So, I really hope someone with the knowledge and a generous heart can help me.
So, I have a HK990 amp and was doing the bias adjustment as my right channel was running hotter than the left. Looking at the manual, it says this:
AMP Adjustment Iding Adjustment
Precaution for handling measuring instrument.
The ground side of the measuring instrument to be connected to the speaker terminal of this unit must be kept in floating condition because this unit is equiped with the floating balanced power amplifier.
So what's the problem, you ask? Yep, I shorted the probe between the pad and something else, maybe normal ground.
The slow blow fuse blew, needless to say, as does the replacement fuse.
Steps taken so far:
1. Disconnect the transformer for the right amp and power up. No problem, everything works as usual with the system and the left amp.
2. Examine fuse board - nothing burned or odd found.
3. Removed and examined RHS amp. Visually, I found nothing. No burn marks. I tested, in circuit, the output transistors and they all behaved the same. I didn't fully disassemble the heatsink so the diodes mentioned in the schematic might not be visible.
So, needless to say, I need help. I've rebuilt numerous amps without doing something so stupid before, so I feel confident in replacing components to fix this, I just need guidance.
Attached are the schematics and parts list.
Please help with any suggestions on where to start. I'm actually out of town but willing to add components to an existing order on the chance that it's the problem.
What do you guys think?
Thanks for your time and take care,
Peter
So, I have a HK990 amp and was doing the bias adjustment as my right channel was running hotter than the left. Looking at the manual, it says this:
AMP Adjustment Iding Adjustment
Precaution for handling measuring instrument.
The ground side of the measuring instrument to be connected to the speaker terminal of this unit must be kept in floating condition because this unit is equiped with the floating balanced power amplifier.
So what's the problem, you ask? Yep, I shorted the probe between the pad and something else, maybe normal ground.
The slow blow fuse blew, needless to say, as does the replacement fuse.
Steps taken so far:
1. Disconnect the transformer for the right amp and power up. No problem, everything works as usual with the system and the left amp.
2. Examine fuse board - nothing burned or odd found.
3. Removed and examined RHS amp. Visually, I found nothing. No burn marks. I tested, in circuit, the output transistors and they all behaved the same. I didn't fully disassemble the heatsink so the diodes mentioned in the schematic might not be visible.
So, needless to say, I need help. I've rebuilt numerous amps without doing something so stupid before, so I feel confident in replacing components to fix this, I just need guidance.
Attached are the schematics and parts list.
Please help with any suggestions on where to start. I'm actually out of town but willing to add components to an existing order on the chance that it's the problem.
What do you guys think?
Thanks for your time and take care,
Peter
Last edited:
I surmise that the DMM and tester may be the entirety of your test equipment?
It's arduous to troubleshoot without being able to probe active circuitry, and applying full power blows the fuse--- a vexing dilemma. If you have adjustable bench supplies, that would be excellent. Lacking that, you might investigate using a Dim Bulb Tester that you'll find discussed here in the forum.
Give us a bit of feedback on these opening thoughts, and I'm sure you'll get lots of help.
OP - Once you get it functioning, here's a closed thread with useful info for the HK990 biasing operation.
https://www.diyaudio.com/community/threads/harman-kardon-hk-990-bias-unstable.308453/
https://www.diyaudio.com/community/threads/harman-kardon-hk-990-bias-unstable.308453/
You didn't specify which resistors you are referring to, but if R26 and R27, you're measuring the transformer winding, so continuity is to be expected.
Also check diodes D1-D4 in the other rectifier section. By unplugging the Faston connectors, you should be able to test the diodes in circuit.
With Faston's still removed, apply power. The fuse should not blow. You could measure AC volts at transformer secondaries to assure all is well with the transformer.
Also check diodes D1-D4 in the other rectifier section. By unplugging the Faston connectors, you should be able to test the diodes in circuit.
With Faston's still removed, apply power. The fuse should not blow. You could measure AC volts at transformer secondaries to assure all is well with the transformer.
