Pioneer SA-7800 - Help replacing transistors!

[groenhen]

According to Sanken's official site, there are only 2- 3 Authorized Distributors:
- one is from Japan
- another one is digikey (USA )
Unfortunately, both above distributors have huge P&P costs (to Europe).

There are also Non-Authorized Distributors and as I'm located in Europe, I am interested what do you know about www . littlediode . com ? Are they selling 100% genuine parts? ...
On a second look, they say nada about the hfe of those complementary outputs

 

[stevieg330]

Littlediode

Hi groenhen,

I'm based in the uk and have used littlediode for a few of the parts on my SA-9800 including the obsolete 2SK34 fets. Everything has measured and worked fine so while I can't say for certain that everything is 100% genuine, I have had no problems.

They are a little expensive though.

HTH

Regards

Steve

According to Sanken's official site, there are only 2- 3 Authorized Distributors:
- one is from Japan
- another one is digikey (USA )
Unfortunately, both above distributors have huge P&P costs (to Europe).

There are also Non-Authorized Distributors and as I'm located in Europe, I am interested what do you know about www . littlediode . com ? Are they selling 100% genuine parts? ...
On a second look, they say nada about the hfe of those complementary outputs

 

[groenhen]

I'm based in the uk and have used littlediode for a few of the parts on my SA-9800 including the obsolete 2SK34 fets. Everything has measured and worked fine so while I can't say for certain that everything is 100% genuine, I have had no problems.

They are a little expensive though.

Regards
Steve

Hi Steve!
Well, happy to hear you had "positive" experience with littlediode.
Even if they're expensive, I seriously doubt they will exceed the bill of an US distributor!

Problem is I recently bought some outputs that turned out to be fakes.
Everything nowadays is literally infected by all kind of chinese crap :thumbsdn:
That's why I became excessively cautious when I buy on internet...

BTW: what major distributors for electronic parts do we have in Europe?
First I thought mouser has a warehouse in Czech republic, but it showed up I was wrong... unfortunately!

 

[rcs16]

mtf stated that ,
512-1N4454 is to be used for 1S2076, that is probably a better choice than a 1N4148

Yes matching hfe for any push-pull or diff pair, usually lowers distortion, canceling 2nd harmonics. If you can't get the exact hfe range, not too much you can do but go with what you can get. You could get a few sets more and do some hfe selection to narrow down the pairs to a closer match. There is some over lap in the ranges.

Those Sanken look to be the best choice, and at $10 ea, that ain't too bad.

Drivers?, mtf cover them, how about Toshiba
yup

q25 q26 2sa913 to-220 150v 1.5a 15w 65-330hfe 120mhz
757-2SA1930(Q,M) pnp to-220 bce 180v 2a 20w 200mhz 50-320hfe 200mhz nsa driver

q23 q24 2sc1913 to-220 150v 1.5a 15w 65-330hfe 120mhz
757-2SC5171(Q,M) npn to-220 bce 180v 2a 20w 200mhz 50-320hfe 200mhz nsa driver

In Europe, Newark/element 14 might be cheaper for shipping.

When you are substituting parts, you are in effect re-engineering the design, so you can possibly upset the apple cart. These days, one would simulate and test the actual design changes. Changing the transistors, will change the gain and phase characteristics, so it could due with some compensation tuning but that is beyond the scope of the thread. First get it running and stable.
You say one channel is dead and the other working, ONLY do the dead channel first!! A reference point. to compare aginst.

So, what is next? finalize BOM, post, spend $

 

[groenhen]

mtf stated that ,
512-1N4454 is to be used for 1S2076, that is probably a better choice than a 1N4148

Well, MTF stated that: 1s2076 replace: 512-1N4454 fast 100v .2a
Yet, 1n4454 seems to have only 75V (BV, according to datasheet).
That puzzles me...

In Europe, Newark/element 14 might be cheaper for shipping.

Thanks for the info!

You say one channel is dead and the other working, ONLY do the dead channel first!! A reference point. to compare aginst.

Thank you! Good idea!

 

[groenhen]

Important news!!!
Uncle Nick comes tomorrow to visit us...
And... he said he'll bring some Toshiba outputs to see if the amp works or not...
Hmmm, some people don't have patience at all !

Well, this is how headaches begin!

 

[rcs16]

Well, MTF stated that: 1s2076 replace: 512-1N4454 fast 100v .2a
Yet, 1n4454 seems to have only 75V (BV, according to datasheet).
That puzzles me...

Well I do not want to second guess mtf, I have never even seen one of these amps, so I only offer what I can. In this case other characteristics of the diode trump breakdown voltage, as long as it is within the necessary margins.

This is what makes every amp sound different. If your ear is good enough, I would think that you will hear the diff between the re-designed channel vs the original. It could sound better, it could sound worse, you might not be able to tell the diff.
Lets see what happens.

 

[groenhen]

Well I do not want to second guess mtf, I have never even seen one of these amps, so I only offer what I can.

SA7800 it's the NSA version of SA708 (I have them both -> 2x65W).
SA7800 it is the kiddo of the NSA family. There are also SA8800 (2x80W) and SA9800 (2x100W) - I never had these two amps.

But I am the proud owner of a SA-3000 which is a true dual mono - best sound ever, a real jewel. :tresbon:

 

[groenhen]

Important news!!!
Uncle Nick comes tomorrow to visit us...
And... he said he'll bring some Toshiba outputs to see if the amp works or not...
...

Well, he has replaced about 5 components:
- one resistor (560 ohm that was burnt). The replacement is a weird triple (?!?) resistors in series (I'll try to attach picture if possible)
- the NSA drivers (were 2SA913 and 2SC1913) with some NECs (2SC2275 and 2sa985) - not genuine NECs I'm afraid (my opinion).
- outputs (were 2SA1075 and 2SC2525) with TOSHIBAs. Bottom plate is back in place, so I can't tell exactly what Nick put there, but I guess 2sc5200 and its pair.

