HK570i Resuscitation issue (amost there but stuck)

mrwasabihead

Fix or restore..fix or restore..fix or restore?
Subscriber
Hello, I found an HK570i at a second hand store that obviously had an issue but for $5.00 US I thought I'd give it a shot. Got it home and plugged it into the DBT (dim bulb tester) and "bright light!". So I started checking things out. I found two of the four output transistors were shot and many of the associated driver transistors were as well. I had suitable replacements on hand and now I've got good voltages (...w/ the dbt of course) all around but the thing that's got me stumped is this: C8, one of the power supply caps will not discharge when the unit is powered down. In fact it'll hold it's negative supply charge of (~ -36V) all day long. I've checked and rechecked, thought and thought some more and now it's time to ask for help. Even though I've got expected voltages from the supply through the preamp and amp circuits currently this cap issue is something I've not run into before and don't want to take it off of the dbt before asking for help. By the way all of the other caps discharge readily. And I've also replaced C8 and it's 3 other comrades just to rule them out.

Here is a link to the SM: https://www.hifiengine.com/manual_library/harman-kardon/hk570i.shtml

Whatever the recommendations my be I'm willing. If this isn't the best way to post a help request please let me know the correct way. It's not a mint unit but it does seem like it wants to live again. If anyone has a minute and could lend a hand I'd appreciate it.
Thanks much, Mark
570i full.jpg 570i close.jpg
 
Don't think you necessarily have a problem here. C8 is not connected in circuit quite the same as the other three power caps. The way C8 is connected in circuit may simply not bleed off the voltage.
 
Don't think you necessarily have a problem here. C8 is not connected in circuit quite the same as the other three power caps. The way C8 is connected in circuit may simply not bleed off the voltage.
Hi Glen, thanks for the input. I had thought that C8 not bleeding the voltage off might be a possibility but I've just never come across that before. Making things work is one thing but understanding why they work, or don't work, is truly another and it's the latter that I'm working on becoming better at. Thanks again. M
 
The amp has separate power supply to each power amp channel.
C8 is the main filter for the left channel, negative(42.4) rail.
C6 performs the same role on the right channel, generally expect
C6 and C8 to behave the same.

Maybe check zener D402 (failed open?) and also D14 as a long shot.
 
The amp has separate power supply to each power amp channel.
C8 is the main filter for the left channel, negative(42.4) rail. C6 performs the same role on the right channel, generally expect
C6 and C8 to behave the same. Maybe check zener D402 (failed open?) and also D14 as a long shot.
Thanks for the input mbz: My first suspect was a diode so I replaced:
D401-402 (RD12EB2) 12v zener 1/2W
D405-408 (ERB12-02) w/ 1N4002
D410 (1S2471) w/ 1N4148
D411 (1S2472) w/ 1N4148
D13-14 (1S2472) w/ 1N4148
The D9 (power led) was originally out though the dial lamps worked which is why D's 410/11/13/14 were replaced.
The power led came back and Q1 (2SA999-F) tested fine and measures proper voltages.
I also expected C6, C8 to behave the same as C7, C9 and checked resistance to ground from all output transistor leads.
Resistance measured from each output trans lead to ground has all emitters @ 23K ohm, all bases @ 6.5M ohm and
all collectors charge C6,7,8,9 and resistance rises to about 5.5 M ohm but (L) Q426 / C8 neg supply, and, Q428 / C9 pos supply,
hold their resistance due to C8 not discharging I assume.
(R) Q425 / C6 neg supply, Q427 / C7 pos supply, return to ~0 ohm within several minutes.
I've also checked C10, C11, R9, R467,468,469,470 all are within spec. The CB401, CB402 breakers measure 2.3 / 2.7 ohm across, in circuit.

This C8 holding not discharging is something I noticed early on when testing components and though voltages with the dbt are what I'd expect pursuant to the SM
this un-discharging C8 has remained consistent. I've checked ground lead connections, screws, etc. to no avail.
Thanks again mbz
M
 
Last edited:
Think you should wait for a tech with a better understanding of circuit theory.
I'm trying to work out why Q408, Q412, R428 and R434 don't bleed the rails voltage.
Their Vbe's will not be identical. Maybe in circuit measure R428 and R434 (100ohms)
 
I had a similar issue with my HK 730, dual power supplies so it was the 2 Caps in one channel that would not discharge.
There is no Relay, but there is a Slow Start Circuit, however with the Main Cap(s) fully charged, C405/6 (Slow Start Circuit in the 730) would also remain fully charged so there was no Slow Start and send 43V to the Speaker Terminals in that Channel.
After about 60 secs the amp would settle down to about 12mV and sounded as good as ever so please check before testing any speakers.

