Toshiba SC335 power supply capacitors

A couple of weeks ago, I started a thread regarding replacing meters in a Toshiba SC335 power amplifier and the new ones should arrive in a week or so.

But in the meantime, I replaced all the small electrolytics since they were accessible, new ones were easy to get and inexpensive. Although I have no reason to suspect there's anything wrong with the 50V 10000uF power supply caps, I was wondering if it would be worthwhile looking for new ones.

The originals are a radial single capacitor but with 4 legs. The reason seems to be that the legs support a small PC board with the bridge rectifier and the whole affair is held onto the chassis by a clamp around the caps with 2 screws into the sheet metal.

I have not exhausted all suppliers but a capacitor of that or slightly higher value in the 50V, 60V, 63V, 75V, etc. range does not seem that common - especially one with 4 legs.

Even though they are 35 or so years old, I am tempted to leave well enough alone and if it ain't broke, don't fix it. But if anyone has some input on the likelihood of these things failing, I will step up the search.
 

Ian Finch

Member
Paid Member
2010-04-11 4:22 am
Coffs Harbour, NSW
If the caps are original, they're coming up to 40 years old. Not many would would suggest they will still be as good as new. Test the value and ESR if you like, but replacement is almost certainly necessary.

The 4 legs (3 will be either -ve polarity or no connection) are for mounting security since only 2 small pins or wire leads would be prone to pull out or fracture in transit, when handled roughly etc. Snap-in types with only 2 heavier, strongly attached pins are specified now, with a generous bead of glue applied around the bottom edge too. Without a manual to see the specification though, it's down to your description to know exactly what you have, how it fits, connections, dimensions etc. Modern caps will be much smaller than these though.

There are a few multi-pin caps, mostly with snap-in pins, that are still manufactured but the type is near extinct and since NOS caps are likely useless, you need to modify the mounting connections to suit standard snap-ins and methods. http://www.nichicon.co.jp/english/products/pdfs/e-term01.pdf

For really large caps, its back to mounting clamps, screw terminals and $$$.
 
Last edited:

irribeo

Member
2013-04-01 2:46 pm
The 10.000uF have two pins with a bracket on each to double them.

All the 30 to 40 year old Macron electrolytics in sc335 I saw had capacitance 10 to 15% above foilvalue and esr well below datasheet. The two for input and two for feedback were relatively high esr, closer to datasheet than all others that like just said, far below datasheet esr.

ce04w ... - Datasheet Search Engine Download
 
Snap in capacitors are unfamiliar to me so had to look up what they were. Has there been some sort of standardization implemented with regards to hole spacing on PC boards?

There is a somewhat local store with 63 volt 10000mfd Rubycons for about CDN $10 each. I hope to pop in there sometime soon to check their dimensions. The PC board they would mount is not crowded and the traces are wide enough I think to permit accommodating a different hole pattern.

There is still the matter of that PC board being suspended from the original capacitor's 4 leads! but as was suggested, some (hot) glue should accomplish the same thing.

I'm assuming that the diameter will be less so I guess I'd need to be creative in using the existing clamp that affixes the whole affair to the chassis.
 
What I see on the trace side of the board is two leads of one cap soldered to a large section of the same trace with no wire attached to either. The other two leads are on separate traces with a red wire wound onto one and an orange wire wound onto the other.

Would this not indicate 2 pos. leads and 2 neg. leads per capacitor and that the 4 leads are for structural purposes?

There is more going on with the 2nd cap in that each lead has a dedicated trace and more wires are involved but my assumption was that it minimized the number of connections per lead.

Since multi lead capacitors do not seem likely to get, I thought a conventional cap with two leads and some "bridging" would be the order of the day.

I am unsure what the bracket between two leads might be. There are no brackets visible on the trace side of the board which is the one I can see. The caps are mounted tight and flush to the board but if there's a bracket or a jumper on the underside, I would not see it without unsoldering and removing.
 

irribeo

Member
2013-04-01 2:46 pm
Or don't think the pcb will be tight to cans but really look, outside your head. There is a gap and light can come through, so you can have a look. But I also told you already the cans have 2 leads going into the can and you see 4 pins above pcb, guess what is under pcb between cans and pcb :D
 

irribeo

Member
2013-04-01 2:46 pm
It is also mentioned in another thread by a another canadian at this website: Recapping vintage Toshiba poweramp. He also questioned for weeks befor he found how the caps looked when he removed them. He also never measured the old caps, I think you could be surprised if you have somebody compare the grey little ones you already removed to for example new Panasonic FC's same values. Very surprised, I was.

Oh, and, the stock caps, after being removed, were simply two pins that branched off into four. I just had no idea! Should have taken the old ones off earlier, and I would have known.
....!
 
I had planned on removing at least one of those capacitors tonight, time permitting. There's not much of that since work days are 14 ~ 16 hours. I inspected the originals more closely this morning and the top of one is bulging slightly. Not obvious to the eye, but it can be felt.

As was recommended, I had also planned on checking the terminals with a meter. I have been fairly certain I was dealing with twin + and twin - leads per cap.