The amp seems to work.
I have a very good ear and I must say the sound doesn't impress me at all...
I expected much more from this NSA amplifier, so I'm rather disappointed...
Sound is worse than my sa708, but ok my sa708 is in top shape - recently recapped, lubed all switched & pots, heat-sinks were Pioneer has forgotten, etc. (and no damages since manufacturing date)

But hey, this was just a test... at least it works, not impressive, but...

So far, no details or explanations from uncle Nick. He was already gone when I came back with my kid from a walk in a park.

Next days I will remove what uncle Nick "planted" in my amp and switch back to my initial plan...

PS: Ok, I managed to attach that picture. See that weird 3-resistor series! :scratch2:

 

[rcs16]

Well, some one either wanted to increase the power dissipation &/or needed to make a certain value and only had a limited supply of values to chose from.

 

[larryderouin]

Am I reading that 3 resistor series right? Resistors add in series, 20K,20K, 36K???

If it's sitting right behind the 2sc2775 it's either R51 or R52 (depending on the side) which is 220ohms. The 560 ohm(R33-34) is the next one down.

 

[groenhen]

Am I reading that 3 resistor series right? Resistors add in series, 20K,20K, 36K???

If it's sitting right behind the 2sc2775 it's either R51 or R52 (depending on the side) which is 220ohms. The 560 ohm(R33-34) is the next one down.

As far as I can see, that 3 resistor series is replacing the original R33 (560 ohm) that was literally scorched...
Kinda weird, right? Nick also kept those ceramic insulators at the bottom!... From this angle, the picture is a little bit confusing for someone who doesn't have a clear global view...
Remark: I saw on FL tube that L & R channels are very well balanced !!!

I didn't change anything Nick has done Saturday...yet. Probably you have some idea / suggestions (based) on current situation/condition of the board ?!

Bad news, my aunt just died, so I do not want to bother uncle Nick for the time...

 

[groenhen]

Well, some one either wanted to increase the power dissipation &/or needed to make a certain value and only had a limited supply of values to chose from.

>>> The second one

Well, even if it sounds unbelievable... the sum of those 3 resistor series is 560 ohms! SAME as original value of resistor R33 !

>>> On the other hand, is there a way to tell if those NECs are genuine ?!?

 

[larryderouin]

Sorry to hear about your aunt. Please extend my sympathy to Uncle Nick.

It's way possible I'm starting to get color blind. The colors look weird tho.

Look at the solder joints for them. Chances are they are original. NEC did color their TO-220's Black for 2SC's and 2SD's, and Green for 2SA's and SB's.

 

[groenhen]

Sorry to hear about your aunt. Please extend my sympathy to Uncle Nick.

Nice of you to say that. Thank you, Larry.

It's way possible I'm starting to get color blind. The colors look weird tho.

Nope, your eyes are just fine. We all got tricked by that brown with orange tint.
The only way to solve the mystery was to unsolder one leg, lift it up and measure that thing with a multimeter.
Uncle Nick's new old stock...

Look at the solder joints for them. Chances are they are original. NEC did color their TO-220's Black for 2SC's and 2SD's, and Green for 2SA's and SB's.

Ooops, you lost me here: "Look at the solder joints for them."...
You mean chinese clones are supposed to be... different... how? Unfinished / imperfect (edges?) or something like that?

 

[groenhen]

ok, I finally ordered the outputs !
Tough decision... but, in the end, I opted for Sanken's.
The reason is quite simple...
I mean there are a lot of candidates for replacements, but I do care a lot about esthetics too... that's why I left OnSemi's aside and I chose the drop-in Sankens!
I made the order directly from Japan, just to be sure I will get genuine parts.
I wish to take the opportunity and thank all the AK fellows here who indicated Sankens as a replacement option.

PS: I realized that Uncle Nick also prefers Japanese parts/components!

 

[groenhen]

I knew it, I knew it, ...
I forgot to add the insulator sheets and heat-sinks on the list of materials!
Thermal paste is not a problem, I can find it at a local store...
What else do I need? I mean, from your experience... what else is it necessary when doing an amp restoration?

 

[groenhen]

All parts that I ordered came today (including outputs).

What should I do now?
I suppose it would be wise to start replacing components on the damaged channel first... (and leave the good channel afterwards).
Should I replace ALL (available) components on the damaged channel or only part of them?
What and how is it better to do in order to avoid blowing-up the amp ?!

Worth mention that I put FL display aside (no zombi wires inside the amp, so no worries about shorts or something).

Somebody with experience, please help me with some guidelines...!!!

 

[markthefixer]

1. document document document
2. 100 % component testing with one lead out of circuit - NO exceptions!!-
3. all diodes and transistors new.
4. methodical deliberate careful work.

 

[groenhen]

1. document document document
2. 100 % component testing with one lead out of circuit - NO exceptions!!-
3. all diodes and transistors new.
4. methodical deliberate careful work.

Tomorrow I'll begin replacing components ONLY on the damaged channel...
I will replace ALL (available) components on the damaged channel, except the power transistors (for the moment I keep the Toshiba's from uncle Nick).
After that, I will do the adjustment of this channel according to service manual and I will see if it works and how it will sound (compared with the right channel which is ok).

Question: Which of TO-126 transistors need a heat-sink? I remember you said q5/q6... Only q5/q6...?
(the area on GWH-118 board is quite crowded, so I don't wanna put heat-sinks if not really necessary).

BR,
Stanley.
PS: I surely hope this wend to be able to come back with good news & nice pix