Hifi amplifier biasing issues and output stage shorting?

[Yamin]

What about the grounding?

The schematic as posted shows the power amp ground and the pre-amp ground separate and dissapearing stage left, and no speaker ground indicated.

If there is a problem here where they don't all three connect properly, then, er, er, I dunno what might happen to the power amp.
Maybe DC offset or something.

Its ground might be floating.

Thanks @Audiorepair - I was hoping that I could run something by you. When checking for DC offset should I have connected a dummy load or not. Some people are against it and some people are for it. In my case when I took the measurement I didn't have a load connected.
Is it wise to connect a resistive dummy load or should I have connected a speaker.
Thanks

[Audiorepair]

If you connect a dummy load and still have DC offset, then you still have a problem, IMHO.

Snubbers, capacitors, RF, are not part of this equation.



Have you checked that Speaker negative, input signal  ground, Power Amp ground and almost certainly chassis, are all connected together by zero ohms?

[dietert1]

Output offset may result from RF oscillation, too.
The schematic shown of the output terminal board has something very unusual once more: 10 nF caps across the output terminals. The usual snubber has a 100nF capacitor in series with a low ohms resistor, like 5 or 10 Ohms. There is a special name for that when used with amplifiers, just don't remember.

Also i would check whether the amplifier has the inductors or the jumpers, the schematic shows both options.

Regards, Dieter

[Yamin]

Right guys, so I may have discovered what potentially has caused the amp to fail -
I had the amp turned on was monitoring the bias test points, the voltage measured was exactly what the service manual states 2.5mV at start-up and goes to around 9mV. After about 5 minutes in I heard the relay click I have mentioned about this in a previous post. The moment the relay clicked the voltage at the bias test point varied a bit (it didn’t go over 9mV) and it started going down! It went down till 2.5mV and remained there.
I have now confirmed which relay it is too! Its the relay which switches the B+ voltage. Here is what I measured across the bridge rectifier for the B+ voltage -when the amp is turned on the relay clicks and it measures around 110V after about 5 minutes the relay clicks and the measurement now is 54V. I turned off the amp and immediately turned it back on again. The same thing happens it starts with 110V and exactly around 5 minutes relay clicks and the voltage is 54V. The relays are controlled by a signal from the controller ’SECH1’ not really sure in what case this is suppose to turn on and off.
There are people who have reported about hearing random relay clicks from Onkyo receivers in AV forums. No one has identified the issue but someone has suggested that messing around with the settings (software) - specifically changing the impedance to 4 ohms gets rid of the clicking even though the speaker that he was using was stated at 8ohms.

Could I be on the right track do you think that this might have caused the amp to fail at the owners place? He has got Left and Right tower speakers as well as the centre and the rear speakers too. While I have only one 8ohm speaker connected on my bench which I’m using to test the amp.

Thanks for the help.

[Yamin]

Also i would check whether the amplifier has the inductors or the jumpers, the schematic shows both options.
Regards, Dieter

It has got the two . I'll try to take a picture tomorrow

[xavier60]

I'd say that the control system automatically switchs to low impedance(low voltage) after sensing low output demand over the 5 minutes.
This causes the output stages to become more power efficient and lower operating temperature.
Any idea what the owner's listening habits are?  Loud?
And his speaker impedance?

[Yamin]

I'd say that the control system automatically switchs to low impedance(low voltage) after sensing low output demand over the 5 minutes.
This causes the output stages to become more power efficient and lower operating temperature.
Any idea what the owner's listening habits are?  Loud?
And his speaker impedance?

I'll have to check his setup's specs. He listens at a moderate level normally. One thing I would remark is that the volume needs to be turned around 50 to get any usable level. I know its an arbitrary number but seems the lower part of the volume range is rather a waste. He also sets his volume around 53.
Couple of questions arises now as to how the amp is detecting lower power draw? Maybe an issue on the  detection circuit causing this problem, but I can't imagine how switching to low voltage could blow up the output transistors.
Both of the times when the output blew the fuses marked on the 'low voltage' was blown - and here's another link that I just thought about. When I was testing the amp I heard the relay clicking on and off and it had no detrimental effect at my place. When I took the amp to the owner's place I didn't run the amp for 5 mins or enough for the relay to click. Its only after when I got back, he would have used it for more than 5 mins. Getting very interesting.
But still can't think why switching to low voltage could take out the transistor
Thanks so much   

[xavier60]

The switching to low voltage is very likely unrelated to the failures.
 Output level would be sensed via the FLVOLH line or something.

