Speaker pop on shut off Sanyo JCX2900

[curtisw99]

I went back to the preamp and measured voltage at shutoff . I measured at the pre-out jacks on the back. My multimeter isn’t fast, but I get at least a 1600mv DC spike out of the preamp at shutoff.
Is this normal?

Have you tried removing audio path coupling caps to help identify area where pop is injected into audio.
The caps of interest are C551,557,562,565,591,592.
Suggest start with C591, need to do both L&R channel, expect no pop, then try C562.

Ok, I’ll try that.
I may also replace all the transistors in the tone boards (6 or 8 I think)

Either way, it’s a bear to get the preamp boards out of the front, so that’s a job for next weekend.

 

[mbz]

I may also replace all the transistors in the tone boards (6 or 8 I think)

I've been playing with the simulators and am getting something of an oscillation at power off around Q553 influenced by C564(100u) and C563(12p). Check both, a large C564 causes problems so don't bump up the values from what was installed. Maybe replace the 12p with MLCC type, something like 80-C322C120G1G5TA

OK, results from simulators are questionable

 

[curtisw99]

Have you tried removing audio path coupling caps to help identify area where pop is injected into audio.
The caps of interest are C551,557,562,565,591,592.
Suggest start with C591, need to do both L&R channel, expect no pop, then try C562.

I've been playing with the simulators and am getting something of an oscillation at power off around Q553 influenced by C564(100u) and C563(12p). Check both, a large C564 causes problems so don't bump up the values from what was installed. Maybe replace the 12p with MLCC type, something like 80-C322C120G1G5TA

OK, results from simulators are questionable

So you suspect C563 is bad? Or is this a design issue? And what does going from one type of ceramic cap to another do?

 

[mbz]

So you suspect C573 is bad? Or is this a design issue? And what does going from one type of ceramic cap to another do?

Thinking it's a design choice to live with some instability at power off since the protect circuit will/should mute it.
The other player is the ceramic C563, no evidence that its bad but if you are willing to replace all transistors on the tone board then consider replacing this cap, lowing the value of the ceramic caused greater instability, so if its drifted out of spec, though ceramics rarely fail.

 

[curtisw99]

Thinking it's a design choice to live with some instability at power off since the protect circuit will/should mute it.
The other player is the ceramic C563, no evidence that its bad but if you are willing to replace all transistors on the tone board then consider replacing this cap, lowing the value of the ceramic caused greater instability, so if its drifted out of spec, though ceramics rarely fail.

Sorry again, meant c563 not c573
I don’t know why I’ve made so many typos in this thread

In the preamp circuits:
Are ceramics and electrolytics for DC filtering and the mylars are in the audio chain?

 

[mbz]

In the preamp circuits:
Are ceramics and electrolytics for DC filtering and the mylars are in the audio chain?

Generally ceramic are used as a noise path to GND, eg on +ve and -ve rails, also used for stability, connected between transistor legs eg, c-b,C563,,,

Electrolytics used for power rail filtering, eg C566 and also dc blocking in audio path eg C592, C551
Mylars tend not to be in the audio path, they do sound harsh, they are found in the tone/filter circuits.
C589, C590 are mylar 0.1uf they are in the audio path, I'd investigate if MKP would fit, something from Vishay (MKP1837??)

 

[curtisw99]

Generally ceramic are used as a noise path to GND, eg on +ve and -ve rails, also used for stability, connected between transistor legs eg, c-b,C563,,,

Electrolytics used for power rail filtering, eg C566 and also dc blocking in audio path eg C592, C551
Mylars tend not to be in the audio path, they do sound harsh, they are found in the tone/filter circuits.
C589, C590 are mylar 0.1uf they are in the audio path, I'd investigate if MKP would fit, something from Vishay (MKP1837??)

On the power supply:
I just measured that I get 54 DC and 118v AC on both sets of filter capacitors at the positive node.
Also, at some of the test points, I can measure significant (volts not millivolts) at the same test point.

