Sunn(0))) Amp Beta Bass Repair

[wirelessliquid]

every time I try to fix this it blows up on me, I need some help

So when repairing this I accidently put the power chords in backwards from the transformer and ruined my already broken amp, After I had done this I checked every components on the board and had to replace CR4, CR10, Q8, Q9, Q10, Q17, Q18, Q19, C16, C17, F2, F3 and the IC 4558.

I tried the amp last night with everything checked out and R17 blew up and caught on fire but I just replaced Q8 so I'm not sure why it would have blew. Does anyone have any ideas why R17 is blowing and what I should do?

I've had a few goes at fixing this amp so far and no luck, its a 1979 Sunn Beta Bass, Drawings and service diagram can be found below any help would be greatly appreciated

https://imgur.com/gallery/KozwgNg

EDIT: Attachments

[nevetslab]

If I understand you correctly, putting the 'power chords in backwards' do you mean you mis-connected the 8-pin power supply connector shown in the photo? If you lost the Drivers & Outputs, I'm guessing you somehow reversed the power supply connections. I don't see the pin-out for that connector on the schematic. That connector on the PCB is polarized, but nothing there preventing mis-connecting it in either direction by one pin.

I'd check to see if the rest of the semiconductors are still good. I assume you have the basic testing skills and can make those determinations with a multimeter. If it has a semiconductor Test mode, all the better. If not, running it in Ohms mode, 2k range generally gives you a 2V potential with a 1mA constant current source, and can turn on semiconductors, showing you the diode drop (Vbe, Vbc). I see you already have flame damage on the bias circuit in the middle of the board by the heat sink. Make sure the bias xstr Q13 is still good. If it is now open circuit, Q5 and Q14 voltage gain stage is probably turned on hard, pulling pre-drivers Q7 & Q16 on hard, which would turn drivers Q8 & Q17 on hard...enough to cause your R17 to burn. If you jumper across the C-E junction of the bias xstr, it would basically turn off the output stage. The bias xstr is normally adjusted to control just how much voltage you want at the Vbe junctions of Q8 & Q17. That voltage is normally less than the normal junction voltage of about 0.6V or so. 0.35V to 0.5V is probably in the range Sunn sets it up at. That voltage is set by the trim pot R42.

So, first, verify all the rest of your circuit semi's are good, and if any resistors look like they've burned open, replace them. The cap C13 that's in the burn region is a signal bypass path around the bias xstr Q13.

It's a very simple & basic complimentary solid state power amp.

When you replaced the TO-3 power xstrs, did you re-use the insulators? They look like an embedded silicon rubber-type insulator, and, over time, thermoset takes place. Taking them off the heat sink can damage the insulation system, now making them a potential short between the collectors (xstr case) and the heat sink (ground). Greased mica works well...a little messy, but gives better thermal conductivity and very good insulation.

[wirelessliquid]

Thanks nevetslab, I will report back in a couple of days when I have a chance to test this on the amp,

So the power connectors I put in incorrectly were the three cables coming from the transformer 0v, +40v, -40V which can be located on the power board in the top right by the two larger capacitors. I'm not sure if revering these has damaged the preamp at all either. (the preamp is attached in the service manual).

Before retesting the amplifier I checked all semiconductors were in good condition and replaced all 6 large transistors which do the brunt of the amplifying. Will these need replacing again as R17 has blown, I've found usually when ever something blows it usually takes out these large transistors but fuses F2 and F3 didn't blow this time so I have a feeling they should be okay. I only have the semiconductor test mode on a Fluke 197 which is generally pretty good for transistor and diode testing but I've found its not so good on high current transistors like these large ones as it cant spot leakage. Is it worth investing in a dedicated transistor testing unit?

I'll have to report back later on this as it's useful for having a look around thanks

Yeah when I replaced the TO-3 transistors I used new silicon pads that were brought from eBay and I checked to see if there was any shorts between the collectors and the large heat sink but didn't find any. I have some transparent insulation pads that look like there made from a more rigid plastic would these be any better, I'm never toe sure if the blue ones are great as there made from a more rubbery material. I've never heard of greased mica, is it a bit like the thermal paste you'd use on the processor to heat sink on a PC?

