Harman/Kardon 330B - burnt resistors, mismatched transistors and what else.

Winters

Active Member
Hello,

I am currently trying to repair a vintage Harman/Kardon 330b stereo receiver in which one channel died burning 100 ohm and 22 ohm resistors on the final stage circuit board.

This is actually the first time I'm repairing solid state gear. I have serviced several vintage stereo and guitar amplifiers but they were all tube. Hence I'm no expert and could use some advice.

I have already measured the voltages around the burnt resistor and there is 50V DC on two legs of the nearest transistor (TR505 which in this case is C1345) as well as on several resistors in the vicinity. This is obviously wrong.

The voltages on the other channel, which works well, are all within specs.

Another big problem I see is that someone has obviously been in there before and replaced all of the transistors on the right channel output circuit board.

The transistors on the left channel output PCB are: (this channel
C1335
C1345
C458
A777
C1509

The outputs are: 2x 1213 (RCA HB7747).

Right channel output PCB has:
2x BC550B
C5468
C639
C640

The outputs are: 2x BD201.

The rest of the circuitry is identical - caps, resistors etc. Capacitors all look good, no leaking, no bulging. I am probably going to replace all of them anyway once I get it to work correctly.

I take for granted that some transistors in the dead channel will require replacement. The question is - what to replace them with? Should I get the same types that are installed on the other channel? Should I replace every transistor on the output board and try to bring this piece to original condition?

Thank you for your help.

Best regards,
Pete
 
I take for granted that some transistors in the dead channel will require replacement. The question is - what to replace them with? Should I get the same types that are installed on the other channel? Should I replace every transistor on the output board and try to bring this piece to original condition?

Hi Pete,

HK330B is a great model to work on. Well documented, easy to get to, nothing exotic or expensive, every part has a modern replacement.

My recommendation would be:

  1. Replace all the transistors, in both channels, on the main amp board (BOM list below)
  2. Replace all the electrolytic caps
  3. Replace the emitter resistors

Beyond the main amp board, the control amp (preamp) features six or eight 2SC458 transistors that will go noisy, if they haven't already. Replace them with KSC1845. You'll be buying several 1845s to fix your main amp, go ahead and buy 10+ so you get the quantity discount and buy 100 (only $6 total) so you can match the gain between the pairs you replace.

The electrolytic caps in the preamp will need replacement, too. There are about 10 electrolytic caps there that would be better replaced with 1 uF 50 V film caps like the WIMA metalized polyester series.

BOM for transistors in the main amp:

component qty old-part replacement-part
Q401,Q402 2 2SC1335-D 512-KSC3503DS
Q403,Q404 2 2SC1345-D 512-KSC1845FTA
Q405,Q406 2 2SC458-C 512-KSC1845FTA
Q407,Q408 2 2SC1509(Q) 512-KSC2690AYS
Q409,Q410 2 2SA777(Q) 512-KSA1220AYS
Q411-Q414 4 2SC1107 863-MJE15030G
 
Good stuff from Dlucy, but I suggest that you do the repairs on the bad channel first, get it working, before doing any replacements in the good channel.
 
Thank you very much for your replies. I find them very helpful. This is one of the truly great things about AK.

I'm going to proceed step by step, get the dead channel working first and then replace the transistors on the other side, then replace the capacitors.

One more concern - do the replacement transistors require any changes/modifications to the circuitry? I come from the tube stuff (The Fisher 500B is my daily driver) and replacing tubes with different types often requires rewiring the sockets / changing some values etc.
 
Its great to listen to all our opinions and then chart your own course. Excellent. Best way to go.

I suggest looking at shipping costs so you can tell if buying all the parts in one order or buying in multiple orders fits your approach better. Realizing 3 more 1845 transistors at $0.21 each but $7.99 shipping is a lesson I still remeber to this day.

Many of us advise keeping the device models identical in each side of the circuit. That isn't a huge deal with this low powered unit. But something to know.

The only circuit adjustment you'll need is to the two trimpots in each channel of the main amp. You may want to physically test them before adjusting. Ive seen more than one just fall apart at the first significant force. That is something you do not want happening when the circuit is powered.

The replacement transistors will operate nearly identically in this circuit.

The one thing to watch for is the varistors hung off or attached to the heatsink immediately above the output transistors. Varistors are nearly impossible to find, the leads can be fragile, and while you can make replacements from common diodes, the fab process is a little tedious.
 
I'm close to finishing the rebuild of the dead channel. I have one more question - can I power this receiver with a variac to make sure nothing explodes in my face if something goes wrong? I always do that with old tube amplifiers but I have never powered a solid state unit this way.
 
can I power this receiver with a variac
Yes if no FET transistors. Dim bulb is better for SS and cheap to build.

I find using my 10amp variac more to regulate vacuum cleaners and blenders house hold appliances and my sander grinder as a polisher... LOL. fun too! :D
 
So, I've replaced most of the dead channel - all transistors, all electrolytic capacitors, emitter resistors and a few suspicious looking resistors. I managed to get all of the recommended replacement transistors.
And it didn't help. Nothing seems to be overheating/smoking/burning but the voltages are all over the place and it obviously doesn't work.

