Harman/Kardon HK3470 noise on the output entering the speakers

Since i'm new to the forum Hello to everybody:wave2: !
I have an old Harman/Kardon hk3470 with a 100Hz noise (yes i live in Europe) on the output with no music, just the speakers attached when it's daytime. When it's night time I don't hear it as much. I don't know what's causing it, but i'm sure after 20 years it should be recapped on the input stage at least. Also there is a 36mVDC at the left and a 6.2mVDC at the right channel, on the output when measuring (no load) with a multimeter (Voltcraft). Should i buy a new amplifier or can this be repaired somehow? When I bought it second hand 6 years ago, i had to do a little work on it since it didn't want to stay on, it constantly switched off. The problem was with a small capacitor hanging off the PCB, i also attached a bigger heat sink to one of the voltage regulators because it was burnin' hot, and also burned the PCB a little over the years. The weakest point i think is the transformer, because it becomes hot to the touch(>60℃ or 140℉) when it's on. Also here is the service manual attached if you wanted to look at it.
 

Attachments

Welcome to the forum!

Your problems don't seem severe and the amp seems likely fixable, unless the transformer is unduly hot due to internal failure of the transformer itself. How about heatsinks--- are they unreasonably hot?

About the hum--- is it "clean" undistorted hum or more "buzz" with rich harmonics? Is the hum present with volume knob at 0, or only with elevated volume setting? Same in both channels?

How stable is line voltage? 220VAC 50Hz, I assume. You might measure AC voltage in daytime when hum is present, and again in evening when behavior is better. Same behavior with all sources? Inputs shorted during evaluation?

Please mention any clues that might be relevant.

Good luck!
 
The buzzing sound is most likely mains distorted sine wave of the 50Hz and it's upper harmonics (around 100Hz). The amplifier is used with an external DAC connected directly to it's main input, thus bypassing the pre-amp completely, but the humming noise is present, even when the input RCA connectors are removed, and only the speakers are connected. The buzzing noise is even present when pressing the mute button on the front. The only way it stays quiet at the speakers is when the speaker button to the corresponding speakers is pressed, then the relay disconnects the speakers from the amplifier completely.
The line voltage isn't stable at all. At daytime the voltage can reach 250VAC and at night it drops down to 220VAC, remember that the nominal voltage here should be 230VAC +-10%. Note: the buzzing is present at night too but it's less noticeable.
 
So in your setup, the amp's volume control isn't used? I assume problem is the same if you use internal preamp with volume control set to 0?

Would you advise what DCV you observe on the amp's supply rails? Then would you switch to AC volts and measure AC ripple on the same rails? Add a series blocking cap if there's any doubt about how the meter handles presence of DC during AC measurement.
 
With the internal preamp connected there is a buzzing sound added to the humming noise even when the volume is at 0, this is because of a bad pre-amp design i think. Can you help locate the supply rails? I don't feel comfortable poking around with my multimeter inside a working amplifier. I have uploaded the PCB pictures from the pdf i uploaded earlier, the schematic is on page 113, and 116. I'm not an expert at amplifier reverse engineering, but i think the two B- and the B+ are the rails that go into the amplifier from the main board right?
 

Attachments

  • hk3470 amp pcb.png
    hk3470 amp pcb.png
    272.6 KB · Views: 79
  • hk3470 main board pcb.png
    hk3470 main board pcb.png
    452.3 KB · Views: 86
But wouldn't that bypass the capacitors? I should measure them with an LCR meter, but their capacitance is too high to measure them on the PCB, so i have to desolder them sometimes in the future. I should also consider buying a protection, because i have an expensive pair of DALI speakers with this old amplifier, a Guardian-86 maybe a good option i think. As for the rails i think i can manage to get my multimeter's test leads into the connector.
 
I was suggesting measuring ripple voltage present across the big electrolytic caps. From that we can draw rough assessment of health of the caps. Knowing DCV on the caps is good for perspective.

Speaker protection is always a good idea.
 