BTW, the D1-D4 diodes do the "heavy lifting"--- they supply +/- 60V power at high currents to the output transistors; the -/-80V rails provide elevated supply voltage to driver stages. We can hope that the problem lies in the D1-D4 area, as they would be easier to troubleshoot/isolate and replace. If D1-D4 prove to be OK, we'll figure out what to try next.
From your first post, I gather that the errant test probe was probably on TP4 or TP8. Any guess which one might have taken the hit?
From your first post, I gather that the errant test probe was probably on TP4 or TP8. Any guess which one might have taken the hit?
First a disclaimer--- I have no familiarity with this amp and am working solely from schematics and the manual. I do have 50 years experience as an R&D EE on the bench. That may be worth something.
I'm uncertain of HK schematic conventions, so maybe you can confirm for me: referring to amplifier power schematic (pgs. 106 manual/ 108 PDF), I guessing that speaker relay contacts on right of drawing are associated with coils on left side? I have the impression that the Fault line goes to the microprocessor and that there's no intervention path to the power amps in the event failure, other that perhaps opening speaker relays. Any contrary knowledge?
I recommend double checking diode junctions of the output transistors (starting near TP4) before we resort to more risky experiments. Looking especially for shorts across transistor junctions that might be responsible for shorts to +/- 60V rails.
If you find no suspicious components, a next experiment: I believe it's safe/nondestructive to disable +/- 60V power by disconnecting Faston JP2 and JP8 and to reapply power--- but be on the alert with the power switch. The fuse shouldn't blow, +/- 80V rails should be active. Of course amp won't work, and there might be a fault indication.
If the results of this experiment are encouraging, for next step in fault isolation, investigate if there is any convenient facility to open the paths between the 60V rails and the output transistors, i.e. path between C12 and transistor array, C3 and its array. The objective is to test for appearance of +/- 60V on power supplies without applying volts to the output transistors.
I'm uncertain of HK schematic conventions, so maybe you can confirm for me: referring to amplifier power schematic (pgs. 106 manual/ 108 PDF), I guessing that speaker relay contacts on right of drawing are associated with coils on left side? I have the impression that the Fault line goes to the microprocessor and that there's no intervention path to the power amps in the event failure, other that perhaps opening speaker relays. Any contrary knowledge?
I recommend double checking diode junctions of the output transistors (starting near TP4) before we resort to more risky experiments. Looking especially for shorts across transistor junctions that might be responsible for shorts to +/- 60V rails.
If you find no suspicious components, a next experiment: I believe it's safe/nondestructive to disable +/- 60V power by disconnecting Faston JP2 and JP8 and to reapply power--- but be on the alert with the power switch. The fuse shouldn't blow, +/- 80V rails should be active. Of course amp won't work, and there might be a fault indication.
If the results of this experiment are encouraging, for next step in fault isolation, investigate if there is any convenient facility to open the paths between the 60V rails and the output transistors, i.e. path between C12 and transistor array, C3 and its array. The objective is to test for appearance of +/- 60V on power supplies without applying volts to the output transistors.
Last edited:
Yes, great!
I replaced the bridge rectifier and connected the fastons except JP2 & JP8. The unit started up and didn't blow the fuse.
I only left it on for a minute, but after the initialization, the relays click and power is sent to the amp boards. That's when the fuse would blow.
No idea about the relays you asked about.
... and to your next conclusion, yes, C3 and C12 have continuity on both negative and positive rails.
I've checked all the output transistors for shorts and found none. I assume there isn't much more I can do with them?
I replaced the bridge rectifier and connected the fastons except JP2 & JP8. The unit started up and didn't blow the fuse.
I only left it on for a minute, but after the initialization, the relays click and power is sent to the amp boards. That's when the fuse would blow.
No idea about the relays you asked about.
... and to your next conclusion, yes, C3 and C12 have continuity on both negative and positive rails.
I've checked all the output transistors for shorts and found none. I assume there isn't much more I can do with them?
Last edited:
Since you haven't found any suspicious components with your DMM, we still need to locate a presumed short. I suggest improvising a way to deliver approximately 10 to 30mA AC into the D1-D4 bridge via the Fastons so that we can trace current flow without making smoke.
