Yamaha A-S500 Amplifier DC protection

[elecdonia]

Connected back up with both L/R channels of CB101 connected to the power main board and when turned on it still clicks off with a flashing power led. It seems the IC401 need isolating more somehow. If i connect it with the R channel disconnected from the power main board, it works as follows.
The volume works up to about number 1(if there were numbers) then the loudness takes over. The input selector still lights up and does need to be on line 1.

There is something odd going on here but I'm still certain it can be resolved.
By "volume" which knob are you rotating?

Your photo showed you rewired the Function(2) PC board correctly.
After the rewiring, then the large "Volume" knob (right end of front panel) should have no effect at all.
The smaller vertical "bar" knob labeled "LOUDNESS" should now be controlling the volume.
Please tell me which knob is now controlling the volume: is it "VOLUME" (large knob) or "LOUDNESS" (smaller vertical bar knob) ?

As for the input selector switch, it is normal for the LED indicators to continue to change when you rotate the knob. That portion isn't changed by the rewiring. However, the position of the selector switch should not have any effect on what you hear. The signal you are applying to the Line 1 input jacks should be coming through all the way  to the speakers regardless of what position the selector switch is in.

I'm hoping that the service manual actually matches your amplifier...
The images of the Function(2) PC board in the service manual look exactly like your photo of the board.
And your rewiring looks exactly the way it should be to send Line 1 through the 4 middle knobs (BASS, TREBLE, BALANCE, LOUDNESS) and then feed it into the power amplifier section.

There is a chance of having an issue located on the board named OPERATION(2). This board is right behind the front panel and down near the bottom of the unit. Gaining access to it requires removing all knobs and then both the outer and inner front panel assemblies. But don't start taking this apart until you can confirm which knob is controlling the volume: 

is it VOLUME knob or LOUDNESS knob ?
   

[tiggerlator]

Connected back up with both L/R channels of CB101 connected to the power main board and when turned on it still clicks off with a flashing power led. It seems the IC401 need isolating more somehow. If i connect it with the R channel disconnected from the power main board, it works as follows.
The volume works up to about number 1(if there were numbers) then the loudness takes over. The input selector still lights up and does need to be on line 1.

There is something odd going on here but I'm still certain it can be resolved.
By "volume" which knob are you rotating?

Your photo showed you rewired the Function(2) PC board correctly.
After the rewiring, then the large "Volume" knob (right end of front panel) should have no effect at all.
The smaller vertical "bar" knob labeled "LOUDNESS" should now be controlling the volume.
Please tell me which knob is now controlling the volume: is it "VOLUME" (large knob) or "LOUDNESS" (smaller vertical bar knob) ?

As for the input selector switch, it is normal for the LED indicators to continue to change when you rotate the knob. That portion isn't changed by the rewiring. However, the position of the selector switch should not have any effect on what you hear. The signal you are applying to the Line 1 input jacks should be coming through all the way  to the speakers regardless of what position the selector switch is in.

I'm hoping that the service manual actually matches your amplifier...
The images of the Function(2) PC board in the service manual look exactly like your photo of the board.
And your rewiring looks exactly the way it should be to send Line 1 through the 4 middle knobs (BASS, TREBLE, BALANCE, LOUDNESS) and then feed it into the power amplifier section.

There is a chance of having an issue located on the board named OPERATION(2). This board is right behind the front panel and down near the bottom of the unit. Gaining access to it requires removing all knobs and then both the outer and inner front panel assemblies. But don't start taking this apart until you can confirm which knob is controlling the volume: 

is it VOLUME knob or LOUDNESS knob ?
   

The big volume control works up to number 1 if there were numbers, then nothing until the loudness control up to full volume. i guess i need to check if leaving the main volume on zero still produces sound when turning the loudness control up. Both channels of the power section work, the problem is the R output from the pre section.

[elecdonia]

The big "VOLUME" control works up to number 1 if there were numbers, then nothing until the "LOUDNESS" control up to full volume. i guess I need to check if leaving the main "VOLUME" control on zero still produces sound when turning the "LOUDNESS" control up.

