Sony TA-5650 v-fet matching and 1N4148 diodes

I got the bug to try a v-fet and Sony vintage equipment, in general, from the many enthusiastic posts here. I recently purchased a working TA-5650. Of course, I've combed many of the threads here while anticipating its arrival. I also acquired a service and users manual. If any of you can help clarify the following questions I would appreciate your response. The Reader's Digest version is at the bottom.

Upon initial inspection I saw that two (D301 & D351) of the trouble prone VD-1221 diodes had been replaced with two 1N4148's in series and that the other two (D303 & D353) had not. I purchased four 1N4148's to replace these. The ones I purchased look a little smaller than the others previously installed and I observed on Digikey that there are ,seemingly, many types of 1N4148 diodes. The ones I purchased are 200A, 100V in a DO-35 package. Will these work or is there a better "type" of 1N4148 for this application?

The next three questions relate to the V-fet power devices. Please see my photos of the four 2SK60 devices. The first photo shows them as they came. It seems obvious that someone has fooled with them previously. This is evidenced by the excessive grease and non-matching mounting hardware. Of particular interest/concern to me was that one of the 2SK60's was not labeled like the other 2SK60's nor the 2SJ18's. The odd one (second from the right in the photo) only has one letter prior to the "rank" number and no hyphen in between. It also does not have "Japan" printed on it like the others. However, the casing appears to have identical shape and construction of the others. Is this a latter replacement, counterfeit or... ?

I've read a post concerning matching the output devices in that each pair must match in rank or be no more than one apart. All of my 2SJ18's are rank 54. The 2SK60's consit of two rank 54 and two rank 56. Oddly, as you can see in my first photo, they came mixed with one in each channel instead of the 54's matched in one and the 56's in the other. I'm assuming I should make the swap for matching pairs in each channel. Advice/precautuions? IIRC did read something from Echowars about the need to change some other components (resistors?) If different ranks are substituted.

I did install the output devices that way (see second photo) and replaced the two VD-1221 diodes with the series 1N4148's. I have not powered it up yet and will wait for the sage advice that typically is shared here. I'm in no hurry to ruin my output V-fets. I also installed silicon rubber insulators under the V-fet outputs instead if using the old mica type with grease. Is there any potential problem using these insted of the mica ones and grease?

I do plan on doing a full re-cap and was wondering about increasing the filter capacity of the main caps. If I can find ones with increased capacity (uf) that fit properly (clamps & PCB) of good quality is this of benefit? It seems like may DIY folks are always increasing the main caps if possible. Is this always a good thing as a rule of thumb? Are there other caps in the TA-5650 that will benefit from substitutions different than the original values? I plan on using nichicons from Parts Connexion while they've got the sale going and since they sponsor this forum. thanks in advance for anyone who can advise.

Reader's Digest Summary of questions:

1. Is there a certain type/spec. 1N4148 diode that must be used to replace the VD-1221?

2. How can you tell if you have a genuine Sony v-fet output transistor?

3. Can I swap my non-matching 2SK60's to have matching pairs and not worry about it or should I take certain precautions needing to make some other parts changes to accomodate?

4. Is it OK to use silicon rubber insulators instead of mica with V-fets?

5. Can the main filter caps or other electrolytics benefit from choosing higher values?

The third and fourth photos show the front after a good scrubbing. She's pretty clean now!

I've nothing to add except proceed with caution~
You seem to have thought things out well. I sure wouldn't want to ruin that fine amp!

Jeff W. said:

Reader's Digest Summary of questions:

1. Is there a certain type/spec. 1N4148 diode that must be used to replace the VD-1221?

Click to expand...

A 1N4148 is a 1N4148. They should all be 100V 150mA devices.

2. How can you tell if you have a genuine Sony v-fet output transistor?

Click to expand...

It's tough to fake a VFET (but I'm sure the Chinese have either tried, or possibly succeeded). If it measures properly out-of-circuit, and doesn't self-destruct when powered up, it's probably a good part.

3. Can I swap my non-matching 2SK60's to have matching pairs and not worry about it or should I take certain precautions needing to make some other parts changes to accommodate?

Click to expand...

I'm not sure I understand the question. I will say that the range of pinchoff voltages on VFET's is huge, and on units where multiple output devices are used, they must all match in order to prevent one device from conducting all the current (and thus negating any current-sharing, which distributes the load). In a pinch (), you can have a VFET that is perhaps one rank different from the others, but it's a non-ideal situation. Looks like you may have some that are two ranks away. Not good...

4. Is it OK to use silicon rubber insulators instead of mica with V-fets?

Click to expand...

