SA-7800 Restoration - Need help

[jcnash]

Hi all,

I picked up an SA-7800 Non-Switching amp along with some other gear a few months back, and finally was able to get to opening up the amp today to try and get it working.

I've worked on its big brother the SA-9800 about 3 years ago and with some extremely helpful advice from markthefixer, was able to get it fixed and working. It has worked flawlessly since and I enjoy it regularly. I'm hoping to get the 7800 fixed up as well.

So essentially, the symptoms on the SA-7800 are as follows:
1. When I flick the power switch on, the power switch LED and the Fluoroscan meters *DO NOT* light up
2. After a few seconds I hear a click from the protection circuit
3. I've tried playing a source through the Aux input, and I get some music on both channels, although the music is very tinny, and extremely distorted. It was so distorted with loud pops and crackling, that I had to shut it off immediately for fear of wrecking my speakers!

I've removed the case and plan to clean and polish it up while I get the unit working. I'm going to de-oxit the controls, to hopefully get rid of the crackling. I've also downloaded the Hi-Res Service Manual and have started to look at it. I'm handy with a DMM, and am hoping that I can use my DMM reading/soldering skills to get this unit in working condition again.

Any help from the experts on AK is much appreciated.

Jacob.

 

[markthefixer]

power supply GWR112 rebuild most likely.

(similar thread: http://www.audiokarma.org/forums/showthread.php?t=200669

V(gr) is ground referenced voltage reading, the black dmm lead goes to a bare metal chassis ground, and probe with the red dmm lead, insulate all but the very tip of the probe in case of slips.

(2) (3): check the power supply first , GWR112 board
pins 15, 16, 17: V(gr) +48vdc regulated
pins 19, 20, 21: V(gr) -48vdc regulated
pins 29, 30: V(gr) +46vdc unregulated, this can vary somewhat
pins 31, 32: V(gr) -46vdc unregulated, this can vary somewhat
pin 22 V(gr) -?? vdc thisis the turn-on/turn-off signal for the protection circuit

(1): The protection circuit is ON the power amplifier circuit board GWH118
and we need to see if it is justified in keeping the speakers disconnected.

V(gr) R71 either side
V(gr) R72 either side

These two resistors are larger than the rest, 2 watts, on either side of the protection relay.

That's enough checking for now.

 

[jcnash]

Mark! I did not expect you to reply to this thread!

Ive been away from the forums for a while, and the last time I was logged on, I saw the thread about your unfortunate hospitalization. I did not realize you have returned to the forum.

It is definitely a very pleasant surprise!

I trust you have recovered fully, and if not, I wish you a successful, speedy and complete recovery.

That being said, personally I would hate for people to continue to refer to my past illnesses, and so I shall not refer to yours any more. :yes:

Welcome back!

I will get the required readings and post back.

 

[markthefixer]

Thanks - notice my status line under my username, it is uncomfortably close to the truth as to what happened this last July.

don't bust your @## getting the readings, take your time - I'm swamped !!!!

I did a SA-7800 a while back, so there is a list (of sorts) and I had not established my "format" for parts, also I (peeked) have to deal with some obsoleted replacements in it...

 

[jcnash]

power supply GWR112 rebuild most likely.

(similar thread: http://www.audiokarma.org/forums/showthread.php?t=200669

Hi Mark,

The following are my readings:

GWR 112:
15, 16, 17: 48.6v
19, 20, 21: -47.4v
29, 30: 32.8v
31, 32: -49.8v
22: -1.9v

GWH 118:
R71, R72: 0v

Considering it has been a while (3 years!) since my previous project, Ive taken a picture to confirm that the resistors whose voltage I measured are in fact R71 and R72. I looked at the schematic and the R71/72 values were 10 ohms, 2W, and I matched that with the values on the actual resistor in the vicinity of the protection relay box. I've attached a picture and circled the resistors whose voltages I measured.

Also, while digging around, I noticed that there is some kind of brownish goo coming out of the bottom of one of the large capacitors on the power supply board. There are 4 capacitors in total, and the one that is leaking is the one on the extreme right while facing the front of the unit.