OK let's try this:
1) Set main "VOLUME" control (large knob) to minimum (set to "0," which is  fully counterclockwise)
2) Then rotate "LOUDNESS" control over its full range. Sound level should get louder as you rotate "LOUDNESS" control clockwise. In fact the normal setting for "LOUDNESS" is fully clockwise. This is where the "LOUDNESS" bar knob is straight up and down. Rotating "LOUDNESS" knob counterclockwise should reduce the volume, but I don't expect it to reduce it all the way down to nothing.

3) It is possible that you won't hear any sound at all with main "VOLUME" (large knob) fully counterclockwise, regardless of what position "LOUDNESS" knob is at.

      Please tell me whether this is case (or not)?

4) If there is no sound with main "VOLUME" fully counterclockwise, then turn it up by 1/4 turn clockwise. Does this get the sound to start coming out?
Also try setting main "VOLUME" control  to 1/2 volume, with pointer on large "VOLUME" knob straight up. What does this do?
Does the sound level change any at all when you rotate the large main "VOLUME" knob back and forth between 1/4 and 1/2?

One possibility is that you won't hear anything at all until main "VOLUME" is above 0. Perhaps main "VOLUME" totally shuts off the sound when at 0, but at all settings above 0 the volume doesn't really change. In other words, no difference between 1/4 and 1/2 setting of main "VOLUME" control knob.

5) Repeat steps 1-2-3-4 with "PURE DIRECT" button both pressed and released. I don't know which position (button in or button out) corresponds to the mode where the BASS TREBLE BALANCE LOUDNESS controls are active and functional.

Details:
-  The "PURE DIRECT" switch is supposed to totally bypass these four controls (BASS TREBLE BALANCE LOUDNESS) in one of its two positions. In this mode none of these four controls should do anything at all
-  In other position of "PURE DIRECT" button then the BASS TREBLE BALANCE LOUDNESS controls should all "do something" when rotated over their full ranges.

We want to set "PURE DIRECT" button so that BASS TREBLE BALANCE LOUDNESS are all working
 

big "VOLUME" control works up to number 1

Can you explain "what changes about the sound" at the time when the main "VOLUME control knob is is moved just above or below position  "1" ?

Is it like this?  When you  rotate large main "VOLUME" knob to "above 1" then the sound abruptly cuts out?

Does this "1" position correspond to the pointer on large main "VOLUME" knob being near the letter "E" on the word "VOLUME" ?

Both channels of the power section work, the problem is the R output from the pre section.

Understood. For now let's do all testing with the R wire disconnected from CB101. Goal is to get the L channel to be as functional as possible. After that then we will sort out the problem which R channel preamp section still has.

[Audiorepair]

Have you actually checked any of these 10uF caps to at least rule them out as the suspect?

[elecdonia]

Have you actually checked any of these 10uF caps to at least rule them out as the suspect?

This unit isn't terribly old ( I'd guess 10 years). Audio coupling caps don't get hot (no ripple current) and can last much longer than 10 years. Although I always use my ESR meter to check every electrolytic, the vast majority of high-ESR capacitors which I find are located in switching power-supply circuits (where there is high ripple current). Also the failure mode of small electrolytics (10uF, etc.) is almost always "high ESR." I can't recall a single case of one shorting out or becoming leaky.

Where I'm feeling a challenge helping the OP diagnose this fault is from the extreme complexity of the circuit design employed in this integrated amplifier. I'm beginning to suspect my previous diagnosis of IC401 being at fault. After reading the schematic several times I discovered a 2-transistor "muting circuit" (1 transistor per channel) is located on a PC board just behind the front panel. These transistors shunt the line-level audio signals to ground when turned on by receiving a logic-level control signal from the MCU. After this muting circuit the 2 line-level audio signals go straight into the power amplifier section. The OP determined that disconnecting the right channel at the power-amp input prevents the "protection" mode from getting activated.  Everything works perfectly when the line-level signal from the left channel preamp section is connected to either the left or right channel of the power amplifier. But the R channel preamp signal always causes "protection" whether connected to L or R of power stage.

