Sony TA-3200F Recap, my first recap!

DrTebi

Active Member
First of all, a big thank you to everyone here at AudioKarma, I couldn't have done it without you!

It was here where I gathered most of my "vintage amplifier repair knowledge", and what gave me the courage to just go ahead and try to recap this Sony TA-3200F...

I got it for under $100, but after running it for a minute or two, it started making "alien noises" (as my wife calls it) in the left channel. A bit of deoxit helped with the scratchy volume knobs, but the noise didn't go away.

Last week I finally decided to just try to recap it myself. If I would screw up, at least it wasn't a very expensive amp I would have ruined.

I examined all the caps, wrote down the values and brand names (all Elna and Nichicon), and found most of the exact same at Partsconnexion. Only two values were a tad higher in voltage. The two huge caps with the Sony lid I decided to keep though (I think they are hard to find anyway).

It was probably not necessary to replace all of the capacitors, and I understand that I did not need the exact voltages or brands either... I also did not have to use "audio-grade" caps... but in the end it was only about $20, so I figured why not. Here is the parts list:

1 x Elna Silmic II Capacitor 220uF / 6.3V, RFS Series
2 x Elna Silmic II Capacitor 470uF 16V
2 x Elna Cerafine Capacitor 1000uF / 10VDC
2 x Nichicon-023-Series KZ 10uf
4 x Nichicon-031-Series KZ 47uf / 100v
2 x Elna RFS Series 0.47µF / 50VDC
4 x Elna RFS Series 33µF / 25VDC

The power supply board with the original caps:
Sony-TA-3200F-01.JPG

... and replaced caps:
Sony-TA-3200F-02.JPG

I used desolder wick to get the caps off. Tried my best to get it all cleaned up, also used some acetone to get rid of some gooey stuff that was all over the board:
Sony-TA-3200F-03.JPG

And here my first solder joints... not amazing yet, but I got better later on!
Sony-TA-3200F-04.JPG

Some caps were smaller (although the exact same values), so a bit of bending was in order, using the old capacitors as a gauge:
Sony-TA-3200F-05.JPG

Here the "before" picture of the amplifier board:
Sony-TA-3200F-06.JPG

And the "after" pic, board installed:
Sony-TA-3200F-07.JPG

And now he is happily singing again:
Sony-TA-3200F-08.JPG


It was quite a feeling when first turning it on again... I am sure some of you must remember this, the first time re-working an amplifier and turning it back on, bringing up the volume... well, it worked, the noise is gone! I am pretty excited. Just hoping that this did the trick, and nothing else is wrong with it; played it today for about an hour, and it sounded nice and smooth, and super quiet when no source is playing.

The plan now is to use it in my active 3-way open baffle system (still under construction) for the mid-range or tweeters.

This was my first dive into electronics, well, besides RCA cables and a passive crossover I once built. It's a lot of fun, and I am planning to do more. Next on the list is Rod Elliot's P125 active crossover... which will be challenging for a beginner I suppose, but I guess I should keep thinking "who dares wins" :)
 
Nice work! I have one of these I recapped. Yours looks to have a different amp board. The board I have has two TO-220 type drivers. They run hot and had no heat sinks. I replaced them with some Fairchild's and heat sinked them. Yours looks to be a more robust design. I like these amps. Fused, a pretty good limiter curcuit and thermal protection. The thing has more bells and whistles than most other power amps.
 
I have one that has been made a complete recap and I really like the full-bodied dynamic sound and detailed
 
This series had a pretty good run. Yours also looks to have better traces than mine. Not much in the way of masking on the boards letting them lift easily with mine.
 
Thank you for all your kind words.
It is indeed very satisfying!

Nice work. Did you check bias and DC offset?
I wanted to check the DC offset today. I do not know how to check the bias... I am new to this, so I still need to figure a lot of things out!