As it happened, my work took me nearby the parts store so I took advantage and picked up the new ones. During the day today I imagined a couple of ways to mounting them but must have everything in front of me to see what might work.
 

irribeo

Member
2013-04-01 2:46 pm
Check if tr5&tr6 have been replaced, most likely you need to see twisted transistor leads there. If you are preparing the Toshiba for another 20 years you will need twisted transistor leads, those original transistors do not last in any vintage amp. Inputpairs can be replaced tr1-4 for that 20 year task too. Iddle current adjusted to 10-12mV.
DC offset can be set with R9 and R10, with matched inputpairs you'll see negative dc offset and by reducing r9/r10 you could bring that to single digits.
 
Check if tr5&tr6 have been replaced, most likely you need to see twisted transistor leads there. If you are preparing the Toshiba for another 20 years you will need twisted transistor leads, those original transistors do not last in any vintage amp. Inputpairs can be replaced tr1-4 for that 20 year task too. Iddle current adjusted to 10-12mV.
DC offset can be set with R9 and R10, with matched inputpairs you'll see negative dc offset and by reducing r9/r10 you could bring that to single digits.

I'll have a look at those transistors. Twisted leads?

You must pardon my naivety but unless something adversely affected them, I was under the impression a silicon transistor had an indefinite lifespan. I had not considered replacing any since the amplifier worked properly (or seemed to) when I first got it.

I am going to put the new power supply caps inside the original cans. The cases uncrimped from the bottom quite quickly and the old innards came out easily. Now I must devise a method of attaching the new cap's stubby little leads onto the old one's base.
 

irribeo

Member
2013-04-01 2:46 pm
Yes that transistor always fails and cause other transistors to fail with them. Reliable replacements have different pinout, thus twisted leads. Replacing the transistor for identical pinout transistor probably won't help, fake or equally unreliable are your options there :)
Accuphase, Akai, Marantz, Pioneer, Sansui well you can find warnings about that transistor in any amplifier. They fail reliably ;)
 
I see, twisted leads = different pin out.

I can't look at the schematic till much later but 2SC458's are they?

An Akai r2r I once owned had a problem and I read about that transistor's reputation and the deck I had used 14 of them so I replaced them all. In reading about it, I thought it interesting that so many manufacturers used that transistor. If what I read was correct, it was made by Hitachi and had a particular case style. Supposedly, either before or after, Hitachi used a different style case for the same transistor and those ones were fine. In the case of those I dealt with, the legs oxidized and had gone totally black.

I can't recall what I used as replacements but they had exactly the same pin out. They may have actually been NOS of the same transistor but of the non affected variety. One of the local shops I buy from has had on occasion, new parts that were long out of production.

Care to share what you would recommend or have used as a substitute?

Whenever I've tried deciphering data sheets, I get lost. Once or twice when I have settled on something, I find out that it is out of production as well
 

irribeo

Member
2013-04-01 2:46 pm
tr5/6 are 2sc1628(-y) original
I remember looking at what I had in stock, and several could fit. I think I used a Toshiba part. All have different pinout except some small Zetex I believe?

bd139, mje15030, but like said I think I used Toshiba here, 2sc3423 maybe, probably obsolete too LOL
 
I don't have their number handy at the moment, but new equivalents for the 4 2SA841's (TR1 to 4) are available locally for 50 cents apiece.

In order to get two matched pairs, do I buy a handful and start comparing with a DMM?

Those two 2SC1628Y's (TR5 & 6) have eluded me - even a data sheet. The only sheet I found online was in Korean or Japanese. If you can suggest a replacement for those, it would be appreciated.

I have not had the time check any of the original transistors. But if your experience says they can be on their way out even though testing fine, I'll put in new ones.

I may not be keeping this amp a long while, it may get given to a friend along with a Hafler pre I'm not using. She has an entry level Yamaha RX396 or something but with decent speakers. She moved into a big house recently and the Yamaha should be improved upon.
 

irribeo

Member
2013-04-01 2:46 pm
Many replacements are mentioned across the web for 1628y. It was npn, 150V, 120Mhz, gain ~200, Cob 50pF. Rail is just below +/-40V probably, so look what you got or what locally is available and is bigger than to92 package like to126. The twisting isn't really much because base pin on pcb is 5mm or more behind emitter and collector, so more like slightly bending here.
 
Finding new for old transistors is really time consuming!

I found a thread on AK where someone used Fairchild 2SC3503 to replace those same 2SC1628Y's. I checked the local parts store's website then phoned them, asking if they had them or something similar. Answer was no, that it too is obsolete but their supplier still had stock so I special ordered a couple.

That same parts shop were who showed stock on hand of 2SA970 which I intended to put in place of those four 2SA841's. I asked for some of those to be set aside for me and I'm thankful I asked. Seems that those are also discontinued but if they do have some leftover as they're supposed to, I can pick all up together.

The power supply capacitor stuffing is going well. The new ones are 25 to 30 % smaller so they fit nicely inside the old cans.