Technics SA-828 Receiver from 70s- Is it worth trying to save this aging patient?

[dorke]

Yes,
Aux and Phono have different paths in the receiver.
Plus, the phono itself can produce problems (needle /dust etc.).

So let's stay with the AUX alone for now.

You should verify that the right channel performs well in all conditions:
Try changing/playing with output related knobs/selectors/filters/volume etc.
Operation should be continuous and smooth, if it isn't cleaning may be needed.

Protection shouldn't occur either,only if it does occur IC1002 should be measured ( DCV) ,
otherwise no need to measure.

 

[Iceman from McMurdo]

Hmmm, well knew it was too good to be true...

Right channel is definitely NOT smooth & continuous - even though it ran fine during that 1st test, further experimentation revealed it DOES kick into Protection Mode pretty consistently as it gets warmed up. At a relatively low 15-20% volume, it rarely goes more than 15-30 seconds before shutting down. Always resets with power switch though, and all output-related knobs/selectors/switches/volume seem to function normally.

What triggers Protection Mode anyway? Will the voltage checks on IC1002 help narrow that down??

 

[Iceman from McMurdo]

Update - Well, in the world of naval aviation where I served for many years, when pilots would bring their aircraft back & write up a maintenance discrepancy they'd observed inflight & wanted repaired before the next flight, it never failed that at least some of them would be annotated by maintenance personnel as, "CND" (or "Could Not Duplicate). Basically, there are some problems that are simply transient in nature.

That's what I think we have going on with the right channel... Just came back from testing again and listening to at least 15-20 continuous minutes of absolutely beautiful sound! Played with every knob & switch I could find, ran the volume up & down (from 0-35%), etc. Never tripped into Protection Mode & sounded AWESOME!

Whatever the problem is, it's intermittent and not worth tracking down right now, at least in my humble opinion.

I'm gonna try & track down those remaining few parts we need to find for the left channel repair & hope to drop a parts order tomorrow.

 

[dorke]

Well,
I agree with you on # 163 ,but it looks like that isn't the case here as per previous posts.
The problem is intermittent indeed.
IC 1002 is the device that monitors for "faults" and puts the receiver in protection if one is found.
There is a test point(on the board containing IC1002) called TP1002,It should have easy access.
In the case the protection does trigger again measure the DCV on it relative to GND.

The fault conditions (on each channel) are:
1. Output stage over-current(final Tr protection),not implemented in this case on IC1002.
2. Presence of DCV on the output stage(speaker protection).
3. Wrong values of power supply voltages.

Let's focus on the left channel than.
One important thing to be aware of :
We may get all parts(replacements) ,insert them gradually and fix the left channel.
But there is a possibility that the calamity could happen again,you should take that into account.

 

[Iceman from McMurdo]

Yes, certainly understand this repair job is a "risky" proposition that could easily go south, but I consider it a learning exercise and willing to risk failure if it should turn out to be a losing battle in the end.

Heck, I was just thrilled you could help me bring the lights back on & get the right channel going again!

Of course, it'd sound so much better in STEREO!

 

[Iceman from McMurdo]

Well, I've got some inquiries out to Mouser & Vetco trying to source alternates for the 4 part numbers I can't find at DigiKey, but haven't heard anything back yet. Hopefully tomorrow.

Been putting the right channel thru some extended play sessions today (30+ min & longer), and it played beautifully each time. No shutdowns.

You mentioned "cleaning may be needed" in #161... What's involved in that? Is that to perhaps remove oxidation from switches & contacts, because I'm beginning to wonder if part of the problem may have been related to the twist-type speaker outputs at the rear of the receiver?

When it was shutting down & not playing, I tried reconnecting speaker wires (at both receiver & speaker) to ensure good solid connections, and that's when I think it started to play more consistently. Maybe the contacts were slightly oxidized & just needed some exercising?

Any other parts sources you can suggest besides DigiKey, Mouser, & Vetco?

 

[dorke]

Cleaning=using Deoxit on contacts of switches, pots etc.
The quick twist-type speakers terminals are known to cause problems.
Can you post a photo?
Cleaning them may help.

Go over the solders with a magnifying glass,any suspicious point should be re-soldered and/or cleaned with a cotton swab and alcohol.
You can also try to find bad joints by taping with a non-conductive tool( like a wooden stick or similar ) at various points of the board or devices when the receiver plays.
Warning: that can be potently harmful!

You have contacted the main suppliers of devices.

I need to find the time for finding the replacements myself,
busy in other things at the moment,hopefully next week will have more time for that.

 

[Iceman from McMurdo]

Yes, I've heard great things about Deoxit - gonna have to invest in some of that on general principle. Have tried tapping around on various components while powered-up (using plastic probe) and got no reaction. Will go over PCB again with magnifying glass to look for suspicious solder joints.

Understand your time constraints - no worries at all. Looking back at the length of this thread, I'm forever indebted to you for your patience & expert advice thus far (some of which it would appear from the timestamps you've even provided in the very wee hours of the morning). You're a dedicated trooper! i salute you Sir.

