Shoebox Sony TA-P7F + ST-P7J, yum, but how do you get DC offset to settle down?

After finding a matching tuner for my Sony TA-F55 (the one with funky volume control) I thought I was done with buying Sonys for a while. Well, that was until I saw a TA-P7F + ST-P7J set for grabs, and naturally I did grab!

The units arrived 2 days ago and I must say I love the sound. TA-P7F is one of the best shoebox amps I've heard, and it looks classy! The tuner's fine, too. Listening to this set is a real pleasure. I did not expect it to sound better than TA-F55, but it kind of does. It has all the clarity without the occasional harshness of the 55. And better bass.

Anyway, I tried to adjust the amp in accordance with the SM (and it did need adjustments). DC bias was as high as 30 mV in one channel, 3 times more than it should be. I adjusted to 10 mV - more or less, it fluctuates between 9.5 and 10.5, I figured close enough and left it at that. The amp got significantly cooler after the adjustment.

But DC offset was not that easy. I first adjusted on warmed up amp (1 hour or so after turning it on. It wasn't easy, but I finally managed to find trimmer settings where DC offset was within +/-15 mV and that was as good as I could get it. Like bias, the values were fluctuating, changing from -10 mV to +15 mV and then back again. I remember my friend had similar readings when he tried to adjust his TA-F70. I figured, OK, close enough, rechecked 15 minutes later (still the same). But then today I checked again after turning the amp on. The values were dropping from over 100 mV (in both channels). 10 minutes later they were around 60 mV and falling, so I suspect they eventually drop down to acceptable levels, but it takes quite long, definitely longer than 10-20 minutes. So is this a normal behavior for this type of amp? PPS, heat pipe, Hi-fT transistors and all?

Other than this suspicious behavior, I love this amp. Small, pretty, and sounds real good. Very quiet background for a 36 years old amp. Speaker terminals are crap, but that's my only complain.

And here are some pics, because, you know, worthless otherwise







And here's the other pair:

Greetings; Glad you got a nice deal.

My exp. w/Sony's with setting the bias is you set when cold as fast as you can as they will drift upwards unless there is a specified warm up period. There is no offset adjustment, at least I haven't seen a sony with that requirement in a SM.

Just out of curiosity was the VR pot direction clockwise or CCW for 0mv?

Hi. Not exactly a scroe story, I paid around $100 for each pair, but the units are worth it to me

TA-P7F does have both bias and offset adjustments. The SM is vague enough, saying adjustments should be made "several minutes" after power on. Here's the relevant page:



I just remembered that in a different FET amp I have from the same period (Inkel MD2200) DC offset should be adjusted with 8 ohm load connected, otherwise it goes crazy, much like here. But the Sony SM does not say anything about connecting load to speaker terminals before adjusting, so I didn't.

From what I had originally (more than +250 mV in one channel) the VR direction was clockwise i think, but the way it changed (the values sort of slowly followed the VR movement and stopped changing good few seconds after I stopped moving the VR) I had to adjust both ways to find the spot near 0 mV. Anyway, as far as I remember it was + left, - right. The PCB with VRs is joints side up, with openings to access the VRs.

Yeah CCW.. most models are I think.

Per the directions, Sony has always been vague about directions (**). The diagram can be interpreted as 'balancing' both meters to be equal in a japanese designers mind, as in balancing both channels. If you don't see an 'offset' procedure that's the only adjustment. I'd be more concerned about drift.

(**) Can you imagine adj. 5 bias pots to 6mv and only have 30 seconds to do it from a cold start? (sony av rec. w/surround) The same "Balance" wording was stated in that SM, too.

DC balance is obviously DC offset. The drift is what I'm worried about, although I don't know how worried I should be. I haven't seen this in other amps, but it does seem to be common for these PPS-powered Sonys. Maybe I'll dig out the TA-F55 later and check how it behaves. The adjustment procedure is the same as for TA-P7F, only the DC bias values are higher (22 mV). I haven't made adjustments in that one, it had other issues and a tech worked on it.

Let me reword this.. Two test points.. BIAS across test point leads and BALANCE between indicated test point and ground as shown in schematic.

(nippon thinking.. hate it when I do that.. )

The other version of SM I have for TA-F55 is in German. In that one they use the German term for DC offset, so it's quite clear. Anyway, I'll try to readjust, check the other amp and see how it goes. The TA-P7J might end up with my tech anyway, I noticed a volume imbalance between channels - slight but enough to annoy me now I know it's there.