 

[markthefixer]

brownish goo if hard is probably glue - electrolyte looks like puke.

pins 29, 30 are radically low. That cap may BE leakage! or maybe the glue doesn't show under the other 3.

the GWR112 D34, D35, D36 D37 SR3AM-8 (400v 2A, 40A surge) diodes are somewhat suspect (512-1N5404)
as well as C39, C41 (8000uf 50v). The diodes could have failed somewhat hurting the caps on that polarity.
Gotta pull one lead loose each to test each one individually.

that "leaking" cap, is it closer to pins 19, 20, 21, 22 ? that would be C39. The other directions don't reflect the positioning in the chassis, unless pictures were available.

measure the diameter and height of the cap (mm) as well as the lead hole spacing (7.5, 10, ? mm) we will probably end up getting 10mm snap in caps, and have to enlarge the holes on the board to 1/16 (0.060") inch on the same (hopefully) spacing.

now take it easier on me...

 

[jcnash]

Haha...sounds good. I was going to say take your time and reply even if I have a post on, but I know how that is. I would not be able to ignore it...and get to it at a later time, so Ill space my posts out some more hereon. Really do appreciate your help!

OK so the goo is brownish and hard. I poked it with a long thin screwdriver and chunks of it broke off. The goo blob is therefore brittle....so that's some good news. Maybe just glue?

The "leaking" cap is indeed the one closest to pins 19, 20, 21, 22. I have attached pictures for reference/confirmation. One of the pictures also shows the underside of the power supply board with a drop of the hard brown goo that has leaked through the tiny empty hole in the PCB in the middle of the three capacitor terminals (lead hole??).

Capacitor information (as read off the actual cap):
50V
8000uF
Height: 63mm
Dia: 30mm

Im assuming the lead hole is the empty hole (now filled with hardened goo) in the middle of the three capacitor terminals.

The lead hole to terminal distance is 7.5mm for each of the three terminals.

If you were looking for the distance between each of the terminals, they are:
Terminal 1-2 - 11mm approx
Terminal 2-3 - 11mm approx
Terminal 1-3 - 15mm approx

 

[markthefixer]

top view picture shows the BIG diodes (4 of them) next to the caps and right where the transformer wires are attached.

you need to read the DC and AC voltages at the capacitor terminals.

but that will just tell us there is a problem, not if it is the diodes or the caps.

you HAVE to either test the 4 diodes with 1 lead each pulled or just plain replace them

the caps almost certainly need replacement and it appears there is room to enlarge the hole to the back, and enough foil to relocate (drill a new hole 2.5mm closer to the heat sink) the other hole to fit snap in caps (10mm lead spacing, 0.62 1/16 inch holes).

I strongly expect a diode failure because if only one of the two caps failed in the pair we still would have 8000uf and the loss of the capacitance would be less noticeable at low volume levels. A failed diode is continuously fighting the charging of the cap, and probably damaging the pair, thus the replacement.

like I said, 4 diodes 512-1N5404
4 caps - 2 choices
647-LGY1H822MELB40 $4.91 ea for 4
647-LKG1H822MESCBK $8.54 ea for 4 "audio grade" and are 35mm so the holes and caps would have to be angled a bit from parallel lines through the leads to clear each other.

 

[jcnash]

Sounds good Mark.

Ill pull 1 terminal of each of the diodes and test them. I would just replace them, but I want to avoid replacing the caps if I can because Im a little concerned about drilling holes in the PCB etc. I've obviously never done anything like that before, and dont want to bite off more than I can chew, and needlessly wreck the entire amp.

I'm hoping that when I test the diodes, Ill find that atleast 1 of them has failed...so I can just replace the set of 4 diodes and not have to replace the caps. If however I find that all 4 diodes are functioning OK, I will then look at replacing the caps and the diodes.

Thoughts??

 

[markthefixer]

diodes ok = bad caps = need cap replacement, but diodes could stay

diode bad = replace diodes, reevaluate by reading capacitor voltages after diodes replaced. caps could have survived.

 

[jcnash]

Sounds good. Will post back with the results of the diode test.

 

[jcnash]

Just pulled out my soldering iron from back in the day to try and remove the diode lead to test, and it wont heat up. She's dead!

Will be picking up a new one, but that will delay my reply by a bit

 

[jcnash]

Hi Mark,

So I was able to pick up a new Weller 60W soldering iron over the weekend and managed to lift one terminal of each diode off the board and test them.

All 4 diode forward voltages read ~0.6v. The exact voltages are as follows:

D35 - 0.589
D34 - 0.603
D37 - 0.597
D38 - 0.592

I am not sure if these are up to spec though. From the SM, I was able to figure out that they are SR3AM-8 diodes, but couldn't find the specifications for that type of diode.

What do you think of these voltages? Assuming they are functioning properly, I will have to replace the caps. If that is the case, would it make sense to replace the diodes as well (in case they are nearing end of life)...considering they are fairly inexpensive and easy to replace?