Perhaps one of these muting transistors has a fault which puts the logic-level voltage onto the audio line?

So what we know now is that the R channel signal coming from the preamp-level circuitry behind the front panel carries a large DC transient which occurs while rotating the main volume control. It is important to note that audio signals do not pass through this main volume control. Rather, its resistance is measured by an A-D converter inside the MCU, and then the MCU takes various actions depending on the detected resistance of the volume control variable resistor. I really haven't encountered this in many many years of repairing/restoring audio gear!

I just don't think this is a case of faulty 10uF capacitors.
 

[Audiorepair]

Have you actually checked any of these 10uF caps to at least rule them out as the suspect?

This unit isn't terribly old ( I'd guess 10 years). Audio coupling caps don't get hot (no ripple current) and can last much longer than 10 years. Although I always use my ESR meter to check every electrolytic, the vast majority of high-ESR capacitors which I find are located in switching power-supply circuits (where there is high ripple current). Also the failure mode of small electrolytics (10uF, etc.) is almost always "high ESR." I can't recall a single case of one shorting out or becoming leaky.

Where I'm feeling a challenge helping the OP diagnose this fault is from the extreme complexity of the circuit design employed in this integrated amplifier. I'm beginning to suspect my pervious diagnosis of IC401 being at fault. After reading the schematic several times I discovered a 2-transistor "muting circuit" (1 transistor per channel) is located on a PC board located just behind the front panel. These transistors shunt the line-level audio signals to ground when turned on by receiving a logic-level control signal from the MCU. After this muting circuit the 2 line-level audio signals go straight into the power amplifier section. The OP determined that disconnecting the right channel at the power-amp input prevents the "protection" mode from getting activated.  Everything works perfectly when the line-level signal from the left channel preamp section is connected to either the left or right channel of the power amplifier. But the R channel preamp signal always causes "protection" whether connected to L or R of power stage.

Perhaps one of these muting transistors has a fault which puts the logic-level voltage onto the audio line?

So what we know now is that the R channel signal coming from the preamp-level circuitry behind the front panel carries a large DC transient which occurs while rotating the main volume control. It is important to note that audio signals do not pass through this main volume control. Rather, its resistance is measured by an A-D converter inside the MCU, and then the MCU takes various actions depending on the detected resistance of the volume control variable resistor. I really haven't encountered this in many many years of repairing/restoring audio gear!

I just don't think this is a case of faulty 10uF capacitors.

OK, but my point is the OP apparently hasn't, or been able to test these capacitors to indicate they are the real source of the problem or not.

Surely it would make sense to replace a few cheap capacitors in order to now have a tool to use to test the rest of the system.

You would now know for a fact it is not a capacitor problem.



As it stands, you are merely assuming that the caps are good, and are trying to find what else could possibly be wrong based on that logic.

[elecdonia]

problem is, i can still only use the L channel. If i connect the R channel of the CB101 connector, the amp still clicks off with the power led flashing.

As I mentioned earlier, let's keep the R channel wire disconnected from CB101.

But...  Let's do something more with this:  Connect your multimeter +probe from this R channel wire and -probe to chassis ground. We need to see if and when any DC voltage appears on this R channel wire.

These 2 wires (one for L, other for R) carry the audio signals into the power amplifier at the CB101 connector. These wires should never have any DC voltages on them. Also the audio signals on these same wires ("audio" is AC, not DC) should generally be no larger than 100 to 500mV (millivolts) unless the volume is really cranked up all the way.

A signal large enough to trigger "protection" would likely measure at least 1V on the DC scale of your voltmeter, either positive or negative in relationship to ground. We do know that rotating the large volume control causes large (and unwanted) transient signals to appear on this R wire. So watching the voltmeter while rotating the large volume knob should be informative.

Do you have any access to an oscilloscope?