Long Story Short, no one was ever able to help me so I put Bleeder Resistors across all the Main Caps.
 
Think you should wait for a tech with a better understanding of circuit theory.
I'm trying to work out why Q408, Q412, R428 and R434 don't bleed the rails voltage.
Their Vbe's will not be identical. Maybe in circuit measure R428 and R434 (100ohms)
"mbz"
I agree on hoping for a tech with better circuit theory understanding having a look at this design. But I can imagine how difficult it could possibly be
to do this stuff remotely. Especially when someone has already started repairs. But everything else seems to be in order, voltage wise at least, except this cap issue. Generally once I get to this point I'd take if off of the dbt and see what I get for DC offset and Bias if applicable. But this issue makes me... nervous.
Both R428, 434 measure 100/101 ohm both in and out of circuit. In fact I've measured every resistor in the power supply and amp section.
I had also replaced C401/401 (100uF-10V) & C409, 410 (1uF/50V) electrolytic caps in the amp circuit and C4,5 -C12,13 in the ps section. There seem to be no other issues except this.
*It also says in the "Precautions" section at the beginning of the Sm "never attempt to do work on the transistor amplifiers without first disconnecting the AC line cord and waiting until the power supply filter capacitors have discharged". Reason for current concern, no?
I'm still thinking over here.
Thanks for looking into it mbz!
M
 
Last edited:
I had a similar issue with my HK 730, dual power supplies so it was the 2 Caps in one channel that would not discharge. There is no Relay, but there is a Slow Start Circuit, however with the Main Cap(s) fully charged, C405/6 (Slow Start Circuit in the 730) would also remain fully charged so there was no Slow Start and send 43V to the Speaker Terminals in that Channel.
After about 60 secs the amp would settle down to about 12mV and sounded as good as ever so please check before testing any speakers.
Long Story Short, no one was ever able to help me so I put Bleeder Resistors across all the Main Caps.
Hey "xrayspecx77" thanks for the info. I'm glad I'm not the only one wondering if this sort of occurrence is correct for these HK supplys or not.
I spoke with an older gentleman that I used to work for and he thought it sounded odd as well. I'll hang in there and see what comes of this inquiry before proceeding and certainly before modifying anything. But I do thank you for your input and your particular solution.
Boy I sure miss those "Description of circuit" offerings included in somewhat earlier stereo receiver SM's.
regards,
M
 
Last edited:
Certainly there are some amps where the main filters don't discharge. The sm states they should be manually discharged with
a resistor... This main caps in this HK appear they should discharge, it's just a matter for finding the expected bleed process.
With this fault present, I'd be against rushing to do anything, best to allow another 1-2 days for a solution to come forward.
That said, the fault is only in the discharge, setting bias and dc offset should be ok, certainly do not connect speakers as
per previous post (xrayspex77)
 
This main caps in this HK appear they should discharge, it's just a matter for finding the expected bleed process.
With this fault present, I'd be against rushing to do anything, best to allow another 1-2 days for a solution to come forward. That said, the fault is only in the discharge, setting bias and dc offset should be ok, certainly do not connect speakers as per previous post (xrayspex77)
Agreed, mbz. I've just attempted the idle current adjustment per the sm (w/ dbt)
R's 464,466 (0.22 ohm) emitter resistors for Q426,Q428 measures +0.041V. Rising to such in about 20 sec. I then stopped to check that the other channel wasn't inordinately high.
R's 463,465 emitter resistors for Q425,427 measures +0.02 and stays consistent over a similar time frame. I had already pulled both and measured them and both were right on at 0.22 ohm to the common lead. The (L ch) R464,466 & (R ch) R463,465 emitter resistors for the output transistors are of a dual design in a 3W package. Would replacing a three lead dual resistor with two 0.22 ohm 3 watt resistor, sharing the appropriate hole be O.K.? I believe I may have them on hand. It doesn't surprise me that this "bad" reading is with the other channel as opposed to the C8, (C9) pair I've been worried about.
A clue? Probably. : )
Thanks again,
M
 
Last edited:
I've also read the Service Bulletins available for this model (570i)
All relevant to this model and or serial number but this one caught my eye given the previous measurements.
SB: 10002 refers to DC offset and the possibility of several problematic resistors (R11,R12) in the "stabilized power supply" for the "control" and "phono" amplifiers.
The two resistors have been replaced with no change in the status of the right channel bias voltage. I'll take a look at R403, the bias trim pot as well. I had planed to upgrade them as well but ordered the wrong form factor from Mouser when I ordered the filter caps. Ugh!
 