Last edited:
  • Like
Reactions: Bobberlite
I've measured the VDC on the connector itself (CN107) on the main board and the B+ to GND and B- to GND were the same 58.4VDC, maybe close to 60VDC, i also measured a 0.002VAC at the same place. The measurements were all measured using a APPA305 TRMS multimeter. Should i get a measurement before the smoothing capacitors?
 
Maybe my speaker and RCA cables are bad, and act like an antenna. When I connect the pre-amp to the amplifier i get a higher noise on the tweeters (violet noise), but the same overall humming of 50Hz and the upper harmonics, that is why i use my DAC as a pre-amp to the amplifier main input.
 
That's very low ripple on the supply rails, maybe even small enough to make me doubt the measurement. Try a 0.1uf cap in series with the meter and repeat AC measurement. If it's the same, I'm more trusting.

You didn't say if hum is still present with volume controls at 0. A very important clue.
 
The hum is present with vol at 0, all my measurements are at vol 0, with and without (there is no volume level just the DAC's gain i think) pre-amp. Without anything attached to the amplifier (and the jumpers removed from pre-amp out main in) there is a buzzing noise on top of the 50Hz humming on the speakers because i think the input leads aren't shorted and it amplifies the surrounding rf signals. With the jumpers attached to the pre-out -> main input there is a greater buzzing sound due to bad pre-amp design. If i connect my DAC to the main input directly, the buzzing sound immediately goes away, but the humming (50Hz) remains the same. Note: bypassing the pre-amp results in no volume control, it amplifies the input whatever it gets.
 
I don't think bad amp design is your problem- HK amps are designed quite well and do not buzz when new.
The main HK problem are failing parts, capacitors are not the best, often a bit on the cheap side.
HK does not care if the unit fails after a couple of years.
The problem is made even worse, as the mains voltage today is higher than 20 years ago.

If you want to repair the amp, you should change all electrolitc capacitors. To be honest, in such a complicated receiver I would not do it.
You have to adjust quiescent current and offset for the higher rail voltage too. This will make it run less hot.

The HK was a nice part, but no high end. Get a separate amp like the AIYIMA A07, add a good power supply and enjoy the improved sound. Cost will be about the same, but result is better and for sure. DIYS repairs of such a large amp can always go wrong.
 
This model was made for 230VAC 50Hz, so it was designed for this "new" mains voltage. I think it's easier to look for the bad capacitor than to buy a completely new system, just because of a handful of bad components. I should get my oscilloscope out and measure the ripple voltage on the rail. Also i measured with no load of course and let the amp reach temp equilibrium the VDC again after 3 years and the 2 channels measure the same 34mVDC offset.
 
Turbowatch2 comments are well taken.

An aside: My own prejudice is reluctance to do wholesale replacement of caps unless there's unequivocal evidence of multiple failures. My wife does crosswords--- DIY troubleshooting is my idea of mental fun. But you do the heavy lifting, so judge my advice accordingly.

Your amp presents interesting symptoms. Bringing your scope into the picture seems wise.

You did some prior repairs. Was everything good until recently? The first post mentioned a hot transformer and a hot regulator. But bias current history sounds encouraging. Any thoughts?

It's easy for me to get lost in the maze of schematics, but it appears to me that some of the low voltage power supplies have fuses present, but the power amp rails do not. If this is correct, it's very surprising. At the same time, it presents the notion of removing fuses and keeping the power amps live. Might change the hum picture and would also allow assessment of heat issues. Divide and conquer strategy.

Comments?
 
If you have an osciloscope finding the defective part should be much easier. You should be able to trace the hum to it's origin, my bet is on a bad capacitor.

I have been frustrated by a handfull of HK amps in the past, after finding the primary fault instandly the next problem got in view.
I restore and repair any kind of technical opject in reach, but sometimes time and possible result do not match up.
I found some HK amps to run quite hot, so not only one cap but a handfull of them
will be off spec. The tuner usualy starts to play up too, so when you finally fixed the amp the next thing needs attention.