Does this mean the rectifiers are failing?
Or is something not grounded?

Or is my multi-meter just too cheap to tell the difference between AC and rectified DC?

 

[mbz]

Or is my multi-meter just too cheap to tell the difference between AC and rectified DC?

Probably.
If the DC voltage is within spec (abouts) then rectifiers working ok.
You may have upto 1Vp-p ripple riding on the 54Vdc rail, some cheap MM may interpret this all as VAC though most I've come across seem to work it out. Try reversing the probe orientation, does it make a difference?

Also, at some of the test points, I can measure significant (volts not millivolts) at the same test point.

example? assume you are talking Vdc.

 

[curtisw99]

Probably.
If the DC voltage is within spec (abouts) then rectifiers working ok.
You may have upto 1Vp-p ripple riding on the 54Vdc rail, some cheap MM may interpret this all as VAC though most I've come across seem to work it out. Try reversing the probe orientation, does it make a difference?


example? assume you are talking Vdc.

I can measure significant AC and DC on the same test point

 

[mbz]

exactly what test point? Give component number or transistor leg or...
Need to work out if it's supplied by the regulated or unreg supply, there will be some ripply from unreg supply
Also give AC and DC measurement result, eg, 2.3Vac, 40mVdc

 

[curtisw99]

Probably.
If the DC voltage is within spec (abouts) then rectifiers working ok.
You may have upto 1Vp-p ripple riding on the 54Vdc rail, some cheap MM may interpret this all as VAC though most I've come across seem to work it out. Try reversing the probe orientation, does it make a difference?


example? assume you are talking Vdc.

Reversing the probes just changed the polarity of the result; no change in the amount of volts AC or DC.

Summary of conclusions (so far):
-1.5-2v measured at speaker terminals at shutoff
-protection circuit and relay working well
Therefore, relay is opening early at shutoff due to the DC spike

-when preamp is disconnected from amplifier, there is no pop at speakers
-1.5v-2v DC spike can be measured at the preamp out
This leads me to believe the DC spike is coming from either the power supply or the preamp itself

All caps in the receiver, except for the tuner, have been replaced.
Power supply transistors have been replaced

I’m now looking at the film and ceramic caps in the preamp and power supply.

 

[curtisw99]

Negative on the triac.
Look carefully at the schematic of the protection circuit. This kinda caught me by surprise. There is a set of NC contacts on the power switch that ground the relay drive when the power switch is moved to off. When turning the unit off, be decisive about it. You may have to take that switch apart and clean those contacts with some Brasso or Flitz. Make sure you get all the residue out.
If this works, it will be the first time I've seen fixing a power switch correct a "pop" condition.

Still working on this pop.

Here’s the switch schematic.
There’s 4 different sets of contacts. Are they all contained in a single switch?

Looks like
-One contact goes to what looks like a bleeder resistor on the relay board
-One goes to transformer 1
-One goes to transformer 2
-One goes to a mute in the tuner

Why would they do this?

 

[mbz]

Schematic shows the power switch has the following contacts,
S15-1, S15-2 mains feed to separate transformers probably to reduce noise. Also you can then
use a lower spec'd switch "2Amp" as opposed to "4Amp".

s14-1 to protect circuit to aid immediate release of spkr relay. Runs down C412

S14-2 to tuner circuit, I haven't studied this, don't understand how tuners work...
Schematic shows tab 7 (MUT Switch?) and tab 8 (MUT REturn?) which goes back to the spkr relays.
C354 is confusing, blocking DC... maybe replace, Nichicon UEP or...

 

[curtisw99]

Schematic shows the power switch has the following contacts,
S15-1, S15-2 mains feed to separate transformers probably to reduce noise. Also you can then
use a lower spec'd switch "2Amp" as opposed to "4Amp".

s14-1 to protect circuit to aid immediate release of spkr relay. Runs down C412

S14-2 to tuner circuit, I haven't studied this, don't understand how tuners work...
Schematic shows tab 7 (MUT Switch?) and tab 8 (MUT REturn?) which goes back to the spkr relays.
C354 is confusing, blocking DC... maybe replace, Nichicon UEP or...