I don't much care for them. Too easy to puncture them and end up with a short. I prefer a nice, thin mica insulator and new grease.

5. Can the main filter caps or other electrolytics benefit from choosing higher values?

Click to expand...

You asked a general question, so I'll give you a general answer: no. New caps, same values. Anything that might cause the pinchoff voltage to rise too slowly can mean the death of the VFET's.

Thanks Echowars.

Q1: My 1N4148 diodes came in a bag marked 200A, 100V DO-35. Maybe mislabled? I'm showing a photo below of two of them installed in the D303 position.

Q3: refers to having received this amp with one channel that has one 2SK60 of rank 54 and one of rank 56. The other channel is the same (a 54 and a 56). Wouldn't it make sense to swap them so that one channel has two rank 56 2SK60 devices and the other has two rank 54 2SK60 outputs? The first photo (first post) shows how they were installed when I received it. The second photo shows them switched having two rank 56 on the left and two rank 54 on the right. Do I need to be concerned about swapping them for any reason to this, seemingly, more logical matched configuration?

Thanks for the advice on the insulators. I have mica and grease. Will do.

Do you have a main filter cap replacement you could suggest? The original is 10,000uf 50v, 40mm diameter X 70mm L. I would guess you've replaced a few of these. Most newer ones of like value are smaller. Thanks again.

Q1: Never heard of a 200A 1N4148. Does that itty-bitty device look like it can take 200A? Question...why the long-arsed leads on your dual diode replacement? If it's for supposed thermal conduction from that transistor below it, that method ain't gonna work.

Q3: Transistors on the same channel need to be matched according to rank. If that's what you've done, cool. That's the way it's supposed to be.

The cap situation is wierd. Sony used a cap that doesn't match anything. The stock cap is a screw-terminal cap 40mm in diameter, but with a non-standard screw spacing. The screw-spacing is very close to a 35mm cap with screw-terminals, but 35mm screw-terminal caps are rare. But...Newark has a very good 10,000µf 50V United Chemi-Con screw-terminal cap measuring 35mm in diameter x 67mm tall. If they aren't lying, and the cap is really a screw-terminal cap, then all you have to do is to increase the diameter at the bottom of the cap with tape so it'll fit in the cap clamp, and you will possibly need to enlarge the holes on the PC board with a small file so you can get the screws threaded in (they will be close, so I'm told by someone who did this, but not quite close enough to get the screw in without a little filing). The Newark part number is 95F4455. And the old screws will probably be the wrong thread, so a trip to the hardware store might be necessary (do not get too long of a screw, or you will rupture the screw mount and destroy the cap...the screw should not bottom out in the terminal).

Q1: Yea, I was trying to duplicate how the original was set up to disipate heat from the transistor below. How can it "work" or is it just worthless to bother? I know in another thread you mused about it really being effective anyway.

Thanks for the lead on the cap, I'll look it up and heed the warnings. Much appreciated!

Long leads on the other side of the diode...so the diode is in contact with the transistor (the way it is, any possible heat picked up by the leads is dissipated before it reaches the diodes).

I don't think this is terribly important. I'd clip the leads and not sweat it too much.

Ah, got cha. I'll clip 'em.

I ordered the Nippon Chemicon main filter cap replacements from Newark this morning. I'm fotunate in that I'm only one interstate exit from their sister company (MCM) and they'll get them in for me with no shipping charge! For those that may be interested, Newark has a special price promotion on them currently. They are reguarly $9.28 ea. and currently they are $7.60. There are hundreds in stock here: http://www.newark.com/united-chemi-con/36da103f050am2a/aluminum-electrolytic-capacitor/dp/95F4455 so it's a good time to grab them and freshen up your TA-5650 main caps.

Cap voltage ratings

Echowars,

One more question on cap values please. I know you advised to stick with original values. Is this critical for BOTH microfarad and voltage? I am searching for caps now and am finding correct microfarad values, but need to go UP on voltage rating for some to get what is not available to match the original value. My understanding is that this is not a problem on typical solid state equipment, but will this have an effect on the rise time of the pinchoff voltage or is it only the microfarad rating? Thank You.

What can't you find? AFAIK, all the caps in the 5650 are pretty standard values.

I should have been more specific. As I mentioned in the last paragraph of my first post, I would like to purchase Nichicon capacitors from Parts Connexion. The reason for this being: I would like to get the best Nichicons available and they carry them (series KZ & FG), They sponsor this site, and I'd like to limit the amount of different places I have to order from (and pay multiple shipping charges). I've already discovered that I'll need to get some capacitors from Mouser in the Nichicon KW series because P.C. does not carry some microfared values (.47uf, 3.3uf). I know Mouser carries some KZ and FG values, but they don't have as many as P.C..