[Martin72]

After reading the schematic several times I discovered a 2-transistor "muting circuit" (1 transistor per channel) is located on a PC board just behind the front panel.

Therefore I´ve wrote :

Two "last chances":
-Observe the solderings on the pcbs for faulty areas.
-Check the transistor Q802 in the mute circuit.

https://www.eevblog.com/forum/repair/yamaha-a-s500-amplifier-dc-protection/msg4792805/#msg4792805

But no feedback.

[elecdonia]

OK, but my point is the OP apparently hasn't, or been able to test these capacitors to indicate they are the real source of the problem or not. Surely it would make sense to replace a few cheap capacitors in order to now have a tool to use to test the rest of the system. You would now know for a fact it is not a capacitor problem. As it stands, you are merely assuming that the caps are good, and are trying to find what else could possibly be wrong based on that logic.

I admire your persistence in suggesting these 10uF capacitors are important. Here's what I just thought of:

   The DC voltage at each end of each 10uF capacitor needs to be measured.

I'm not thinking only of faulty 10uF capacitors, but rather I'm thinking these 10uF capacitors are the perfect location to probe for unwanted DC voltages. They are "through-hole" parts which are easy to locate on the PC boards and it will be easy touch their connecting wires with the probe of a DC voltmeter.

I fully agree that small electrolytic capacitors are "prime suspects" for many faults. But I also have a strong suspicion that "something else" is wrong with this Yamaha amplifier.

I've looked at the schematic for this amplifier over and over: All of its audio circuits operate on balanced +/- power supply rails, either +/-15V for the opamps or +/-7V for IC401. If anything is wrong with these power supplies it will affect both channels equally. However the L channel always works perfectly yet the R channel always generates this horrific loud pop whenever the volume control rotates. Therefore I don't suspect a power supply fault.

The entire audio signal path itself should be at 0V DC due to the use of these equal +/- supplies. So while reading your posts  about the 10uF capacitors the thought popped into my head that "both ends of each 10uF capacitor should be at 0V DC."

When a non-zero DC voltage is located this will pinpoint a bad opamp or muting transistor. Or perhaps IC401 is actually at fault?

So I think the OP will need to test both the + and - wire of each 10uF capacitor for DC voltage. There should be 0V on both ends of all of them.

Unfortunately some of these 10uF capacitors are hidden away on the OPERATION(2) PC board which is just behind the front panel. It might be difficult to probe the capacitors on the OPERATION(2) PC board, but this is what needs to be done next. 

[elecdonia]

Two "last chances":
-Observe the solderings on the pcbs for faulty areas.
-Check the transistor Q802 in the mute circuit.

BINGO!
I don't know how I missed you mentioning Q802 previously. I discovered it myself while going over the schematic. These muting transistors are precisely at the other end of the cable which feeds audio into the power amp board. Of course they are tiny SMT devices tucked away on a PC board OPERATON(2) which isn't exactly easy to reach. But they are certainly in a place where a failed transistor could put major DC transients into the power amp input. It is also possible that the MCU sends brief muting pulses to Q801/Q802 whenever the main volume control is rotated or when the selector switch moves.

Logic-level control signal PREMT which drives Q801/Q802 is large enough to trigger protection if it gets onto the POR line. That will never happen if Q802 is good.

[Martin72]

I looked over again...
Hm-hm...
Yes, a DC potential could be there if the transistor is broken, but the mute signal is for both mute transistors.
Means the 5V (or whatever is the H level) could only appear on the pre-out if the circuit is in mute state(and the transistor is broken).
"We" strongly need a scope for pointing the failure out...

[mbz]

Working with OP on another forum.
Confirmed the following,
DC spike at CB101 right channel, no spike at IC802 pins 5 & 7 (right channel opamp input, output)
Q802b -3.2Vdc., Q802e -2.6Vdc. Suggest replace Q802

[elecdonia]

Yes, a DC potential could be there if the transistor is broken, but the mute signal is for both mute transistors. Means the 5V (or whatever is the H level) could only appear on the pre-out if the circuit is in mute state(and the transistor is broken). "We" strongly need a scope for pointing the failure out...