In nippon thinking its balancing the two channels. that's why you have that annoying imbalance L-R or the R-L.. whatever.

Ideally a scope would- should be used.. seen a few tol sony's sm's that it was the only way to get certain models to get the ideal balance and mv. .. sounds like you have a little more gain on one channel.. make sure the balance knob and all other knobs are in their upright positions.. easy to bump.

basically if you can get the two reasonably balanced within 2mv on the test points? You shouldn't see a big difference audibly.. I'm saying if you get 8mv or 12mv and it balances.. no big deal. If you can get 5mv and it sounds good? take it! nother thing is VR's do wear out.

hope you get it fixed.

I think that the difference in gain between channels is elsewhere, preamp section probably. It stays the same regardless of DC offset and bias adjustment. The difference is there regardless of volume, so probably not a dirty contact. With dirty contacts, they usually work fine at high volumes when there is more current. And here the sound is centered at all volumes, DC offset and bias adjustments, all sources etc. when the balance pot is a bit to the left. Tonal balance of both channels is fine, no missing or weaker frequencies. All functions work for both channels.
It could be a bad resistor somewhere, or cap, anything. Without disassembling the amp and comparing the channels point by point I will probably not find it.

I was able to get DC bias to the same value on both channels. Offset still drifts, but it stays within acceptable levels.

VRs wear out and get dirty just like pots, I know, I've seen that before. I would use Deoxit on them (I did that with other amps that had 'jumpy' VRs with good results), but they're under the PCB with no easy access, and I don't have time to disassemble this amp now. It's really tightly packed, so that's not easy. Here's how it looks under the hood:





The four round holes in the paper screen are access points to VRs.

I did one more test. I played 1 kHz mono tone and checked AC on speaker terminals at various volume pot positions. There was a clear difference (about 10% and more at some positions) until about 2 o'clock position. After 2 o'clock things started to get balanced, with voltages perfectly equal in both channels from 3 o'clock to volume pot turned to the max. Then I did the same test with an external meter and 1 kHz played from an mp3 player on low volume, and than with music material. Right now I'm listening with volume pot at 3 o'clock, music playing from the mp3 player with its output volume turned down, and I have perfectly balanced channels.

So it looks like the volume pot is the culprit!

Sounds like you got it. Can you clean the pot?

I did. Getting to both sections of the volume pot is not easy in this amp, but I managed, I also cleaned the balance and bass pots. Treble pot was buried too deep.

Judging by meter readings it didn't remedy the problem completely, but I think it helped. The difference between channels seems smaller. Seems, because my meter is less sensitive than the one I used yesterday, so I'm not 100% certain. At full volume the readings are still the same. I'll do a listening test tomorrow. Plus it will give the spray some time to dissolve more dirt.

The pot is an obsolete and a bit strange ALPS so I would love not to have to replace it. Finding good replacement seems unlikely, and there's no space for a big blue ALPS.

After daytime listening and few more measurements: no joy, cleaning didn't change anything at all. Still about 10% difference between channels until volume almost at max. At least it's consistent and can be corrected with balance pot, so I can use the amp without much fuss until I decide what to do.

After measuring the volume pot it's confirmed, it is the culprit.

It is a 2 section volume pot. 2x250k controls volume before the tone amp, 2x20k after the tone amp. Looks like this:



I'm guessing the second section is there to attenuate noise from the tone amp.

The 2x20k section measures fine, the 2x250k has clear imbalance. The problem is, finding the original part or similar replacement is impossible, so I think the best thing I can do is to replace it with a plain 2x250k pot and bypass the other section altogether, or replace it with small resistors. The questions are: will it work and where to put the pot? Before or after the tone amp?

My instinct tells me to put it where the 2x250k section was originally, before the tone amp, and then cut the signal by 30-40% where the 2x20k section was (after the tone amp) to remove at least some tone amp noise. I know I probably will not have full power of the amp available this way, but I'd rather have a clean 30-watter than a noisy 50-watter.
I'm not sure which section of the volume pot affects loudness, I'm not good enough at reading schematics to figure it out. So that's a potential problem - I don't normally use loudness, but I would prefer it to work.

What do you think?

This is what it looks like on schematic:

Sorry to dig up such an old thread...

Please let me know if you decide to sell the TA-P7F and ST-P7J combo - i've been after these in good condition at a sensible price, for ages!

Cheers! Jim