I'm sure you are right in that I could get all the exact voltages if I mix more of the different Nichicon series together or go with a lower grade series. I was just wondering if I needed a 47uf-10v and P.C. has a 47uf-25v (and it fits) can I use it without worry, having to source it elsewhere and/or use a lower grade series? I don't mean to be difficult. I would like to have a better understanding of which (uf or v) can effect the rise time of the pinchoff voltage or if it is both.

I know you (Echowars) have worked on these. Would you share what brand and series of capacitors you would use in this unit? IIRC (actually I printed this info and saved it!) you like Panasonic FC & FM and Nichicon UPW & UHE. Are these a better choice than the highest grade "audio" Nichicons because of the high temperature rating and long life? Ultimately, I want my TA-5650 to sound the best it can and be reliable. I don't mind spending a little extra to get it there. However, I don't want to waste money if there is nothing to be gained from the using the Nichicon KZ's & FG's. If exact voltage value matching is critical, of course I want to do it. I'm just an amatuer trying to learn and bring out the best in this little gem. I appreciate your responses.

This pretty much sums it up: http://audiokarma.org/forums/showpost.php?p=2526347&postcount=12

Signal Path?

OK, thanks. I don't know enough to know which are in the "signal path". I know it is too involved to cover with a quick answer, because this has come up before for me. However, I see in the manual that the B+ and B- patterns are highlighted. By chance, are these the signal paths? Any hints? :scratch2:

Signal path is just that..........the signal from any source input, through the preamp & amp to the outputs to the speakers. B+/B- are the power-supply high voltages.

The only signal path caps in the 5650 are already Mylar, and should not be changed.

Re: Signal Path

It might be easier for we electronics newbs to understand where to look for signal path caps by defining what is not signal path. It seems to me that signal probably runs through about everything other than the main power caps, but that's just a guess.

Thanks for the input on the signal path gentlemen. Echo, thanks for cutting to the chase. If anyone else can reference some material on this subject for the uninitiated I'd be grateful. MX117.MC7150, your explanation sounds logical. I'm sure there are plenty of nuances, but it gives me a place to start.

In order to understand what is, or is not, in the signal path, you first have to understand how a particular amp stage works, and how it is passing the signal along to the next stage. This is not something I care to tackle through forum posts...it is something probably best addressed through electronics classes at a local university or community college, although you can certainly learn a lot though books and online sources with a bit of added self-discipline.

Understood. I'm grateful for all the time and knowledge you've shared with me thus far. I'm currently compiling my order for Digikey during my lunch hour!

Back to work now. It always gets in the way of fun! Although, it sure helps in affording stereo equipment and parts (not to mention food, shelter...)!

Re-capped and running!

I've got the TA-5650 running this evening. I had finished the re-cap last week with the exception of the main filter caps. I finally got them yesterday and installed them last night. They are screw terminal caps and came with the screws! I used a good quality foam with an adhesive back to increase the diameter of the bottom of the caps to fit the original mounting clamps (pics. below). This worked quite well. As mentioned previously, the originals were 40mm in diameter whereas these are only 35mm. Most important was that the screw lugs were very close to the originals in spacing. They were just slightly closer together which did require a little bit of filing on the inside of each mounting hole as Echowars forewarned. However, this was a minimal task that only took a few minutes on each hole with a fine, jewelers type, rat-tail file.

Below are photos of the power supply board and driver baord with the new Panasonic caps. I found panasonic FC's for most and some FM's. For a couple values I had to go to the Nichicons because the Panasonic values were out of stock. I'll post the complete list of caps for each board soon for the convenience of others.

I ran through all the adjustments outlined in the service manual:

1. (2-1) I adjusted RT401 to get 20v at the emitter of Q403

2. (2-2) I confirmed Q401 showed 97v (96) at the emitter

3. (2-3) On the DC balance check I'm a bit out of spec. on the right channel. It measures 56-57mv. The left channel is at 4mv.

4. (2-4) I went ahead and adjusted the DC bias to a 50mv reading for both channels. This did not change the DC balance readings above. Any advice?

I know Echowars mentioned taking readings from locations other than those shown in the service manual, so in the meantime I will seek out this thread. I'm so close here and I'd really like to "dial it in" to specifications. Any assistance would be greatly appreciated and I'll "dig" in the meantime.

I did hook her up to some old Rat Shack minimus 12's and all sounds good. It's great to have this baby to this point. It's been a lot of fun working on this nice piece of equipment. It's modular design and layout make it very easy to get to everything without a struggle.