the monoaural must be working.  This audio goes into u504, pin 4 (still good as monoaural).  It should be coming out pin 5, still monoaural.  This, finally, is the point at which it separates.

Make sure we're correct to this point, as the next stuff gets complicated.

(If the sound does NOT disappear when you do the temp jump, we're already going the wrong way.)

Charlie

[Duo_R]

just to clarify ADPCM works (well in the left channel at least). So the issue now is just no sound in the Right sound channel.