I suspect the MCU sends a brief "mute" command whenever the main volume control or the input selector switch are rotated. By "brief" I mean <20msec. Still, 20msec is a long enough pulse to make a big loud pop in the audio signal if Q802 C (collector) is open circuit. The B-E junction will act like a forward-biased diode to couple the logic level pulse onto the audio line. Note that on the schematic Q802 is drawn as an NPN transistor with C grounded and E connected to the audio line.  Also the schematic shows both E and C are 0V when B is +0.7V.  The +0.7V on B implies the "mute" state is active.

These "muting" transistors are kind of strange. Their emitter and collector are interchangeable. Some even have a spec named "reverse hfe" which implies that when the B-C junction is forward biased then the reverse-biased B-E junction will conduct. Again, this Yamaha has Q802 C (collector)  grounded with E (emitter) is connected to the L (or R) audio line. This  seems strange, but it's plausible if the B and C terminals on this particular transistor are interchangeable.

I don't recollect needing to replace many of these muting transistors. It would be fascinating if the root cause turns out to be something like poor soldering on the emitter of Q802 or a broken foil on the PC board. Earlier I was all worried that IC401 was faulty (switching and attenuator IC) . IC401 is an insanely complex chip and the line input RCA jacks go straight into it through 470 ohm series resistors which don't provide much ESD protection.

[elecdonia]

Working with OP on another forum. Confirmed the following, DC spike at CB101 right channel, no spike at IC802 pins 5 & 7 (right channel opamp input, output)
Q802b -3.2Vdc., Q802e -2.6Vdc. Suggest replace Q802

i wonder how the B or E voltage on Q802 could get negative?
The schematic shows +0.7V on Q802 B. I expect this to be the case when the amplifier is in the "muted" state.
Schematic also shows both E and C should be 0V DC.
One (very odd) thing about this muting circuit is that C is grounded, with E connected to the audio signal line that gets muted. Evidently these muting transistors are designed so that E and C pins are interchangeable.

As a quick test to confirm this Q802 fault I suggest the OP should remove Q802 from the OPERATION(2) PC board and then test the amplifier again. There won't be any muting on the R channel with Q802 removed, but there also won't be any DC getting into the R channel audio.

[mbz]

yep, remove Q802 would be a good test

[tiggerlator]

Thanks guys.

I will read through these replies and test what you have suggested and report back today ok.

Thanks a lot for your time

Simon

[tiggerlator]

problem is, i can still only use the L channel. If i connect the R channel of the CB101 connector, the amp still clicks off with the power led flashing.

As I mentioned earlier, let's keep the R channel wire disconnected from CB101.

But...  Let's do something more with this:  Connect your multimeter +probe from this R channel wire and -probe to chassis ground. We need to see if and when any DC voltage appears on this R channel wire.

These 2 wires (one for L, other for R) carry the audio signals into the power amplifier at the CB101 connector. These wires should never have any DC voltages on them. Also the audio signals on these same wires ("audio" is AC, not DC) should generally be no larger than 100 to 500mV (millivolts) unless the volume is really cranked up all the way.

A signal large enough to trigger "protection" would likely measure at least 1V on the DC scale of your voltmeter, either positive or negative in relationship to ground. We do know that rotating the large volume control causes large (and unwanted) transient signals to appear on this R wire. So watching the voltmeter while rotating the large volume knob should be informative.

Do you have any access to an oscilloscope?

There is  -2.87v DC on the R wire of CB101 with the volume turned up. the voltage is constant, amp did not go protect as i had the wire disconnected from the CB101 connector before the power amp. There is 0v at the L signal wire.

[tiggerlator]

After reading the schematic several times I discovered a 2-transistor "muting circuit" (1 transistor per channel) is located on a PC board just behind the front panel.

