SCR randomly triggers - amplifier protection circuit

Thread Starter

tttapa

Joined Jan 31, 2016
22
Hi,
I'm having trouble with my Sansui AU-777 solid state amplifier.

TLDR:
The amplifier has a protection circuit with two current sensing resistors on the emitters of the output transistors. When the voltage across one of these resistors is too high, it turns on an SCR that switches on the "protector" light, and mutes the main amp. However, the SCR switches on all the time, pretty randomly, without high voltages across the resistors.
Could the SCR be broken?

History:
A couple of weeks ago, the protection circuit randomly kicked in while the amplifier was idle (my computer was connected, but wasn't playing any sound). After a few hours, I tried to turn it on again, and the protection kicked in again. I just unplugged it, and put it away until a few days ago.
I didn't see any visible damage, so I plugged it into my dim bulb tester to measure some voltages. It worked again, and all voltages were normal, idle current of the main amp was normal, etc.
So I put everything back together, connected some speakers, all fine, no distortion, tried a stress test at high volume, no problems whatsoever.
However, later that day, while playing at low volume, the protection circuit kicked in again.

Schematic:
Sansui-AU-777.png
The protection circuit works by connecting B6 to ground, taking away the power from one of the first stages of the amplifier, basically muting the main output stages.
The official schematic is attached as a PDF file.

What I've done so far:
I've measured the voltages of the power supplies, and they all are pretty close to what they should be according to the service manual.
I've adjusted the biasing voltages and main amp current, as explained in the service manual.
I've hooked up my scope to the gate of the protection circuit SCR, and waited for it to trigger. This is what that looks like:
SCR.png
The SCR triggers when the gate voltage is around 0.6V, then it turns on and the gate voltage rises to 0.9V (within a couple of microseconds).

However, if I look at the voltages right before the two diodes (at the emitters of the output transistors), there is nothing there to trigger the SCR. Sometimes when turning on the amplifier, there's a small peak of 400mV at most, but nothing as high as 0.8V (0.6V trigger voltage + 0.2V diode forward voltage). Sometimes the SCR just triggers a couple of minutes after turning on, without anything before the diodes.
Even when I disconnect the power to the main amp (Vaa), the protection circuit often kicks in.

I have absolutely no idea why this happens. I've tried adjusting VR901. That seemed to work at first, but eventually, it triggered again. Triggering seems to be independent of the position of VR901, except when turning it down completely (i.e. tying the gate directly to ground).
Right now, it triggers pretty much every time I turn it on, about an hour ago, it worked just fine.

What could be the cause for this problem? Is the SCR broken? Should I look for another culprit?
Thanks in advance,

Pieter
 

Attachments

Last edited:

crutschow

Joined Mar 14, 2008
34,285
It would appear to be something funny with the SCR, although that's an odd type of failure.

You could reduce the sensitivity to any transients by adding a small capacitor between the SCR gate and ground (say 0.1μF) and see if that makes any difference.
 

Thread Starter

tttapa

Joined Jan 31, 2016
22
Thanks for the reply. I added a 0.1µF ceramic capacitor between gate and ground like you suggested. At first, it worked fine, but after two minutes or so, it went in protection mode again. I turned it off for a couple of minutes
Right now, it works fine. It only goes into protection mode at the peak of my 1kHz sine wave test tone if I turn up the volume too high (at a peak collector current of a little over 2A).

It's been 10 minutes since I wrote that first paragraph, and the SCR is acting up, again. I give up.

Does anyone know where I can find a replacement for the 2SF-656?
A quick Google search shows the NTE5402 as an alternative, but I can't find a seller that ships to Belgium (without astronomical shipping costs or minimum orders).

Thanks a lot,
Pieter
 

ian field

Joined Oct 27, 2012
6,536
Hi,
I'm having trouble with my Sansui AU-777 solid state amplifier.

TLDR:
The amplifier has a protection circuit with two current sensing resistors on the emitters of the output transistors. When the voltage across one of these resistors is too high, it turns on an SCR that switches on the "protector" light, and mutes the main amp. However, the SCR switches on all the time, pretty randomly, without high voltages across the resistors.
Could the SCR be broken?

History:
A couple of weeks ago, the protection circuit randomly kicked in while the amplifier was idle (my computer was connected, but wasn't playing any sound). After a few hours, I tried to turn it on again, and the protection kicked in again. I just unplugged it, and put it away until a few days ago.
I didn't see any visible damage, so I plugged it into my dim bulb tester to measure some voltages. It worked again, and all voltages were normal, idle current of the main amp was normal, etc.
So I put everything back together, connected some speakers, all fine, no distortion, tried a stress test at high volume, no problems whatsoever.
However, later that day, while playing at low volume, the protection circuit kicked in again.

Schematic:
View attachment 134669
The protection circuit works by connecting B6 to ground, taking away the power from one of the first stages of the amplifier, basically muting the main output stages.
The official schematic is attached as a PDF file.

What I've done so far:
I've measured the voltages of the power supplies, and they all are pretty close to what they should be according to the service manual.
I've adjusted the biasing voltages and main amp current, as explained in the service manual.
I've hooked up my scope to the gate of the protection circuit SCR, and waited for it to trigger. This is what that looks like:
View attachment 134664
The SCR triggers when the gate voltage is around 0.6V, then it turns on and the gate voltage rises to 0.9V (within a couple of microseconds).

