Hartke B90 no sound

[oleskool]

  Oh well the line fuse went as soon as I hit the power. will get another tomorrow to see if it happens again.

[Roly]

Oh well the line fuse went as soon as I hit the power. will get another tomorrow to see if it happens again.

Don't do that again - it will, and you will only cause more damage.

Do this;

You will also need to knock up a limiting lamp, as per;

http://www.ssguitar.com/index.php?topic=2093.0

[oleskool]

Thanks Roly!!! Will do.

[Roly]

Okay, so if it's blowing fuses, once you get your limiting lamp in series you will find that it lights brightly because the amp is shorting.

With limited power on the amp the thing then is to try and isolate what, exactly, is shorting.

You also need to disconnect the loudspeaker, and a simple way to do this in this particular amp is to plug a blank plug, or lead with nothing connected to the other end, into the headphone socket.  However you also will need to get inside the chassis to take voltage measurements so it may be easier to disconnect it altogether, however.

With the limiting lamp all the voltages will be very low compared to their normal values, but what we are looking for is some indication of what is shorting to what.

You will need to locate the main electrolytic caps C34 and C35, 4700uF/63V.  Using the chassis as the reference we hope to see a few volts positive on the +ve end of C34, and a few volts negative on the -ve end of C35.

We also need to find the voltage on the output line, the ungrounded speaker connection.  What is of particular interest here is if the output line is the same as the +ve supply, the -ve supply, or somewhere in between. (D12 and D13 are between these three points, +ve, -ve, and output).

If the output is +ve this suggest you have one or more shorted transistors on the +ve side, Q11/12/13, if on the -ve side Q5/8/9.

If there isn't much voltage at all it is possible there are shorted transistors both top and bottom, and the bridge rectifier D9 may also be shorted as a result.

The next step is to remove any suspect transistors from the circuit, carefully noting/marking where each came from, and testing out of circuit.  While Q13/14 and Q8/9 are out of circuit you can also power up for a few seconds and see if the short has cleared (the lamp settles to a dim glow or less).  This will clear the bridge rectifier D9 (and you should get tens of volts on both +ve and -ve supply rails), but if the short remains then D9 has most likely failed as well and will require replacement.  You can do this and retest before replacing any of the transistors.

Anyhow, using the limiting lamp, take some measurements and report back.

[oleskool]

Hello again.
Well the first test C34=1V, C35=-1V, speaker=1V.
Pulled Q8, Q9, Q12, Q13 They all test good with a multi meter.
Plugged back in switched on, very dim bulb for about two seconds then it fades out.
D12=   -.384V, D13= 104.7V ?
I guess I should test Q5 & Q11?

[Roly]

  >Well the first test C34=1V, C35=-1V, speaker=1V.


+/- 1V supplies tell us we have a short right across.

Speaker at +1V says we have an output transistor shorted between the +ve rail and the output, i.e. Q12 and/or Q13.


  >Pulled Q8, Q9, Q12, Q13 They all test good with a multi meter.


Given the measurements above it seems to me unlikely that Q12/13 are actually okay.

Firstly you need to measure the Base-Emitter and Base-Collector junctions.  These should both look like diodes, between 500mV "500" and 700mV "700" forward, and open "1___" in reverse (on your multimeter diode test range).  Then you need to measure the Collector-Emitter path which should be open both polarities.


  >Plugged back in switched on, very dim bulb for about two seconds then it fades out.
  >D12=   -.384V, D13= 104.7V ?


Now something significant has changed because you no longer have the short between supplies, but the output rail (join of D12 and D13) is still stuck to the +ve rail, again strongly suggesting that Q12 and/or Q13 is short C-to-E.


  >I guess I should test Q5 & Q11?


Yes, this is a good idea, but in any event I think you still have a shorted transistor at Q12 and/or Q13 (and it's possible that you also have Q8 and Q9 blown open).

Because this amp uses pairs of output transistors it is possible to remove only Q13 and Q9, then Q12 and Q8 to help isolate which are faulty, just as long as you don't try to drive the amp hard into a load (but you should have the speaker disconnected anyway).

I have to say that this power amp circuit is not well drawn; it's actually quite conventional but drawn in a confusingly offset manner.

This is not identical - it only has one set of output transistors instead of two, and the initial input stages are "upside down" to yours, but this is the same general arrangement, and seeing this may help you find your way better around the Hartke circuit.

[oleskool]

OK I hope i didn't mess up, but earlier in this post there was a spot I needed to jump from collector to collector at Q8 and Q9. Well when I pulled Q8 and Q9 I also pulled that jumper. I replaced the jumper Q8,Q9,Q12, and Q13 are still out. Across D12=  -51.3v. -51.3v. Black lead connects at end of diode with the band for both. At first these were jumping up and down. Now they are fairly stable with these two measurements. I was talking to a guy he said the zener D5 is most likely bad. He also advised to change all the transistors. He also noted which does sound correct, that if J5 was bad it would cut the speaker out. But he is getting said parts.
I'm getting 20mv at the speaker. C34=51v and C35= -51v.

[Roly]

> I'm getting 20mv at the speaker. C34=51v and C35= -51v.


I'm having a bit of trouble following what you are doing, but again the conditions have changed (for the better) and you are now getting very reasonable readings - plus and minus supplies, and very little DC at the speaker; this is all good.  As you may have worked out, D12 is connected between the +ve supply and the output rail, and D13 is connected between the output rail and the -ve supply, making these two a handy point to find the three most important voltages.


> I was talking to a guy he said the zener D5 is most likely bad. He also advised to change all the transistors. He also noted which does sound correct, that if J5 was bad it would cut the speaker out.


Measure the voltage across the zener -  measurement trumps a wild guess any day.  If you have 15 volts it's okay.

Given that you now have good supplies and only 50mV across the speaker, advice such as "replace all the transistors" is, at best, let's say, not very well informed.  ( :duh)  This is what I call "blunderbuss" servicing and often leads to more trouble and expense than doing it properly.  "Properly" means locating which component(s) are faulty, and replacing them.  The blunderbuss mode is often used by people who can't actually locate faulty components, it's hit and miss, and not at all professional or reliable.

While he is right that J5 could interrupt the output, it in no way explains a) why both supplies were being loaded down to only one volt, or b) why the output rail was previously stuck at the +ve rail.  This guy isn't even considering the available evidence, and when you're done you will be able to give him (better) advice.

[oleskool]

Thanks. I was hopeful that these measurements were more reasonable. There is a spot on the circuit board that is fried. It is a very short run. The connection between the collectors of Q8 and Q9. I used a piece of wire to bridge that gap. The first measurements when the supplies were 1v, and more important stuck + were taken after I removed Q8 and Q9. When I removed them I also removed the wire that bridged the gap. Another thing I learned along the way, if something changes and doesn't make sense, look at the work you did. Look at the schematic and imagine the entire line that connects the collectors of Q8 and Q9 erased. Well that is what my board looks like. I have placed a piece of wire there to bridge the gap, and then took new measurements. I will try to put up a few pictures. My camera is eating batteries.

[Roly]

Ahh so - track blown off.  Happens.

May I then suggest that you fit a couple of fuse holders somewhere between C34, +ve rail, and C35, -ve rail, and everything else?

P=150 Watts
R=8 Ohms

P=I^2*R
I^2 = P/R
I = root(P/R) = (P/R)^0.5
(150/8)^0.5 = 4.33012702 amps

So initially try say 4 amps in each rail and see how you go.

These won't prevent transistor failure, but they will help to minimise damage if that should happen again.