You would install two current-limiting series resistors between the JP2 & JP8 Faston terminals and their respective transformer leads. 1.8K, 2W resistors would work, but may not be available in your junk box. Other improvs are possible: eg. you could power JP2 & JP8 via 220 Ohm resistors tapped into pins 1&3 of J11 connector on PDF page 113. An external, center-tapped low voltage transformer is another possibility.
In any case, with suitable AC current available at the D1-D4 bridge, look for anemic DC output voltage produced at C3 and C12. We'll explore from there.
Steve
You would install two current-limiting series resistors between the JP2 & JP8 Faston terminals and their respective transformer leads. 1.8K, 2W resistors would work, but may not be available in your junk box. Other improvs are possible: eg. you could power JP2 & JP8 via 220 Ohm resistors tapped into pins 1&3 of J11 connector on PDF page 113. An external, center-tapped low voltage transformer is another possibility.
In any case, with suitable AC current available at the D1-D4 bridge, look for anemic DC output voltage produced at C3 and C12. We'll explore from there.
Steve
I'm on a road trip, so my presence may be erratic.
You should probably use your ohmmeter to preset R53 on both channels to minimum resistance; this will minimize bias currents and give you the best chance of having enough supply voltage to produce useful clues.
So you install the two current resistors in series with the Faston clips and their respective transformer windings and apply power. With any luck, you'll observe some DC generated at C3 and C12--- with luck, at least a few volts. If there's no B+ or B- voltage, suspect the bridge rectifiers D1-D4.
Since you are blowing fuses, the presumption is that something is shorted. I suggest connecting your black voltmeter lead to ground and poke around with red probe. Be most suspicious of the power supply rail that seems the most heavily loaded. Use the undamaged (presumed) channel for comparison.
When the amp is working, there should be about 2.4VDC across Q32. If the Vbe multiplier (aka bias spreader) isn't showing activity, confirm that front end is enabled. You might have to install a temporary short across pins 3 & 4 of Mute optocoupler U1.
If you can't find anything suspicious, you can annotate observed voltages on a schematic and post a picture of the data. Or tabulate data. Measure voltages re ground:
+80V rail
+60V rail (will be only a few volts in present condition)
Base Q18
Base Q17
Bases Q12 through Q16
TP1 through TP4
Amp Output
TP5 through TP8
Bases Q1 though Q4
Base Q7
Base Q6
-60V rail
-80V rail
Report negative voltages as negative numbers.
But if you don't observe at least 3V produced between the +/- 60V rails, report a few salient voltages rather than all that tedious data.
Good luck. TTYL
You should probably use your ohmmeter to preset R53 on both channels to minimum resistance; this will minimize bias currents and give you the best chance of having enough supply voltage to produce useful clues.
So you install the two current resistors in series with the Faston clips and their respective transformer windings and apply power. With any luck, you'll observe some DC generated at C3 and C12--- with luck, at least a few volts. If there's no B+ or B- voltage, suspect the bridge rectifiers D1-D4.
Since you are blowing fuses, the presumption is that something is shorted. I suggest connecting your black voltmeter lead to ground and poke around with red probe. Be most suspicious of the power supply rail that seems the most heavily loaded. Use the undamaged (presumed) channel for comparison.
When the amp is working, there should be about 2.4VDC across Q32. If the Vbe multiplier (aka bias spreader) isn't showing activity, confirm that front end is enabled. You might have to install a temporary short across pins 3 & 4 of Mute optocoupler U1.
If you can't find anything suspicious, you can annotate observed voltages on a schematic and post a picture of the data. Or tabulate data. Measure voltages re ground:
+80V rail
+60V rail (will be only a few volts in present condition)
Base Q18
Base Q17
Bases Q12 through Q16
TP1 through TP4
Amp Output
TP5 through TP8
Bases Q1 though Q4
Base Q7
Base Q6
-60V rail
-80V rail
Report negative voltages as negative numbers.
But if you don't observe at least 3V produced between the +/- 60V rails, report a few salient voltages rather than all that tedious data.
Good luck. TTYL