Thanks.

 

[jcnash]

Mark - Update on the 7800:

Ive placed an order for the 4 caps and the 4 diodes.

I de-soldered the old caps from the board and noticed that they are the 3-terminal caps. But the 3rd terminal isnt connected to any foil or wire on the board, so Im assuming that it is redundant/ground.

The caps I've ordered (the parts you provided me with) are two terminal caps.

Will solder in when I receive it and post back.

Thank you very much for your help so far in identifying the potential culprits.

 

[jcnash]

Hi Mark,

Had a quick question about the replacement cap. Ive put my order on hold, and will confirm once I hear back form you.

Considering the lead spacing on the PCB is 15mm and the caps nowadays are 10mm snap in's with not enough lead to allow for stretching, I did some research on Mouser and found this:

http://ca.mouser.com/ProductDetail/Nichicon/UKW1H103MRD/?qs=kArNe9LFxXlXU/0N8Vy%2byg==

This is a 10000uF, 50V cap, but it is the "terminal type" which has enough terminal length to allow for bending and stretching to fit the 15mm lead spacing.

So essentially my question is: Can I replace the stock 8000uF, 50V cap with a 10,000uF, 50V cap? I know that replacign with higher voltage caps are perfectly fine, but I wasnt sure about capacitance.

The 8000uF caps usually have +/-20% tolerances which brings their max capacitance close to 10,000uF anyway, so Im hoping this will be OK and I can order these caps instead, to avoid having to drill new holes in the PCB and other such madness.

Do you think this would work??

 

[markthefixer]

more basic issue, IS the distance between the two terminals of ONE capacitor 15mm? Comparing the distance with the Phillips screw heads in the picture makes me dubious.

and can the caps go side by side or do they interfere with each other?

all for the sake of avoiding drilling.

I'm giving 3:1 odds that you'll have to drill holes for just about any cap.

as to the original question, 10,000 uf is fine, the diodes are beefed up a bit with the replacements anyway.

p.s I have that same flashlight!

 

[jcnash]

Re-took the terminal distance measurements after de-soldering the old caps to be doubly sure.
The old cap specs are:
30mm Dia
65mm Length
Distance between two active terminals - 15mm
There is a third terminal on the old caps which I'm assuming is a dummy because it isnt connected to anything on the PCB.

The new caps that Im considering:
25mm dia
50mm length
Distance between two terminals - 12.5mm.

And the caps "stand up" next to each other. So without seeing the new caps, Im guessing that I should be able to do the same, and there should be more distance between each because they are smaller in size. Ill have to rig up a system to make sure that the weight of the caps are not supported by the terminals which will have to be bent a little bit to fit the 15mm PCB holes.

Sounds good...Ill go ahead and order them along with the new diodes. Now Im really looking forward to getting these parts in the 7800 .

Haha yeah.. I've had that flash light around for years. Its built pretty solid...

 

[markthefixer]

sounds good.

my flashlight has taken a beating (dropped many many times), and has a screw through the center of the back to hold the back cap on, as well as extra wires to fast charge the ni-cads or run with external power for long periods of time.

 

[jcnash]

Hi Mark,

So I finally received the parts and plugged them into the amp.

Keep in mind, I replaced the 4 original capacitors (8000uF, 50V) with new 10,000uF, 50V ones.

I also replaced the 4 diodes with the 1N5404 diodes.

Plugged her in and...still the same result. Upon flipping on the power switch, no power indicator lamp and no fluoroscan meters. However, just as before, I hear the protection circuit click after a few seconds.

I re-took the original readings you had asked for and here they are after the new parts went in:

GWR 112:
Pins 15, 6, 17: +48.6V
Pins 19, 20, 21: -47.4V
Pins 29, 30: 44.7V
Pins31, 32: -44.7V
Pin 22: -1.6V

GWH 118:
R71, R72: 0V

Any thoughts?

 

[jcnash]

Also just made a very weird observation. While messing about, I accidentally touched the heatsinks that are attached to transistors Q34 and Q35 and felt a slight current. That was pretty weird...so I put my voltmeter leads to it and sure enough, one heatsink indicated +62.5V and the other indicated -62.5V.

I looked up the SM and one of the leads of Q34 and Q35 are supposed to hold +/-63V.

However is that as per design to allow voltage to flow into the heatsink surrounding a transistor? Maybe its not a heat sink and a "voltage sink" instead?