Therefore I´ve wrote :

Two "last chances":
-Observe the solderings on the pcbs for faulty areas.
-Check the transistor Q802 in the mute circuit.

https://www.eevblog.com/forum/repair/yamaha-a-s500-amplifier-dc-protection/msg4792805/#msg4792805

But no feedback.

So sorry i missed that.

[tiggerlator]

So i took Q802 off, reconnected the R channel to the power amp input and the amp works, on both channels with no protect.

So does it need that Q802? Can i use it without it?

If so i can probably find one somewhere.

Can't thank you guys enough.   

Only paid £25 for this amp, and it is seemingly a very good one. I have tested and there is seemingly no problem without Q802, pure direct works too. .

[mbz]

Possibly Q802 is used to mute transients during power up, eg, those annoying thumps through speakers that's somewhat common in vintage gear. Try power up while monitoring the voltage at CB101 right channel, you may see a dc spike. In any case the item should be replaced, you might need it one day to protect your speakers.

[tiggerlator]

Possibly Q802 is used to mute transients during power up, eg, those annoying thumps through speakers that's somewhat common in vintage gear. Try power up while monitoring the voltage at CB101 right channel, you may see a dc spike. In any case the item should be replaced, you might need it one day to protect your speakers.

Just checked the speaker outs, there is -11mv on R and nothing on L. Turning the volume up does not increase it. The amp works seemingly fine, no pops on power on or off, sounds really nice, the best amp i have now by miles.

[elecdonia]

So i took Q802 off, reconnected the R channel to the power amp input and the amp works, on both channels with no protect

This is good news.

It is OK to temporarily use the amplifier without Q802. But it should eventually be replaced.
Q802 is a specialized “audio muting” transistor. Original part number is 2SD2704.  I’ll check and post a list of part numbers which will work to replace it in case it is difficult to find an exact replacement.

And if you haven’t already done so, you should reverse the modification of FUNCTION(2) PC board, restoring it back to its original wiring:
  Unsolder and remove the twisted pair of jumper wires that were added
  Reconnect the free ends of C460 and C461 back to the PC board as they were to begin with
  Use short pieces of wire to reconnect J406 and J407 as they were originally

[tiggerlator]

So i took Q802 off, reconnected the R channel to the power amp input and the amp works, on both channels with no protect

This is good news.

It is OK to temporarily use the amplifier without Q802. But it should eventually be replaced.
Q802 is a specialized “muting” transistor. I’ll check and post a list of part numbers that will work to replace it in case it is difficult to find an exact replacement.

And if you haven’t already done so, You should reverse the modification of FUNCTION(2) PC board, restoring it back to its original wiring:
  Unsolder and remove the twisted pair of jumper wires that were added
  Reconnect the free ends of C460 and C461 back to the PC board as they were to begin with
  Use short pieces of wire to reconnect J406 and J407 as they were originally

I have reversed the mod of function 2. The amp works fine.

There is -11.2mv on right speaker terminal, but no protect and amp works and sounds fine. Will it be ok to use it like this for a while?

Thanks

[elecdonia]

I have reversed the mod of function 2. The amp works fine. There is -11.2mv on right speaker terminal, but no protect and amp works and sounds fine.

Will it be ok to use it like this for a while?

Yes, go ahead and use it.

[tiggerlator]

So i took Q802 off, reconnected the R channel to the power amp input and the amp works, on both channels with no protect

This is good news.

It is OK to temporarily use the amplifier without Q802. But it should eventually be replaced.
Q802 is a specialized “audio muting” transistor. Original part number is 2SD2704.  I’ll check and post a list of part numbers which will work to replace it in case it is difficult to find an exact replacement.

And if you haven’t already done so, you should reverse the modification of FUNCTION(2) PC board, restoring it back to its original wiring:
  Unsolder and remove the twisted pair of jumper wires that were added
  Reconnect the free ends of C460 and C461 back to the PC board as they were to begin with
  Use short pieces of wire to reconnect J406 and J407 as they were originally

Thanks again.

I will replace that Q802 as soon as i find a replacement.