Pioneer SX-850 with High DC Offset

[Brivan]

Hi, everyone

I have a Pioneer SX-850 that's exhibiting high DC offset in both channels (well, maybe). The receiver was service previously and Q5 had been replaced by soldering a replacement transistor to the remainder of the legs of the old transistor (sheesh). It was the correct type (exact replacement), but there was no voltage at pins 8, 9 & 10. The transistor tested bad and was replaced with a KSC2383. All power supply voltages now tested correctly and the protection relay engaged.

I was able to set the DC offset and idle current, but I noticed the idle current could only be set to about 20mV minimum (where it should be) and the voltage would shoot back up if I continued to turn the pot in either direction. I also notice an extra ground wire had been soldered between pin 20 of the flat amp to a nearby ground post. I removed that wire, being pretty sure it shouldn't be there. All the controls were then cleaned.

On the next power-up, the protection relay would not engage and I measured about 39v DC offset on each side of the amp. I checked pins 19 & 21 of the flat amp and pins and measured about 20v. I then separated the audio cable between the flat amp and the tone amp and measured about 4v on pins 19 & 21 of the flat amp and around 22v on pins 1 & 3 of the tone amp board. All these measurements were difficult to read, since the values would rapidly climb as I held the probe on the pin, so my readings are a rough average.

When I grounded pin 20 of the flat amp, the protection relay would reengage and the DC offset would stabilize to near 0v. I followed the schematic and found that the ground to pin 20 of the flat amp is actually supplied from the tone amp via pin 2, which is connected internally to pin 6, which leads over the the pre-out jack where it's connected to ground via a ceramic cap. This is where the confusion is - the schematic doesn't show the ground at that point as being connected through that capacitor, so I'm stumped, right now. I did try grounding the frame-side of that ceramic cap to see if the silly ground screw it was attached to had lost contact with the frame, but it didn't make any difference.

One more thing - when ground is removed from pin 20 of the flat amp (or pin 2 or 6 of the tone amp), the idle current jumps up about 110mV on each side. I haven't tied to adjust it in this condition to see if it acts more-normal.

Any help you could give would be appreciated. Thanks!

 

[markthefixer]

what does it do with the pre-out / power amp in jumpers removed?

also, can you get Hfe and Vbe's for each of the output transistors?

 

[Brivan]

what does it do with the pre-out / power amp in jumpers removed?

also, can you get Hfe and Vbe's for each of the output transistors?

Hi, Mark. I forgot to include that with my original post. Removing the jumpers makes no difference. It's odd, with the jumpers removed, I can still make the protection relay engage/disengage by grounding/removing the ground on pin 20 of the flat amp (or pin 2 or 6 of the tone amp).

 

[larryderouin]

Check the pre-out / Main in RCA's and ensure someone didn't internally jumper them.

 

[Brivan]

Check the pre-out / Main in RCA's and ensure someone didn't internally jumper them.

Hi, Larry

Yeah, that's what it acts like. Here's a photo of the pre out / main in connectors from the back side. The yellow and green cables go directly to the amp board, with no other connections to those points. You can see the capacitor from the ground to the frame ground screw.


I've done more isolating and testing and here's my results. I've removed the audio in and out cables from the tone amp, effectively isolating the flat amp, tone and main amp from each other. Most voltage readings aren't very steady - more of an average.

Flat amp with pin 20 ungrounded:
pin 19 (audio out to tone amp) = 4v
pin 21 (audio out to tone amp) = 4v

Flat amp with pin 20 grounded:
pin 19 (audio out to tone amp) = -3v
pin 21 (audio out to tone amp) = -3v

Tone amp with pin 2 / pin 6 ungrounded
pin 1 (audio in from flat amp) = 30v
pin 3 (audio in from flat amp) = 30v
pin 5 (audio out to preamp out RCS connector) = 30v
pin 7 (audio out to preamp out RCA connector) = 30v

Tone amp with pin 2 / pin 6 grounded:
pin 1 (audio in from flat amp) = 0v
pin 3 (audio in from flat amp) = 0v
pin 5 (audio out to preamp out RCS connector) = 0v
pin 7 (audio out to preamp out RCS connector) = 0v

Power amp with pin 6 / pin 21 ungrounded
pin 3 (audio in from main in RCA connector) = 30v
pin 11 (audio out to protection relay) = 13v
pin 18 (audio in from main in RCA connector) = 30v
pin 26 (audio out to protection relay) = 13v

Power amp with pin 6 / pin 21 grounded (protection relay engages)
pin 3 (audio in from main in RCA connector) = 0v
pin 11 (audio out to protection relay) = 0v
pin 18 (audio in from main in RCA connector) = 0v
pin 26 (audio out to protection relay) = 0v

 

[markthefixer]

edit:
in your original post, the "pot" being adjusted and the readings aren't explicitly named.
there is DC offset voltage in millivolts
there is Idle Current also a voltage in millivolts
there is the DC offset adjust pot VR1(2) at 10,000 Ω
there is the idle current adjust pot VR3(4) at 100 Ω
who does what to whom is not clear, and the worst could be assumed.


Have you looked at what the power supply voltages are doing while all this is going on?

with pre-power link broken, the power supply is your commonality.