[nevetslab]

The Transformer wires are NOT +/-, they are AC CenterTapped leads that plug into the bridge rectifier. The C/T is normally a different color, such as a pair of RED leads and a Red/Yel wire (the centertap). If, instead, you managed to connect the OUTPUT from the power supply, which there you WILL HAVE +/- 40VDC, reverse connecting those to the circuit will do major damage. THAT sounds like what you did....though all I could tell from the photo was seeing an 8-pin polarized connector, which COULD be mis-connected by missing a pin on either side instead of properly aligned. Plugging it in upside down, the PCB male header with it's lock would fight you.

Actually, the amplifying of the signal is done ahead of the output stage. The output stage, from Q7/Q16 on is unity gain, and instead provides the output current. I'm thinking Q7 & Q16, the pre-drivers, failed. With Q7 shorted, it would pull enough current thru R17 to do damage, and would also turn on Q8, Q9 & Q10, potentially causing them to fail again.

Do you have a Variac (variable ACV Autoformer, providing 0-120VAC) to work from? This is an essential tool when servicing an amp like this, as you can bring up the AC mains from 0V and watch to see if the output voltage immediately increases in either positive or negative direction. With a Variac, having an AC Ammeter to show you mains current is, to me, and essential tool as well, since you'll see immediately if the unit under test is pulling fault current or not.

For general purpose work, your Fluke 197 is adequate for your needs. Only a few of us have semiconductor curve tracers. Usually, we will find the leakage problems in-circuit, either under idle conditions when things aren't making sense and the circuits aren't sitting where they should voltage-wise.

In looking closer at your schematic, after downloading it and zooming in, I see they have provided enough voltage levels thruout the circuit to let you figure the rest out. Such as, seeing 1.2VDC at the base of Q8, you'd know to expect about 1.8VDC at the base of Q7/collector of Q5, and the mirror of that on the base of Q16 (-1.2VDC), with similar negative voltages at the base of Q16/Collector of Q14, about -1.8VDC. And what to expect at the bases of Q9/Q10 & Q18/Q19 (+/- 0.6VDC or in that area). The output should be near 0VDC, might be +/- 50mV.

[wirelessliquid]

Yeah sorry haha the two wires coming from the full wave rectifier before C16 and C17 and the ground wire were the ones I mixed up when blowing my amp to bits, The 8 Pin connector physically can't be put in the wrong way and that connect to the pre amp board.

Unfortunatly I dont have a variac, I might be able to find one lying around or cheap on ebay but for the mean time my mate told me that building a plug socket with a light bulb in it to take short currents should work, I think there called load lamps / light bulb current limiters. If theres a short in the circuit the light bulb should go bright to take up the current. will this work for the mean time?

To stop me going through these transistors like there's no tomorrow is there a circuit I could build to test the old ones for leakage?

Okay sweet, I'm away at the moment should be back on the amp fixing monday so I will have a look then

[nevetslab]

is there a circuit I could build to test the old ones for leakage?

Nothing occurs to me at the moment for this task, but hopefully one of the other members can assist on that one. And yes, the light bulb current limiter method will work, and very popular among the forum members.

Ah......it was the full wave bridge output wires you swapped. Yes, that would do catastrophic damage to all polarity-dependent devices down stream, sorry to say. I gather both 10A supply fuses blew, while the mains current was still flowing and trying to charge C16 & C17 in the wrong direction. Those too could well have been damaged, depending on how long it took for the mains current to stop (mains fuse blows, you turned it off). I'm now not sure you'd want to try and re-use any of the transistors in the amp that would have had high current flowing in the wrong direction. Those that did have sufficient resistors in the path to limit the reverse-current flow, then those would probably be the ones you'd want to check for leakage.

[Jazz P Bass]

"To stop me going through these transistors"

Without an attached load, which you DO NOT WANT when testing a 'just repaired' amp', the only real way to 'blow' the output devices is if both sides are turned on at the same time.

That would be indicated by a bias circuit, driver circuit or limiter circuit problem.

[wirelessliquid]

"To stop me going through these transistors"

Without an attached load, which you DO NOT WANT when testing a 'just repaired' amp', the only real way to 'blow' the output devices is if both sides are turned on at the same time.

That would be indicated by a bias circuit, driver circuit or limiter circuit problem.