I have written down my voltmeter readings:

Power output transistors - MJE15030G
Q414:
E: 0V
C: 53V
B: 0V

Q412:
E: 53V
C: 53V
B: 53V

KSA1220 (Q410):
B: 52V
C: 0V
E: 52V

KSC2690 (Q408):
B: 53V
C: 53V
E: 53V

KSC 1845 (Q406):
E: 54V
C: 54V
B: 54V

KSC 1845 (Q404):
E: 2V
C: 53V
B: 0V

KSC3503 (Q402):
E: 4,7V
C: 2,1V
B: 0V

I didn't measure all of these voltages before the replacement but what I did measure was very close - about 50V DC wherever I put the probe This makes me believe that the problem is not on this PCB but somewhere else. But where can it be?

DSCF0466.JPG


Best regards,
Pete
 
Last edited:
How do the voltages you're getting in this newly-rebuilt channel compare to the exact same points in the working channel?

When you swap the pre-out and main-in jumpers on the rear panel (switch the jumpers out for RCA cables, then connect pre-out right to main-in left and vicea versa), does the dead channel stay the same?
 
How do the voltages you're getting in this newly-rebuilt channel compare to the exact same points in the working channel?

When you swap the pre-out and main-in jumpers on the rear panel (switch the jumpers out for RCA cables, then connect pre-out right to main-in left and vicea versa), does the dead channel stay the same?

1. Totally different. The voltages in the working channel are very close to those in this schematic (page 17): http://manuals.harman.com/hk/Service Manual/HK330b sm.pdf One thing that strikes me is that the correct voltages in several places are 26V and I'm getting 52-54V which is double the correct value.

2. I will do that but as far as I understand, this bridges only the signal path and wouldn't have caused damage, would it?. This receiver belongs to a friend of mine and was working fine until the day something burned inside (one resistor was completely burnt, another one had signs of overheating).
 
1. Totally different. The voltages in the working channel are very close to those in this schematic (page 17): http://manuals.harman.com/hk/Service Manual/HK330b sm.pdf One thing that strikes me is that the correct voltages in several places are 26V and I'm getting 52-54V which is double the correct value.

2. I will do that but as far as I understand, this bridges only the signal path and wouldn't have caused damage, would it?. This receiver belongs to a friend of mine and was working fine until the day something burned inside (one resistor was completely burnt, another one had signs of overheating).

Ok, you are getting the full power supply voltage (~53VDC) at places you should not. The next step I would take is tracing the power coming in from the power supply (at FU3) to the failing side of the power amp board. You are looking for the spot where 52 VDC is supposed to be dropped to (much less). Something is allowing all that voltage through (sounds like a shorted component) to stuff that is expecting much less.
 
After writing all the voltages you've reported on a copy of the schematic and then overlaying the expected voltages, it seems to (hobbyist) me that your bad voltages start at Q406. That is partly controlled by the trimpots.

Try pulling your trimpots out and checking their values as you move them. If one is shorted or open, I think you'll deliver the wrong voltage to the Q406 and subsequently all those after it.

The trimpots in there are open frame, prone to oxidization, and I've even seen one set on a 330B just fall apart as they were touched.
 
After writing all the voltages you've reported on a copy of the schematic and then overlaying the expected voltages, it seems to (hobbyist) me that your bad voltages start at Q406. That is partly controlled by the trimpots.

Try pulling your trimpots out and checking their values as you move them. If one is shorted or open, I think you'll deliver the wrong voltage to the Q406 and subsequently all those after it.

The trimpots in there are open frame, prone to oxidization, and I've even seen one set on a 330B just fall apart as they were touched.

Will do that. Thank you very much!
 
The suspects are R418, R420, R422, RV402, C414, and maybe C410. One of those is shorted... I bet. The voltage 53v is coming in but it is not getting dropped to 24v.

IMG_1164.JPG
 
The suspects are R418, R420, R422, RV402, C414, and maybe C410. One of those is shorted... I bet. The voltage 53v is coming in but it is not getting dropped to 24v.

I have checked these components with an ohmmeter without removing them from the circuit - my thinking is that although it is not the best way of measurement, it would show a short as the signal from the meter would go right across, right?. Not only nothing seems shorted, everything measures within spec. In this case is it possible that something shorts only when voltage is applied? Bad varistor? (It measures 17 mega ohms, just like the one in the other channel but who knows what happens when voltage shows up?).

I have also measured the variable resistors - no shorts there either.
I have also looked at the traces in this area - all looks good, it doesn't seem that I shorted anything by using too much solder.

I'm already scratching my head... with most of the channel replaced, maybe it would be worth a shot to replace EVERYTHING in there? I mean, variable resistors, film/ceramic caps, remaining resistors...
 
Totally understand.

The way to check resistors and capacitors in-circuit is to "lift one leg". In other words, desolder one leg, lift that one leg out of the PCB hole, then measure the component.
 
Totally understand.

The way to check resistors and capacitors in-circuit is to "lift one leg". In other words, desolder one leg, lift that one leg out of the PCB hole, then measure the component.

Sure. I will do that. Just wanted to test this quickly.
 
Here is a simple test you can do without desoldering anything.

53 VDC should be arriving at one side of R418. That resistor is there to drop the voltage down to or part way to 25 VDC. Read the voltage on the supply side of R418 and then on the other side of R418. There should be a big voltage drop.

The same is true for R420. It will be getting whatever R418 is giving it, then R420 should be dropping the voltage down to ~23 VDC. It supplies voltage to collector of Q406. If either of R420 or R418 aren't working, at the correct ohm values, they won't be dropping the 53 VDC down to 23 VDC.

Or something else is letting the full 53 VDC of the supply get to the collector of Q406.
 
Back
Top Bottom