Harman Kardon HK 690i Suggestion Needed

[mbz]

Hiding in the corner of the schematic, D405 is a maybe, likewise D406 on other channel.
Maybe these are breaking down under load which means they will probably test ok.
Measure Vdc on anode side, both channels, however any voltage would appear as dc offset...

Like mentioned before, the rails come from the unregulated supply, it will take a sizeable short
for it to drag the rails down. Have you replaced the main filters C6, C7 4700uf/50V in schematic.

No buzz from transformer like it's working hard?

 

[ivandezande]

Hiding in the corner of the schematic, D405 is a maybe, likewise D406 on other channel.
Maybe these are breaking down under load which means they will probably test ok.
Measure Vdc on anode side, both channels, however any voltage would appear as dc offset...

Like mentioned before, the rails come from the unregulated supply, it will take a sizeable short
for it to drag the rails down. Have you replaced the main filters C6, C7 4700uf/50V in schematic.

No buzz from transformer like it's working hard?

No buzz. Only a few small caps that I had on hand were replaced. I would be happy to go gung-ho on replacement however. D405 and D406 just show a few steady mV.

 

[mbz]

Rather not go gung-ho just yet, however if you have replacements for C6, C7, it would be worth a try.ationale being the main caps are not holding their charge
and the voltage quickly collapses. Alternativel if you can measure them that's good also.

 

[ivandezande]

Rather not go gung-ho just yet, however if you have replacements for C6, C7, it would be worth a try.ationale being the main caps are not holding their charge
and the voltage quickly collapses. Alternativel if you can measure them that's good also.

I have 3,300uF 50V, I obviously wouldn't leave them in long term but would they work for the time being?

 

[mbz]

Hold off on replacing the main filters.

The power supply is not as simple as I thought. Those main filters also
feed into a regulated supply, point D is for the positive rail that's
collapsing. Suggest you verify the regulated supply at point D, ie,

measure at Q623emitter (point M) and also at Q909 collector (point L)
Zeners D903 and D605 being candidates or downstream from L&M. Maybe
R655, 7 are getting warm.

This supply is for the MM&MC amps, it can be isolated by lifting the connector
P603 on PCB-3, see pic, removes both +/- supply at points D&E

 

[ivandezande]

Same behavior with the phono power supply disconnected. Point M: Steady 22V Point L: 14.3V Zener diodes check fine in circuit. Resistors mentioned have air clearance and look to be fine, in their location it is hard to get to them to check temperature.

 

[mbz]

The regulated voltages are almost perfect on DBT, difficult to blame the
feed (unregulated) but there is nowhere else to go. Looking at main filters
C6,7 4700uf/50V and rectifier bridge and transformer winding/AC supply.

If you have an oscilloscope can you measure the AC ripple at C6, point D.
Expect less than 500mVpeak-peak. This should indicate the health of C6.
If you have a true RMS DMM you could try that. My DMM (elcheapo) would
think AC ripple on DC is large AC, maybe 50VAC.

Since the problem is somewhat intermittent, the rectifier bridge would
probably pass diode test.

Regarding replacing C6/7 with 3300uf, It should be ok since the amp is doing
nothing. You should check connections on that D6 cap.

 

[ivandezande]

No oscilloscope with me and I probably get some weird looks in the physics lab if I brought the amp there. I replaced just C6 and 7 and the same behavior occurs. My DMM shows very low AC ripple across either cap.

 

[mbz]

The sm is not real friendly but looks like the only place that D feed goes to is the power amp (MC/MM stage was disconnected, problem persisted).
Think the only course left is to track down the extra current by measuring voltage drops across resistors and comparing with the other channel.

Bias was zero, set the bias to 15-20mV(on DBT)
Measure volt drop across R461 (earlier you said about 1Vdc)
Compare with other side.

If R461 volt drop ten percent high on problematic side the problem after R461.
Measure at R451, R443, R439, R427, R409, R407, R435.
C413 replaced?

If R461 volt drop is lower or about the same then check earlier branches
Measure at R465, 469 and 483

1 volt across R461 is 0.05Watts (1 W resistor), so problem probably R465, R469, Q441 or D405 branch

 

[ivandezande]

The sm is not real friendly but looks like the only place that D feed goes to is the power amp (MC/MM stage was disconnected, problem persisted).
Think the only course left is to track down the extra current by measuring voltage drops across resistors and comparing with the other channel.

Bias was zero, set the bias to 15-20mV(on DBT)
Measure volt drop across R461 (earlier you said about 1Vdc)
Compare with other side.

If R461 volt drop ten percent high on problematic side the problem after R461.
Measure at R451, R443, R439, R427, R409, R407, R435.
C413 replaced?

If R461 volt drop is lower or about the same then check earlier branches
Measure at R465, 469 and 483

1 volt across R461 is 0.05Watts (1 W resistor), so problem probably R465, R469, Q441 or D405 branch

Sorry for taking forever, life got pretty busy for a bit. One thing - not sure if there's a problematic side to this amp. May be safer to assume it could be anywhere.
L R461: .43 V
R R462: .43 V
R465: .66V
R469: .025 V
R483: Same as bias

I think I'm going to bite the bullet and buy new caps, some small transistors for those who were corroded a little, and replacing those bias diodes with 1N4148. In the meantime I'll happily keep measuring.

 

[skyegreen424]

Sorry for taking forever, life got pretty busy for a bit. One thing - not sure if there's a problematic side to this amp. May be safer to assume it could be anywhere.
L R461: .43 V
R R462: .43 V
R465: .66V
R469: .025 V
R483: Same as bias

I think I'm going to bite the bullet and buy new caps, some small transistors for those who were corroded a little, and replacing those bias diodes with 1N4148. In the meantime I'll happily keep measuring.

Did you ever figure this out/was it a cap problem? I've got one on my queue with similar symptoms (dim bulb gets brighter with inputs being changed)

 

[ivandezande]

Did you ever figure this out/was it a cap problem? I've got one on my queue with similar symptoms (dim bulb gets brighter with inputs being changed)

Can’t remember, think I gave up.