I normally test the amp without a load I've just had alot of goes repairing this amp and if one of the transistors before the output transistors blow, they seem to damage the output transistors in some way. Usually they will seem fine with a test on a multimeter but then when brought up to voltage they don't seem to work properly and blow F2 and F3, replacing the output transistors in sets of 3 seems to fix this problem as I cant tell which on of the three output transistors is the one causing the fuse to blow.

[The Dude]

FWIW: A variac and current meter are invaluable if your doing much of this kind of thing.

[nevetslab]

With more hints in trouble-shooting this amp, you CAN make this work without the Drivers and Outputs installed (Q8-Q9-Q10 & Q16-Q17-Q18). That leaves Q7 & Q15 as your output stage, which still has the feedback network connected, so if all was well, you would have an output that should settle around 0VDC on the output. You just can't drive any kind of power from it, and, it may also have oscillation, not having the rest of the load on those pre-drivers. You can also lift the limiter circuit off of the pre-driver's bases, disconnecting CR2 & CR8 from the bases of Q7 & Q15, eliminating the current limiter circuit as culprits.

IF Q13 (bias xstr) is open, there will be NO current flow thru the voltage gain stage (Q5-Q13-Q14), which will turn Q7 & Q15 on hard, as their bases will be pulled towards the two power supplies by Q5 & Q14).

You can even go one step further....take Q7 & Q15 out of the circuit, jumper the collectors of Q5 & Q14 together, and tie that point to the output, so there's feedback from this voltage gain stage. It may oscillate, but, circuit-wise, it leaves you with a high voltage amplifier stage, with limited current drive capability.

While I have the luxury of lab power supplies, I've been known to do without the brute force supplies that can destroy output stages when something hasn't yet been found, and run the power amp on a variable-voltage/variable current dual tracking power supply so I can see what is happening, and NOT have to keep replacing failed xstrs. You are sort of limited to the light-bulb current limiter not having either a variac or external variable supplies in this case. But, at any rate, these are some methods of tackling your problem, while taking some of the expensive parts out of the equation.

We still need to learn WHAT the DC potentials are in the circuit, when you DO get it to stop blowing up. As I said before, there are enough clues on the schematic to guide you in finding what is and isn't working.

[wirelessliquid]

Thanks for the tips man, that will save plenty of time as well I've done a few years in college studying electrical and electronic and this is my first go at fixing a full electronic circuit of this size so any tips are greatly appriciated

Now I've got a power supply similar to this https://bit.ly/2N3tEWx that I found while snooping around at work a while ago, can I test the amp board with this in any way and work across the circuit in sections? I think this has a current limiter and short protection on it aswell so I could remove the output transistors and set the current really low to test it. My mate gave me a digital ossiloscope so when testing with the output transistors removed I should be able to see the 0VDC with this.

I'm going to have a snoop around today to try and find a variac I remember seeing one in the back of a cupboard a while ago and no one seemed to know what it was so that tells me no ones uses it so I might be able to take it home with me.

[nevetslab]

That dual supply would do nice. If it has the ability to slave one side to the other, all the better...then one knob adjusts voltage for both, so you can slowly raise the DC supplies while monitoring the circuit. Nice to have the lab supply fold current back when something goes wrong than see more $$ go up in smoke!

Without the driver and output xstrs, it looks like the circuit draws around 20mA thru the bipolar 45V supplies. Now, it looks like Q1 & Q11 are set up to supply +/- 15VDC to the rest of the amp's LV bipolar needs (op amps, discrete xstr circuits in the preamp). That's a little misleading, since it would really be +/- 14.4VDC, accounting for the Vbe drop on Q1 & Q11. No matter...in the ballpark. Looking at the voltage figures around Q1, it's collector stated to be 30VDC, and the supplies being 45VDC, that's 15V across 150 ohms, being 100mA for the preamp. Without it, the collector would be much higher, since the front end Q2 & Q3 only draw 3mA each.

Anyway, best of luck on the hunt!

Make sure the variac you found in the back of a cupboard is 0-240V rated (might be 0-280V), since you are in the UK.