FYI - twist-type speaker outputs shown below:

 

[Iceman from McMurdo]

Well, good news (I think)... Got some potential substitute parts info.

2OA99 = NTE109 (per Vetco, which was ONLY source I could find for this little diode)
http://www.vetco.net/catalog/product_info.php?products_id=1708&keywords=2OA99

As for transistors, got some mixed recommendations from Mouser, Vetco, and a different DigiKey rep than I spoke with before.

2SA1124 =
1) KSA1013YTACT-ND (per DigiKey, though it's only rated .9w, vice OEM's 1w rating)
2) ZTX795A (per Mouser, though it has a different pin-out we'll have to consider)
3) NTE32 (per Vetco, though physically very different with wide round flat head on top?)

2SB745 =
1) 2N4403-ND (per DigiKey, though it has a different pin-out to consider)
2) NTE234 (per Vetco, though physically very different with wide round flat head on top?)

2SC2632 =
1) KSC23830TACTND (per DigiKey, though it's only rated .9w with 100MHz trans freq, vice OEM's 1w & 160 freq)
2) NTE399 (per Vetco, though physically very different with wide round flat head on top?)

Kinda inclined to stick with DigiKey parts if specs are deemed acceptable, as the NTE parts from Vetco are totally different than OEM parts (though I believe there's room to fit them on the PCB, if needed).

Any advice or recommendations? I can send links to above parts if you need to review specs more closely. THANKS!

 

[dorke]

Nothing suspicious in the photo of the speakers terminals.

Yes,please provide links to the devices offered .
Please post a photo of the original Trs,so all part number detailed may be viewed clearly.
I will go over them next week.

 

[Iceman from McMurdo]

Got some Deoxit on the way and will spray down any suspicious terminals, pots, switches, connectors, etc. Safe to use on/around the PCB if needed, correct?

OK, here's the problematic parts, with photos & links to potential sources, incl spec data :

2SA1124 - needed for BOTH Q613 AND Q629, but you'll see they have slightly diff case markings (photos 1 & 2 below). Not sure if that makes a difference or not, since they have same part #s.

DigiKey can supply with .9w power dissipation rating, vice OEM's 1w rating:
http://www.digikey.com/product-detail/en/fairchild-semiconductor/KSA1013YTA/KSA1013YTACT-ND/4553035

Mouser appears to have one with full 1w power rating, but with a diff pin-out config that now appears problematic:
http://www.mouser.com/ProductDetail...GAEpiMZZMshyDBzk1/WizF0eVyYE44Z5Zqv%2bUsbU6M=

Vetco says NTE32 will work, but it has a completely different physical shape (plus power rating is only .9 too):
http://www.vetco.net/catalog/product_info.php?products_id=3421

I'm thinking the DigiKey option comes closest to OEM part, assuming we can live with the slightly lower .9 power dissipation rating for both locations on the board.


2SB745 - needed for Q625 (photo 3 below).

DigiKey recommends this, but it has diff pin-out config (Base in ctr vice edge) that we'd have to be creative with:
http://www.digikey.com/product-detail/en/fairchild-semiconductor/2N4403BU/2N4403-ND/1418

Vetco says NTE234 will work, and (surprisingly) when I revisited their site I see it has a traditional head shape (vice the round flat head of the other NTE products), AND a pin-out config that matches the original
http://www.vetco.net/catalog/product_info.php?products_id=2862

I'm gonna go with the NTE234 part, since I have to order the obsolete 2OA99 diodes from Vetco anyway.


2SC2632 - needed for Q627 (photo 4 below)

DigiKey recommends this, but it's only rated .9w with a 100MHz transition frew, vice OEM's 1w & 160 MHz freq
http://www.digikey.com/product-detail/en/fairchild-semiconductor/KSC2383OTA/KSC2383OTACT-ND/4213848

Vetco says NTE399 will work, but (again) very different in appearance
http://www.vetco.net/catalog/product_info.php?products_id=3529&keywords=nte399

I'm inclined to go with the DigiKey part, assuming we can live with the slightly lower .9 power rating & transition freq.

Each of the links above has internal links to the spec datasheets if you need to double-check anything.

Although I'd obviously like to get parts that are "exact" matches for the obsolete OEM parts, beggars can't be choosers at this point in the supply chain. As I've reviewed the datasheets closer during preparation of this update, think I've already got a pretty good idea what will work.

Biggest question I have for you (or other lurkers on this thread) is whether the slightly lower power rating (.9w versus 1w) and/or transition freq (100Mhz versus 160MHz) of the above parts will cause any MAJOR problems in our rebuild. If we can live with the diff specs, I've got the full parts list ready to order now.

 

[dorke]

The issue of replacements isn't that simple:

1. There are Trs which must be complimentary pairs
i.e if you don't have one of a pair you have to replace both with another suitable complimentary pair.
The pairs are:
Q635 and Q637.
Q631 and Q633.
Q627 and Q629.
Q623 and Q625.