[Yamin]

The switching to low voltage is very likely unrelated to the failures.
 Output level would be sensed via the FLVOLH line or something.

Thanks @xavier60 I would imagine so too, but during the relay click the voltage at the bias test point does a funny 'wobble' and then goes down. Would under biasing cause a high current draw enough to take out the OT and the fuse?

[xavier60]

The switching to low voltage is very likely unrelated to the failures.
 Output level would be sensed via the FLVOLH line or something.

Thanks @xavier60 I would imagine so too, but during the relay click the voltage at the bias test point does a funny 'wobble' and then goes down. Would under biasing cause a high current draw enough to take out the OT and the fuse?

I can't fully explain that except that it would have something to do with temperature being sensed in 2 places, changing at different rates as the rail voltage changes.
The only things I would suspect are the Pot which has been already covered, and solder joints. I would identify all joints in the bias circuit that are capable of causing loss of control, examine them, then re-solder using  a safe flux.

[Yamin]

Finally I have figured out the why the relay randomly clicks - I was wrong, my speaker actually is rated at 4 ohms. When I changed the settings for 4 ohms from the amp the relay issues was solved. When 4 ohms is selected the the B+ voltage remains at the lower voltage setting that is 52V across the rectifier.
The owner uses a KEF Q900 the impedance rating is stated as 8 ohms and the amp was set at the proper setting for that, but I do have a feeling the relay would have clicked at his place too because the fuses which were blown was on the low voltage line.
I was able to get my hands on a scope and I measured the outputs of the DAC since the problem seemed to affect the left channel only and the HDMI inputs were used when the amp failed. I was hoping that there was an issue on the LPF and some high frequency content was going into the amplifier section. But this was not the case a clean signal was seen on the outputs of the DAC-damn! Can't think of what else to look for :S

[Audiorepair]

Have you checked that Speaker negative, input signal  ground, Power Amp ground and almost certainly chassis, are all connected together by zero ohms?

[Yamin]

Have you checked that Speaker negative, input signal  ground, Power Amp ground and almost certainly chassis, are all connected together by zero ohms?

Yep they all seem to be connected. However I have got the back panel removed at the moment some chassis connection are not made. I'll assemble the unit and re confirm.

The amp threw another curve ball today. Discovered this charring near to the +/- 15V regulators. The +/-7V for the R2A15218FP IC (used for switching between sources) is also derived from here. Couple of capacitors around that area had gone out of value and reading high ESR value. This was hidden under all the boards had to disassemble everything to discover this. Going to try and replace them all before assembling.

[TurboTom]

These are two zener regulators and the corresponding ballast resistors. They may get (very...) warm. If the voltage is on spec, the charring should be tolerable. Actually, you don't need that high temperature to cause the observed kind of discoloration over a few years' time. One thing that would worry me much more are the electrolytics in close proximity of the "hot" resistors. This kind of arrangement can almost be considered to be "planned obsolescense" since the caps will dry out over a few years and become high-ESR. This may or may not cause additional trouble, depending on their usage. I'ld probably simply go and replace them all, they are small enough to be no-brainers price-wise. Preferably, you should use 105°C rated types as replacements. Just noticed -- in the second photo, you already removed some. Right so!

Did you check the mains voltage present at you friend's home (also for surges)? Often, this consumer electronics isn't designed for a high margin of mains overvoltage. This may also contribute to too much dissipation in these shunt regulators and possibly contribute to the failure of the output stage. The reported 110V total supply voltage to the power amps appears a little high as well, though only marginally.

[dietert1]

It's a common design problem, selecting resistors with a nominal max power of say 2 W. Running at 1 W continuous the resistor body will reach several hundred °C. The resistor does not unsolder itself only because its wires are made of steel with low thermal conductivity. But you never know..
One nice mod i have done often is using a chassis mounted resistor as a replacement that dumps the heat into a large piece of metal.

Another common problem in many so-called high end amplifiers is the lack of low voltage transformer secondaries. The typical construction derived the +/- 15 V for the preamplifier from the +/- 60 or +/- 70 V supplies of the power amplifier, which caused those heat related problems, even when running the amplifier at low volume. A shame, since this was about saving some bucks during production of equipment to besold for thousands.