However, if I look at the voltages right before the two diodes (at the emitters of the output transistors), there is nothing there to trigger the SCR. Sometimes when turning on the amplifier, there's a small peak of 400mV at most, but nothing as high as 0.8V (0.6V trigger voltage + 0.2V diode forward voltage). Sometimes the SCR just triggers a couple of minutes after turning on, without anything before the diodes.
Even when I disconnect the power to the main amp (Vaa), the protection circuit often kicks in.

I have absolutely no idea why this happens. I've tried adjusting VR901. That seemed to work at first, but eventually, it triggered again. Triggering seems to be independent of the position of VR901, except when turning it down completely (i.e. tying the gate directly to ground).
Right now, it triggers pretty much every time I turn it on, about an hour ago, it worked just fine.

What could be the cause for this problem? Is the SCR broken? Should I look for another culprit?
Thanks in advance,

Pieter
Crowbar protection circuits usually have about 0.1 - 0.22uF shunting the SCR gate against spikes or noise.

The inductor in series with the resistor being monitored for volt drop will tend to accentuate noise or spikes.

It might be worth firing up an oscilloscope to see if there's more than just the wanted signal at work - once I had an amplifier burn up the output stages because a parasitic 100kHz was getting in somewhere. The fault had made it not so loud when running flat out, but no other symptoms to warn me of the impending smoke signals.
 

GopherT

Joined Nov 23, 2012
8,009
Thanks for the reply. I added a 0.1µF ceramic capacitor between gate and ground like you suggested. At first, it worked fine, but after two minutes or so, it went in protection mode again. I turned it off for a couple of minutes
Right now, it works fine. It only goes into protection mode at the peak of my 1kHz sine wave test tone if I turn up the volume too high (at a peak collector current of a little over 2A).

It's been 10 minutes since I wrote that first paragraph, and the SCR is acting up, again. I give up.

Does anyone know where I can find a replacement for the 2SF-656?
A quick Google search shows the NTE5402 as an alternative, but I can't find a seller that ships to Belgium (without astronomical shipping costs or minimum orders).

Thanks a lot,
Pieter

A BT169D is essentially the same and commonly available with the same pinout. The BT169D can handle higher voltage but all is good. Send me a PM if you cannot fine one cheap. I have a whole drawer full of them.
http://www.mouser.com/ds/2/302/BT169D-527397.pdf
 

Dodgydave

Joined Jun 22, 2012
11,285
Hi,
I'm having trouble with my Sansui AU-777 solid state amplifier.

TLDR:
The amplifier has a protection circuit with two current sensing resistors on the emitters of the output transistors. When the voltage across one of these resistors is too high, it turns on an SCR that switches on the "protector" light, and mutes the main amp. However, the SCR switches on all the time, pretty randomly, without high voltages across the resistors.
Could the SCR be broken?

History:
A couple of weeks ago, the protection circuit randomly kicked in while the amplifier was idle (my computer was connected, but wasn't playing any sound). After a few hours, I tried to turn it on again, and the protection kicked in again. I just unplugged it, and put it away until a few days ago.
I didn't see any visible damage, so I plugged it into my dim bulb tester to measure some voltages. It worked again, and all voltages were normal, idle current of the main amp was normal, etc.
So I put everything back together, connected some speakers, all fine, no distortion, tried a stress test at high volume, no problems whatsoever.
However, later that day, while playing at low volume, the protection circuit kicked in again.

Schematic:
View attachment 134669
The protection circuit works by connecting B6 to ground, taking away the power from one of the first stages of the amplifier, basically muting the main output stages.
The official schematic is attached as a PDF file.

What I've done so far:
I've measured the voltages of the power supplies, and they all are pretty close to what they should be according to the service manual.
I've adjusted the biasing voltages and main amp current, as explained in the service manual.
I've hooked up my scope to the gate of the protection circuit SCR, and waited for it to trigger. This is what that looks like:
View attachment 134664
The SCR triggers when the gate voltage is around 0.6V, then it turns on and the gate voltage rises to 0.9V (within a couple of microseconds).

However, if I look at the voltages right before the two diodes (at the emitters of the output transistors), there is nothing there to trigger the SCR. Sometimes when turning on the amplifier, there's a small peak of 400mV at most, but nothing as high as 0.8V (0.6V trigger voltage + 0.2V diode forward voltage). Sometimes the SCR just triggers a couple of minutes after turning on, without anything before the diodes.
Even when I disconnect the power to the main amp (Vaa), the protection circuit often kicks in.

I have absolutely no idea why this happens. I've tried adjusting VR901. That seemed to work at first, but eventually, it triggered again. Triggering seems to be independent of the position of VR901, except when turning it down completely (i.e. tying the gate directly to ground).
Right now, it triggers pretty much every time I turn it on, about an hour ago, it worked just fine.

What could be the cause for this problem? Is the SCR broken? Should I look for another culprit?
Thanks in advance,

Pieter
I would remove or monitor D901 Anode, or D902 Anode and see which pair of transistors is triggering the thyristor, ....

If all goes well replace the thyristor with a C106D...
 

ian field

Joined Oct 27, 2012
6,536
I would remove or monitor D901 Anode, or D902 Anode and see which pair of transistors is triggering the thyristor, ....

If all goes well replace the thyristor with a C106D...
The SCR is probably OK until it fails S/C on activation.

Isolating the diodes one at a time should identify the dodgy channel, then comparative voltage checks could be enough if the TS doesn't have a scope. The front end inputs should be connected in parallel for true mono operation.

There's some direct coupled stages as you get nearer the final, so intermittent dry joints are a likely suspect.

The transistor numbers look pre RoHS - but proper solder was never 100% perfect.
 
Top