2. The "suffix letters" like R,S,T are important ,you have one "2SA11024R" and another "2SA11024S " .
they are actually different devices(in their amplification factor hFE).
Since we don't have the parts list of the receiver, the only real references are the blown devices .
So this applies to all of them, not just the ones you couldn't source.

3. There may be other issue,need to go over the circuit to find out,and don't have the time now.

I advise that you wait till next week.

 

[Iceman from McMurdo]

Hmmm, I knew Q635 & Q637 were matching, and DigiKey has sourced those (Sanken 2SA1215 & 2SC2921).

And, you told me previously that Q631 & Q633 were matching, which DigiKey has also sourced together (Sanken 2SA1667-ND & 2SC4381-ND).

However, I wasn't aware the other two pairs you mention had to be matched, so agree we may need to revisit parts selection there, although the parts houses seemed pretty limited in options for these obsolete parts.

WRT Q623 & Q625, it appears NTE234 is the ONLY option for Q625, so I may need to try and find a matching NTE part for its partner transistor Q623...

Dunno about the two KSA part subs for Q627 & Q629, especially because of the diff transition freq on the sub for Q627. Definitely may need to do so more searching for hat pair.

Agree, will postpone parts order until you have some time to help ensure we get complementary pairs No worries on timeline- I can certainly wait, so please do so at your convenience.

 

[Iceman from McMurdo]

Still patiently waiting on parts ID concurrence...

No rush - but each time I pass by the work bench I keep wondering if I'm ever gonna remember how to put all that stuff back together again!

 

[dorke]

Hi Iceman,
I'm very busy lately.
My apologizes on the delay,it will take some more time,hope you have the patience.

BTW,
I have just got a Technics receiver model SA-1010 ,it is a 120W monster(the bigger brother of the SA828).
Guess what?
It has one working channel(great sound) and one bad channel with blown bat-wings Trs. and probably more...

 

[Iceman from McMurdo]

Ha, what are the odds! Well, that SA-1010 should be a piece of cake to fix that one now that you've already walked me thru all these issues with its little brother!

No sweat on the delay, just wanted to make sure you hadn't lost patience & abandoned me. Happy to wait awhile longer. Thanks for letting me know you're still there.

 

[dorke]

Iceman,I'm back.
I shall have more time for a while.

About the SA-1010.
It is a more complicated unit,it has a uC in the loop for setting the final stage into proper working conditions (a "pre-heat" state).
But in general the work on the SA-828 may help.

Just to be come back on the same level :
Could you please put together a list of the Tr's and Diodes identified as "bad" and the suggested replacements for them ?
Did you inspect (visually for over heating marks,and in ohm test)resistors: R685,R683,R681,R679,R677,R673,R671,R649,R647,R643,R633 ?

 

[Iceman from McMurdo]

Thanks, will try and get you caught back up on the status of the SA-828. As I'd mentioned before, many of the OEM parts are no longer avail, so the tech reps at DigiKey & Vetco did their best to find suitable alternates. I'm going to try and upload my planned "shopping list" of parts based on either what I could find listed on their parts pages or (if unavail) what they recommended instead (see attachment below).

I think the biggest wrinkle in the list will be confirming the suitability for those TRs you mentioned had to be "matching pairs." The only pair I knew to ask about when searching was Q631/Q633 and, of course, the 2 batwings Q635/Q637 - and I think we're covered there...

But, in #172, you also said Q627/Q629 and Q623/Q625 had to be matching, so I'd ask your help in confirming whether the ones I have listed will work, or maybe helping me identify some preferred matching pairs that are still avail today.

As for the resistors you mention, I have not yet checked them, but will endeavor to do so later today.

PS - Regarding diodes... NONE of the sources I checked with could supply the original 2OA99, and only one (Vetco) had something listed as a potential substitute (NTE109), so I went ahead and grabbed a few of those while they were cheap & avail. Hope they will work ok...

 

[dorke]

We need to look at the suffix letter of the Trs as well.
If visible/readable on the bad device that would be fine,
if not then look at the corresponding Tr in the bad channel.

e.g the batwings D845 R ,B755 R the suffix letter here is (R).
That letter in Japanese Trs determines the Hfe range of the Tr,
we should match the replacement Tr as close as possible to it.
Please go over the bad parts and add that letter to the part number like so
SD845 ==> 2SD845-R etc.

 

[Iceman from McMurdo]

OK, examined all the bad TRs and added suffix letter to all of them I could read. Q625 was barely visible, but I think the suffix is T (there's a closeup photo of it back up in #171 that might help if you don't think a T suffix would be correct).

Updated parts list (see attachment 1 below) has summation of TR suffix info. Didn't add suffixes to Q631/Q633, because they appeared to have some kind of numerical suffix - maybe you can decipher by focusing in on them from my souvenir "group photo" of dead TRs (see attachment 2 below).

Getting ready to try and check those resistors you mentioned earlier. Will advise results soon. Thanks!



.