[wirelessliquid]

So my power supply is this

Its got the slave on it which is good but it only goes to 30V, I think I can still use this to test a bit but all the calculations will be out

I couldnt find the variac today so the hunt will go on but I did find some pieces and mashed this together

Now hopefully I wont burn this amp up again in the mean time

I started having a look at what damage the blown R17 has done and semi conductor wise its wiped out pretty much everything
Q13 CE short
Q15 Shorts and OC
Replaced C13
Replaced CR10 CR4
Q16 OC
Q7 OC
Q6 OC
Q17 Short
CR2 Short
CR6 Short
Q19 Short
Q18 OC
Q9 Short
Q10 Short
Q8 OC
This is only what I found so far and replaced what i could but theres a couple diodes and Q7 2n3440 that I dont have spare so I've ordered them but they will take a couple weeks to arrive so I can get to testing at that point

I still have to check the preamp board (the one with all the pots on it) as when I reversed the 40VDC the other board is fed off of this one from the 8 Pin plug. I'm assuming there is damage on there aswell.

What would be the problem is the Variac was 280V, would it not work plugged into a 230V socket?

[nevetslab]

I figured Q7 & Q16 failed if you lost the drivers and output xstrs. The Q13 CE short would seem to prevent Q7 & Q16 from going fully north/south, but with the supplies reversed, it takes a bit of thought to see just what would fail and what wouldn't. Since both limiter xstrs Q15 and Q6 failed, along with CR2 shorting, I'd expect CR8 to short. You state CR6 is short, which is an odd part to see fail with that condition. Is that a typo?

You can replace Q13, as well as Q7 & Q16, leaving out the drivers and outputs Q8-Q10, Q17-Q19. The filter caps C2 & C7 were reverse-biased and should be replaced. C3 and C8 are probably ok, as I don't think they took a hit. C6 is probably ok, since you had OC on Q7 & Q16. The 15V zener diodes just became steering diodes when the supplies were reversed, and current-limited by R1 & R23, so probably ok.

If you have any BD139's and BD140's, they would suffice for Q7 & Q16 while working off of +/- 30VDC, but need to be more than twice the power supply voltage when everything is back together. These are only 80V parts.

I don't think all the circuits running off of +/- 15VDC got hurt, since these supply regulators weren't running when the voltages wre reversed. The supply fuses F2 & F3 would have opened anyhow with the flyback diodes CR4 & CR10 conducting, shorting out the supplies. I am concerned about the main buss caps C6 & C17, as those remained reverse-charged until the power supply was turned off (or the mains fuse opened).

On setting up your dual 30V supply, the Master supply being the one on the right, its' (+) is the pos supply output, then you'd connect (-) on it to the (+) on the other, which also is your GND connection. And the (-) of the Slave is your neg supply. Then, with the unit in Series Tracking mode, I assume the Slave supply's voltage pot isn't used, but the current pot would be...you'd set it high, and adjust the Master current pot for limiting the current from both. +/- 30VDC is more than enough to see the circuit work in terms of getting the high voltage amp up and running. Once you're there, THEN you can add the drivers, check again, and finally add the output's. It will still be adequate to run everything, open circuit.

You could leave the limiter circuit out while checking the rest of the front end. Once you know the front end is all working...up to and including the pre-drivers, then you could add the limiter circuit parts.

I'm assuming the variac that's there is a 280V autoformer....producing 0 thru 280VAC. If there is a second multimeter, it would be handy to make if function as an AC Ammeter (probably limited to 2A AC). Cobbling the wiring for that is of course tedious, to make it safe-and-sane.

[wirelessliquid]

No I found a short on CR6, I've had to order the diode to replace it, stange component to blow in this case I know but I still have to check the rest of the circuit so I may later see why.

Yeah C6 is all good, I was thinking to replace all electrolytic 22uf and 10uf anyways as I have a million spares, theres like some thermal white tac on the back of the Q11 and Q1 which stick to the caps C2 and C7 and also small heat sinks. Do you know what this stuffs called as I'll need some more to stick back to the caps after

I've ordered some of these for the mean time as Q7 is going to take like 20 days to arrive in the post ¬.¬

Hopefully the other board is all okay then C6 seems all good but I havn't got round to C17 yet and everything thats blown seems to be opposite on the circuit so far so hopefully C17 is good aswell but no problems if it needs changing.

I'll have to give this a try to a multi meter first just to make sure I get it right. What about the common GND pin on the power supply, I remember setting up an OP AMP in college that had to have the dual supply and we used the common ground then.

Just to be clear what makes up the limiter circuit

All is good as I have 2 multimeters just need to swipe the variac when i have a chance