5000a keeps blowing fuses, help

[Liljohnjo]

Hi Y’all, I just picked up a sansui 5000a on Craigslist yesterday, got it all plugged in today, sounds amazing….at least it did. After about 20 minutes the protection lamp came on and my speakers went quiet, which is fine. So I turned it off changed to my other outlet, everything seemed to be normal except now my left speaker isn’t working (headphone port too). So I checked the fuses, the left channel quick acting fuse popped. Again if there’s a surge, to be expected no biggie. So I just got back from Lowe’s with some replacement fuses, and as soon as I turned it on, it popped again, so I tried another one, popped again. So I’m at a loss, all the while the right channel is fine and still works completely normally. I’m new to working on receivers but have worked on vintage guitar amps so soldering caps and transistors on old equipment isn’t new or scary to me necessarily. Any tips or answers or tests would be highly appreciated, thanks

 

[sansui 74]

More that likely the left output transistor is blown and needs to be replaced, also check the tightness on mounting screws become loose over time and will not disparate heat properly I had same problem on a 5000 I had years ago and will need to be recapped as well.

 

[Guido57]

Okay, do you have any idea of what amplifier boards it has? there are two possibilities F-1040 or F-6013. Using a flash light look through the vents at the back top. There will be a pair of circuit boards situated between two pairs of large capacitors. On the right top rear of the boards the number will be at least somewhat visible. (alternatively you could take the cover off) Check that out and report back. If you open it up you might as well include pictures of the boards. Note that as far as I'm aware most 5000a's that are still working have had the original F-1040 boards replaced with F-6013 boards.

 

[dly66]

But a lot of them don't. Been running an F-1040 5000A in the living room for at least 6 months. Has the @skippy124 mod and some other updates on the amp driver boards. I believe it was the square 'chiclet' bias diodes mounted on top of the pre-driver transistors that tended to go open and blow the amp circuit up. Sansui did a replacement driver board for the 5000/5000A/5000X series but I get the impression a sizable number of them probably never got the F-6013 replacements. Been impressed by the sound of the 5000A. Been looking to swap a restored Sony STR-7065 into that system but keep putting it off.

If it still has F-1040 boards in it with the black square bias diodes sitting on top of a pair of metal can transistors on each amp driver card you probably want to do some updates there. The F-1040 boards look like this and the green arrows point to the diodes in question. There should be another transistor with square diode on top farther down on each board. If you don't see those you probably have either the F-6013 or someone has done some work on it before.

 

[Guido57]

If it still has F-1040 boards in it with the black square bias diodes sitting on top of a pair of metal can transistors on each amp driver card you probably want to do some updates there. The F-1040 boards look like this and the green arrows point to the diodes in question. There should be another transistor with square diode on top farther down on each board. If you don't see those you probably have either the F-6013 or someone has done some work on it before.

That's why I asked them to post pictures if they opened them up. I see from your picture the F-1040 boards had their identification screen printed in a different location than the F-6013. If one wants to get picky there is also an f-1040-1 version of the board which shares all the problems of the F-1040. My first 5000a had the F-1040-1 boards, the diodes and their mountings were in much worse shape than what is in your picture. One of the diodes had completely separated from the transistor.

 

[Liljohnjo]

Thank you sansui 74 for the advice, and I left it at a buddies so he could look at it, and he just texted me saying mine’s got the F-1040’s. I’ll get back over there on Thursday and I can post pics thank you guys so much for all the help so far.

 

[EastPoint]

You can update the F-1040s, you don't have to replace them with the F-6013s.

 

[Liljohnjo]

So they’re actually the F-1040-1’s and here’s some pictures. Also, there’s a sticker on the side saying it was serviced in 79’ so maybe they were already updated?!?!?!?

 

[Guido57]

Sansui was replacing the boards with F-6013 until 1982, so I while someone has been in there it may not have been a Sansui tech. Since the metal clips holding the diodes to the transistors are missing I'm guessing that there was an attempt to rectify the situation, but considering the corrosion on those transistors it may not have been enough for a robust permanent fix. Not knowing how many transistors are blown, it could be every one on that board, I see two options: the Skippy mod or reproduction F-6013 boards. There is at least one member on AK who does not like the Skippy mod and may have reproduction F-6013 boards available. There are others quite happy with the Skippy mod. If you go with the Skippy mod I have the power and output transistors from one channel of my 5000a to help with the rebuild if needed, but the transistors with the corrosion on both boards have to go.