I noticed tab 7, mute switch.
So the question is: is it a mute for the tuner? Or a mute for shutoff?
I’m looking into the same thing.
There’s a muting switch on the front that drops everything 20db

 

[mbz]

There’s a muting switch on the front that drops everything 20db

Yeah, that caused me some confusion earlier on.
Corresponds to switch S6-1, S6-2, forces the audio through additional 47k resistors (R542, R642) so attenuates the audio, this is how most(?) amps perform the 20dB muting.

I noticed tab 7, mute switch.
So the question is: is it a mute for the tuner? Or a mute for shutoff?

Initially I was leaning(guessing) that it was a Tuner mute for the same reasons, ie, kill any transients but then I noticed pin 8 going back to the speaker relay coils but then C354 function confuses me.

Plan to tinker with the simulator, this is what I'm getting around Q553 and output to tone stage.
I've added a 20mS delay before shutting down and the transients happen about 7mS after that

 

[mbz]

Since pop also occurs when TONE DIRECT is on then that eliminates tone control switches and Q553 and surrounds. So audio goes from C557, R568, S8-1, to the vol pot then back to S13-1...

 

[mbz]

Some more things to try if you have the time,

- With low filter on, note pop, would you say it's about the same/no difference as previous.
low filter switch S13 adds additional caps (= dc blocking) C589 and C590

- With tone direct on again is pop about the same level?

- Check solder on caps C557, C591, C592 no solder bridges shorting out the caps?

- Did you try powering up/down with C557 removed? Likewise with C592 removed.
The "pop" finds it's way into the audio path usually as a dc offset or a arc/spark and played through
the spkrs. Removing these caps in turn breaks the audio path hopefully isolating the source of the pop

 

[tarior]

Still working on this pop.

Here’s the switch schematic.
There’s 4 different sets of contacts. Are they all contained in a single switch?

Looks like
-One contact goes to what looks like a bleeder resistor on the relay board
-One goes to transformer 1
-One goes to transformer 2
-One goes to a mute in the tuner

Why would they do this?

I have to apologize. I have been so busy lately, I scarcely post on AK. This is, IMO, an odd problem. I would like to figure it out, if I can get some time, in case it ever comes up again.

 

[tarior]

New switch didn’t fix the pop.
The old switch was also a replacement.
Video added to show the “pop” sound

I’ll start looking in the protection board.

You should really turn the volume all the way down before power off, and leave it down at power on until after the speaker relay closes.

I think the problem is the relay closes at the same time as the power switch. My other receivers have a delay.

Which is the timing capacitor?

Did you fix this problem? If so, how?

I put in a new switch and no change
But it looks like someone was in there before and might have changed something.
I need pictures of a switch board from someone who doesn’t have the pop.

Just because the switch is "new" doesn't mean that it's good. Those SDU3P switches have been out of production since the early 1980s, and can get cruddy contacts in the slide portion just from storage.
You need to check that S14-1,2 make a good connection to ground. The end of R420 connected to the switch should show an ohm or less, including the resistance of your meter leads, to ground.

I found the root cause of the pop. The relay is opening too fast. There needs to be a delay before the relay opens at shutoff.

This is incorrect. The way this unit is designed, the relay should open immediately. Voltage at the base of Q431 should be zero pretty much the instant you release the power switch when powering off.
I'm leaning toward something really odd like the relay driver transistor is somehow latching up, or there is just something you aren't seeing to tell us about.

 

[soccerfan]

Had a very similar issue on a sansui au 7900. Pop on turn off. The negative preamp supply was collapsing at a much lower rate than the positive supply causing a pop . Putting pre amp/main in separated mode was not causing any pop.
I changed the ac sensing cap from 47uf to 4.7uf which sped up the relay drop and stopped the pop
I haven't checked the schematic of the sanyo, but see if you can play with the ac sensing cap spec and it eliminates the pop