Last edited:
Question: I pose this question here as opposed to the general "Solid State" area because it may be more specific to the problem I'm having than not.
The question, relative to having this cap that's not discharging for whatever reason, is whether or not leaving this unit running through a dbt for long a bad idea.
I want to re-take voltage readings on the many transistors in the pre/power amp section without turning the unit on and off every few (~15-30) seconds.
Which is what I'd been doing based on the cap observation. The dbt is 100W and does not glow at all at power up or thereafter.
And are there any concerns for possible damage one might have when taking measurements via the dbt? The particular unit's operating power "rating" is probably one of the only considerations I make when using the dbt. I haven't had any issues previously but this particular issue has me considering things I hadn't necessarily considered before. There is no information regarding wattage on the rear of this unit and none in the SM.
So I guess the real question is: can damage still occur even though current is being limited through a bulb with say a 50% to 100% increase of the units over all power consumption?
 
Somewhat puzzled why you want to replace the 0.22 ohm 3W emitter resistors.

Bias spec is 33mV. Left channel is at 41mV, right ch at 20mV, ok they need adjusting
but keep in mind you are still on DBT. Possibly the trimmers have become fragile making
it difficult to adjust. Bourn multi turn series (3296) are about USD2.50, need to get
the correct adjuster/pin orientation, maybe the 3296Y??? You can use separate 0.22/3W
resistors if you wish...

Wise to update amp with any applicable service bulletins. These are known "faults".
R11,12 are for the regulated supply, the C8(?) issue is for the unregulated supply
to the power amp.

Running on DBT will limit any current surge and give a good indication of that. It
also means that the amp is not exposed to full voltage so you should not set bias
while on DBT. However since the left channel is high (41mV), it will almost certainly
be higher off DBT, I would wind it back in line with the other channel to 20mV.

Failure of the cap to discharge is not causing any damage during normal power ON/operation.
The amp can remain on for extended periods. Based on the DBT, it appears safe to
disconnect the DBT. Certainly do not connect speakers and remove headphone before
power off and only insert minimum 5 seconds after power on.

I will try and have another look at the bleed sequence.
 
Somewhat puzzled why you want to replace the 0.22 ohm 3W emitter resistors.
Bias spec is 33mV. Left channel is at 41mV, right ch at 20mV,...
resistors if you wish...
Hi mbz,
Apologies, I mis-represented the measured voltage at dual R's 463,465 (emitter resistor) it should have been +0.002mV. I've reordered proper trim resistors.
The dual resistor replacement issue popped into my head when I first pulled them both and measured them but both measured slight above 0.25 ohm.
I was wondering about options if either or both were bad.

Wise to update amp with any applicable service bulletins. These are known "faults".
R11,12 are for the regulated supply, the C8(?) issue is for the unregulated supply to the power amp.
The SB (#1002) mentioned that "through the control amplifier a signal (DC) may enter the power amplifier..." should the two resistors in question fail or begin to do so. So I just replaced them to be sure they weren't a factor in the current circumstance.

Running on DBT will limit any current surge and give a good indication of that. It also means that the amp is not exposed to full voltage so you should not set bias while on DBT.
I was less trying to "set the bais" as opposed to checking existing offset and bias voltages and looking for obvious problems. It would seem I found one ; )
Note: while checking the many jumpers in the amp section I found that one was missing on the board but is shown on the SM. It connects C406, C408 midpoint to a circuit trace that leads to C424 and then to the ground plane. I replaced the jumper but there has been no change to the C8 discharge issue. I've run into no other SM vs. existing layout discrepancies so I assume that someone must have removed it previously. No idea why one would tho.

I'm still pondering this one and waiting for others to hopefully chime in as you mentioned. As soon as the new trim resistors arrive I'll be checking voltages again for the semi's in the pre/power amp section.
Thanks again mbz, I really do appreciate your help.
M
 
Back
Top Bottom