Looking at the awr-101 board there are two ground arrays.
One is ground reference for regulator controls
The other is grounds for the power supply storage caps , also where power supply return current goes.

some boards run on + and - voltages, and the schematic lists + current and -current in the lines.
ground is just a formality and reference point

there's no -36.5v so it's power return is ground. The 51.5 is paired with the -51.5 (like in the power amp) and +28 paired with -19 in the phono amp.
you get the idea. Check all those boards, their power and ground requirements.
Then check how those requirements are satisfied. Eventually a hidden broken wire (or trace, or bad solder joint etc..) will be found.

Then be sure the voltages ARE getting to the boards, if a - voltage that is a current return is broken, then the return current could go, for example, on a shield of a shielded cable.

 

[Brivan]

edit:
in your original post, the "pot" being adjusted and the readings aren't explicitly named.
there is DC offset voltage in millivolts
there is Idle Current also a voltage in millivolts
there is the DC offset adjust pot VR1(2) at 10,000 Ω
there is the idle current adjust pot VR3(4) at 100 Ω
who does what to whom is not clear, and the worst could be assumed.


Have you looked at what the power supply voltages are doing while all this is going on?

with pre-power link broken, the power supply is your commonality.

Looking at the awr-101 board there are two ground arrays.
One is ground reference for regulator controls
The other is grounds for the power supply storage caps , also where power supply return current goes.

some boards run on + and - voltages, and the schematic lists + current and -current in the lines.
ground is just a formality and reference point

there's no -36.5v so it's power return is ground. The 51.5 is paired with the -51.5 (like in the power amp) and +28 paired with -19 in the phono amp.
you get the idea. Check all those boards, their power and ground requirements.
Then check how those requirements are satisfied. Eventually a hidden broken wire (or trace, or bad solder joint etc..) will be found.

Then be sure the voltages ARE getting to the boards, if a - voltage that is a current return is broken, then the return current could go, for example, on a shield of a shielded cable.

The last thing I was going to do today was recheck my power supply voltages, but I ran out of time. Just now, I started thinking about the ground issue and realized I didn't look upstream from the flat amp. Looking at the schematic, I see pin 5 of the flat amp is supposed to be tied to ground, but I never checked it. The schematic also shows pin 6 of the tone amp being tied to ground, but that connection doesn't exist (at least not on my receiver).

I won't be able to verify this until tomorrow, but I've got to be missing a ground somewhere in the tone amp or flat amp (the last person to service the receiver added a ground from pin 20 of the flat amp to a grounding post, which was probably okay, but I was suspicious and removed it. Now I want to find the ground that's missing and connect it properly. Right now, I'm hoping it's pin 5 of the flat amp. I may ground pin 6 of the tone amp, simply because the schematic shows it and it's probably not a bad idea.

Once I get that grounding issue straightened out, I might be back to the goofy idle current adjustments where I can get it down (barely) to 20mV on both channels but, no matter which way I rotate the adjustment pot from there, idle current goes up.

 

[markthefixer]

the amp uses 2sa726 transistors. are they still in there?

to disconnect the dc offset adjust open up R19 & R20

Get Vbe's and Hfe's for all your output transistors. I suspect some strong Vbe mismatches.

If you can't get the Vbe , measure the 0.5 ohm resistor voltage which will show current flow's for each output transistor at that 20 mA null.

Then repeat for an increase of pot, and then for a decrease of pot.

the factory idle current reading is a resistor sum of those 0.5 ohm resistor voltages.


the real stinker is that it's supposedly happening in both channels.
make sure the big cap voltages feeding the drivers and outputs is correct.
That's the only commonality...

 

[Brivan]

the amp uses 2sa726 transistors. are they still in there?

to disconnect the dc offset adjust open up R19 & R20

Get Vbe's and Hfe's for all your output transistors. I suspect some strong Vbe mismatches.

If you can't get the Vbe , measure the 0.5 ohm resistor voltage which will show current flow's for each output transistor at that 20 mA null.

Then repeat for an increase of pot, and then for a decrease of pot.

the factory idle current reading is a resistor sum of those 0.5 ohm resistor voltages.


the real stinker is that it's supposedly happening in both channels.
make sure the big cap voltages feeding the drivers and outputs is correct.
That's the only commonality...

Today, I tried tracking down grounds to pin 5 of the flat amp, but there was nothing there (with exception of the original ground wire from that pin to the nearest ground terminal that I removed, since it looked like it was soldered in by an amateur). I finally Googled the SX-850, looking for images of the tone amp and flat amp boards which would hopefully show a ground wire connected to one of them. Although none of the images were crystal-clear in those areas, they were clear-enough for me to make out what looked like a ground wire from pin 5 of the flat amp to the ground terminal next to it. So, it turns out that original ground wire I removed was supposed to be there - it just looked like a hack job. There was a couple other solder connections that looked dubious, but were obviously untouched, so maybe somebody in the factory assembly line was having a bad day, that day. I reinstalled a ground wire to those points.

Next, I rechecked the idle current on both channels. Oddly, the the left side was now adjustable down to 0v, but the right side was still shooting up on either side of the 20mV minimum adjustment. I had an epiphany - what if those pots are flaky? I ran the right side idle current pot back and forth several times and tried, again. Voila! I can now reduce it to 0mV. The fact that both sides originally shot up in either direction of 20mV threw me off. I exercised and cleaned all four adjustment pots on the amp board and reset offset and idle current. All is well.

Thanks to Mark and Larry for your help!

 

[markthefixer]