Regards, Dieter

[Yamin]

One nice mod i have done often is using a chassis mounted resistor as a replacement that dumps the heat into a large piece of metal.

That sounds like a nice Mod, would be great to see a photo if you can share . Is these resistors used as a fusable resistor?

Thanks so much everyone for the help. Learned alot! Just an update - I changed the caps which have gone out of value. The random relay clicking has gone. Come to think of it the random relay click could have been an issue of not installing the back cover as it serves as a ground connection. I have been using the amp and so far so good

[dietert1]

Those chassis-mountable resistors are aluminum housed and named RH010(ohms-value), cost about US$ 3 or 4 each. One example:
https://de.rs-online.com/web/p/widerstande-durchsteckmontage/8459655/?sra=pmpn

The 10 W version is pretty small and it is easy to find a good mounting position, maybe on the main heatsink of the amp.

Regards, Dieter

[Audiorepair]

Those chassis-mountable resistors are aluminum housed and named RH010(ohms-value), cost about US$ 3 or 4 each. One example:
https://de.rs-online.com/web/p/widerstande-durchsteckmontage/8459655/?sra=pmpn

The 10 W version is pretty small and it is easy to find a good mounting position, maybe on the main heatsink of the amp.

Regards, Dieter

I've always had a bit of an issue with this type of resistor.
They claim 10 Watts, but then if you look at the datasheet you will find this is only if they are atually chassis mounted.

In free air, they are actually rated 6 Watts at 25 C, and should they get hot, only 4.8 Watts at 70 C.

I've always thought claiming them a 10 Watt resistor is a little misleading at best.

https://docs.rs-online.com/b9cf/0900766b81410739.pdf

[dietert1]

What? He was asking for the chassis-mountable resistors i proposed. Those resistors are not meant to be used in free air.

Free air resistors are made from completely different materials, like ceramics, cement and the like that withstand 800 °C. But nobody wants such temperatures inside an amplifier, possibly close to electrolytic capacitors and semiconductors.

Regards, Dieter

PS: They are wire wound with welded copper terminals and low TC, typical 5 ppm/K, sometimes with a flat min/max at temperatures of 20 to 40 °C (selected ones). Also interesting for precision applications where you want to keep a resistor at a well defined temperature.

[Audiorepair]

Calm down, Dieter, I wasn't attacking you.

I was pointing out that on first appearance these appear to be 10 Watt resistors.

One might think one could buy one, stick it in free air, and it will withstand 10 Watts.

It won't.  That is my point.

[Yamin]

Those chassis-mountable resistors are aluminum housed and named RH010(ohms-value), cost about US$ 3 or 4 each. One example:
https://de.rs-online.com/web/p/widerstande-durchsteckmontage/8459655/?sra=pmpn

The 10 W version is pretty small and it is easy to find a good mounting position, maybe on the main heatsink of the amp.

Regards, Dieter

Thanks @Dieter for the link and info, so in the circuit what purpose does the resistors  serve? They are not suppose to act like fuses right?

[dietert1]

On your image Char1.jpg we can see that the circuit board gets very hot and is near black around the four resistors R4175.. R4177... Those are resistors that burn power when deriving the OpAmp supplies (+/-15 V) from the main power rails (probably around +/- 50 V or even higher). If you replace those resistors by chassis-mountable resistors, you take the heat away without creating high temperature spots. Nearby semiconductors and electrolytic capacitors will work better/last longer.

The bad design shown in your photo is present in many so called high-end amplifiers. The best way to avoid it is a mains transformer with lower voltage secondary windings (helpers). Chassis-mountable resistors are a way to fix the problem later on.

Regards, Dieter

[Renate]

... the lack of low voltage transformer secondaries. The typical construction derived the +/- 15 V for the preamplifier from the +/- 60 or +/- 70 V supplies of the power amplifier...

Yeah, the low voltage sections often seem an oversight.
OTOH, preamp sections take less current these days with low power op amps.

There is one silver lining to using high V for low V.
It gives you more time to see that the voltage is collapsing when the power cord is pulled and mute before the op amps start making weird noises.

[dietert1]

Yes, a lot of last century equipment did not include power status monitoring. They used  timers and the like, so decent behaviour on power on or power off or brown out is not guaranteed.

Regards, Dieter