Nice work! I have one of these I recapped. Yours looks to have a different amp board. The board I have has two TO-220 type drivers. They run hot and had no heat sinks. I replaced them with some Fairchild's and heat sinked them. Yours looks to be a more robust design. I like these amps. Fused, a pretty good limiter curcuit and thermal protection. The thing has more bells and whistles than most other power amps.
According to The Vintageknob there are two or even three different versions of this amp. I was already confused when I looked at a service manual from a different version, which has an extra board that mine doesn't have at all?!?!

By the way, this amp did not run hot after playing at about 1/2 power on one channel for an hour or so. It's warm, heat sinks can still be touched. I could check the temperature if anyone is interested.
 
One question:
It this amp "capacitor-coupled or transformer coupled"? If so, does the DC-offset measuring still work as described Amplifier Distortion, DC-Offset, and You! ?

When measuring the right channel, I get a pretty steady 9.1 mV. On the left channel however, it jumps around from 0 to up to 35 mV.

Does that mean something is wrong with the left channel? I did notice that one diode looked a bit odd... couldn't tell if it was broken or not (lack of experience :)).
 
It's not cap coupled, it's a DC amp, so the offset is relevant. Before investigating the offset jumping, I would verify the bias (idle current) is correct.
 
One question:
It this amp "capacitor-coupled or transformer coupled"? If so, does the DC-offset measuring still work as described Amplifier Distortion, DC-Offset, and You! ?

When measuring the right channel, I get a pretty steady 9.1 mV. On the left channel however, it jumps around from 0 to up to 35 mV.

Does that mean something is wrong with the left channel? I did notice that one diode looked a bit odd... couldn't tell if it was broken or not (lack of experience :)).
Without any signal or source In, put a meter across the speaker terminals. 0 to 15mv should be ok. Even 30mv isn't horrible. I would think the dc offset would be very close or the same for the different versions. The extra board in one version may be an international version with voltage selector. Yours looks to be a later version with a better amp board. Mine is an early version using proprietary parts, like the drivers that ran hot enough to burn the board itself. New modern replacements with heatsinks solved that problem. One other note, there is one tantalum capacitor on the power supply board and two right at the input of the amp board. I replaced these with low esr electrolitics.
 
Does the dc offset jump around even after about 20 minutes of being on? It can take a few minutes to settle in, but thats kind of a wild swing. You may want to check those resistors that looked like they may have gotten hot along with that diode. If the trim pot is dirty or has carbon build up you could try cleaning the one for that channel's offset setting. And adjust it with your meter connected to the speaker terminals. Mine had some build up, but they cleaned up nice.
 
Generally when I've seen DC offset swings like this on these units, the fix always required changing out some semiconductors - like the diff pairs.
 
I think it was running for ten minutes and still showing the "swings" on the left channel.

Now how do I adjust bias? Sorry, I am new to this... any pointers to a thread or tutorial explaining that would be great.

I will also run it again for longer, and see if I still have DC offset swings.
 
I think it was running for ten minutes and still showing the "swings" on the left channel.

Now how do I adjust bias? Sorry, I am new to this... any pointers to a thread or tutorial explaining that would be great.

I will also run it again for longer, and see if I still have DC offset swings.
The bias adjustment procedure is described in the service manual.

If you haven't yet download the service manual, the link in NAD80's post above will get you to one source.
You can also download if from hifiengine. (You need to register, but it's free.) Hifiengine has 2 versions, one listed as early model and one as later.
 
The service manual will tell you where to attach your test leads for bias and dc offset adjustments. Two trim pots per channel. One for bias and one for dc offset.
 
I checked again today after running it for an hour or more. Now the left channel settled at about 0.00mV. So that looks good I would say? I noticed that the jumping occurred again, but only when I held the probes with my hands... once stuck into the speaker terminals and "hands-off", it settled to about 0.
The other channel was somewhere around 6mV.

I suppose these values are pretty good, and I should probably not mess with the bias and DC offset now?
 
Back
Top Bottom