 

[dly66]

It has probably been worked on. Looks like someone replaced those problem diodes. Job looks nice, may have been a good tech who knew what they were doing.

I am pretty sure Fred Longworth did warranty work on these way back and I think is a member here. I have seen old posts of his on this and other websites talking about issues on the 3000/3000A and 5000/5000A/5000X series back in the day. I found this comment on this site from back in 2006:

On the 5000A, the actual failure is the two dual forward bias diodes (black things) mounted on the top of the TO-5 driver transistors. These were poorly manufactured. They open and cause both levels of class-AB drivers and the output transistors to go into saturation. This will usually kill the outputs. In my opinion, this is the one and only flaw in the first-generation board design, and it CAN be corrected.

I used to use the MSD7000 (Motorola) as a straight-across replacement for the "black gizmo." Unfortunately, this has been out of production for nearly twenty years. These days, the NTE605A is a good substitute for the "black gizmos." I will epoxy the 605A's to the tops of the driver transistors and use fine wire to hook the leads to the pc board.

If the amplifier has failed, however, you will need to replace the outputs, and check the two pairs of driver transistors (TO-66 and TO-5) out of circuit. Look for leakage.

Replacement outputs will likely have wider bandwidth, so you may need to install small compensation capacitors in the class-A driver stages to prevent oscillations.

There is also a high-failure (leaky) transistor used on the 5000A and similar models: the 2SC458. The 2SC2240 is an excellent replacement for this.

Best of luck.

Fred Longworth
San Diego, CA

 

[dly66]

Mine was acquired as a distorting fix and repair. Has MJE15032 as drivers and MJ21194 outputs. Had the F-1040 boards as in the picture I posted, no TO-66 type drivers. Put small heatsinks with isolator kits on the driver transistors, not much excess room there. Was able to get the amp voltage and bias current to adjust properly. Initially adjusting on the dim bulb found out my bias trimmers were set backwards at max instead of min did not blow anything up. No oscillation issues are apparent and it has not getting very warm in my setups.

 

[Guido57]

Perhaps this tread may be interesting reading:
https://audiokarma.org/forums/index.php?threads/tech-question-for-echowars-etc.74559/
Particularly post #8.
If I had to do it over again the first thing I would do is remove the filtering and coupling caps. They are guaranteed to have dried out electrolyte, even if they measure in spec, and should be replaced anyway. This will give plenty of room to carefully remove the F-1040-1 boards so they can be upgraded: new caps, diodes and modded to prevent future failures, or replaced with F-6013 boards. I cut the braided grounding straps on the F-1040-1 boards as they were to much of a PITA to de-solder. Replaced them with 14 gauge wire.

 

[sethl]

Excited to dig into the Skippy mod, as I just picked up a 5000a with similar issues. Question: Do the 2SD118 output transistors typically fail as well? Easy enough to check for shorts, just curious as I see little written about this possibility....

 

[Hyperion]

Do the 2SD118 output transistors typically fail as well?

Only when they are taken out of their SOA - which all transistors, and especially Power Transistors have as an important part of their specification.

Turning the unit up too much, especially with too many speakers or otherwise difficult speaker loads stresses power transistors. Also blown driver transistors, out of control, or incorrectly set bias, or dare I say it 'wrongly orientated components', and 'probe slips' may/will also cause problems for output power transistors.

 

[bubhardy]

Only when they are taken out of their SOA - which all transistors, and especially Power Transistors have as an important part of their specification.

Turning the unit up too much, especially with too many speakers or otherwise difficult speaker loads stresses power transistors. Also blown driver transistors, out of control, or incorrectly set bias, or dare I say it 'wrongly orientated components', and 'probe slips' may/will also cause problems for output power transistors.

Or, when one of the speaker outputs is shorted.