Hi,
i am restoring a vintage Harman Kardon HK6500 amp and i have encountered a bad ground loop situation. I am running just the amplifier board without any other boards since that only makes the problem worse. If i short the inputs, hum disappears/becomes *incredibly* quiet with ear on the speaker driver. With normal inputs (Magni 2 Uber pre-out) attached, its also quiet. If i have no inputs connected, it's quite a bit louder (maybe 6dB).
I changed all the important capacitors on the board.
Not sure what to do now. The amp has 2 pin plug (tried reversing it too) and uses center tap as ground. All of my other gear uses floating ground too. I saw some pictures on "ground lifter" circuit, which would connect the amplifier to mains earth, but i am not sure about that (i have good amount of SMPS in the same room running on the same live phase).
Does anyone have any experience solving such an issue?
What would you recommend to do?
Thanks!
i am restoring a vintage Harman Kardon HK6500 amp and i have encountered a bad ground loop situation. I am running just the amplifier board without any other boards since that only makes the problem worse. If i short the inputs, hum disappears/becomes *incredibly* quiet with ear on the speaker driver. With normal inputs (Magni 2 Uber pre-out) attached, its also quiet. If i have no inputs connected, it's quite a bit louder (maybe 6dB).
I changed all the important capacitors on the board.
Not sure what to do now. The amp has 2 pin plug (tried reversing it too) and uses center tap as ground. All of my other gear uses floating ground too. I saw some pictures on "ground lifter" circuit, which would connect the amplifier to mains earth, but i am not sure about that (i have good amount of SMPS in the same room running on the same live phase).
Does anyone have any experience solving such an issue?
What would you recommend to do?
Thanks!
Not a ground loop at all - more like electrostatic hum pickup, maybe from transformer E fields. Probably normal. If you have audio on high impedance (input) nodes running all over the place on simple ribbon cables, that is just bound to happen. As long as the hum is gone with an actual source attached, I would consider this a non-issue.
You can make sure that all the case parts have good electrical connection (paint does not qualify as such). Other than that, it gets tricky. You could implement some extra shielding for the connections in question (e.g. wrap problematic cable(s) in Cu foil and connect to ground - Al also works but connecting to it is a lot more difficult). If you want to be absolutely sure that such issues do not appear, you'd need input buffers on every single input.
BTW, be glad you didn't turn on the loudness. That would give you a real reason to complain about hum - it's a real princess on a pea type circuit with precious little inherent PSRR and still not as hum-free as it could be despite having a dedicated (if simple) post-regulator. The first reg stage also is just a simple EF + zener type circuit, so I guess total PSRR of both just isn't all the much to write home about. (C529/530 and C981/982 definitely need to be in good shape as well.)
You can make sure that all the case parts have good electrical connection (paint does not qualify as such). Other than that, it gets tricky. You could implement some extra shielding for the connections in question (e.g. wrap problematic cable(s) in Cu foil and connect to ground - Al also works but connecting to it is a lot more difficult). If you want to be absolutely sure that such issues do not appear, you'd need input buffers on every single input.
BTW, be glad you didn't turn on the loudness. That would give you a real reason to complain about hum - it's a real princess on a pea type circuit with precious little inherent PSRR and still not as hum-free as it could be despite having a dedicated (if simple) post-regulator. The first reg stage also is just a simple EF + zener type circuit, so I guess total PSRR of both just isn't all the much to write home about. (C529/530 and C981/982 definitely need to be in good shape as well.)
C529/530 and 981/982 are all changed for good quality japanese capacitors, but that is irrelevant now since i have all of those boards disconnected (incl. phono).
After i wrote this thread, i tried some more experiments and i think the EMI pickup might be a good theory. I have basic RCA cables soldered on the input of the main amplifier board. If i set the output to speakers and move them around, in one spot they hum more, in other they hum less, though if i completely unsolder them it still hums. After this i took my Brymen 235 and measured electric fields on high sensitivity. In that area, the meter shows 1-2 segments on electric field scale, which is quite a lot, going in different part of the room shows no fields.
I could try making a simple piece of aluminium shield that would separate the audio circuitry and transformer with mains power lines.
The transformer also audibly buzzes (probably slighly delaminated core) but i have not registered that through the speakers.
I am not an expert on this circuitry, but if you see C32/C31 on the schematic (220nF cap), this cap is not present on the actual amplifier, the solder pads are blanked out and the spot is empty. Could that have any effect on this? I have a Wima film cap i could put there.
After i wrote this thread, i tried some more experiments and i think the EMI pickup might be a good theory. I have basic RCA cables soldered on the input of the main amplifier board. If i set the output to speakers and move them around, in one spot they hum more, in other they hum less, though if i completely unsolder them it still hums. After this i took my Brymen 235 and measured electric fields on high sensitivity. In that area, the meter shows 1-2 segments on electric field scale, which is quite a lot, going in different part of the room shows no fields.
I could try making a simple piece of aluminium shield that would separate the audio circuitry and transformer with mains power lines.
The transformer also audibly buzzes (probably slighly delaminated core) but i have not registered that through the speakers.
I am not an expert on this circuitry, but if you see C32/C31 on the schematic (220nF cap), this cap is not present on the actual amplifier, the solder pads are blanked out and the spot is empty. Could that have any effect on this? I have a Wima film cap i could put there.
You have to find out where the noise is coming from - externally, internally or a cross- channel ground loop.
Disconnect all inputs.
Take a 2 meter RCA cable and plug one end into one channel and the other end into the other channel so that you have effectively shorted the two channels together through the cable. If you move the cable around, does it hum and if so does the level change with the position of the cable?
Disconnect all inputs.
Take a 2 meter RCA cable and plug one end into one channel and the other end into the other channel so that you have effectively shorted the two channels together through the cable. If you move the cable around, does it hum and if so does the level change with the position of the cable?
I tried shorting them like that, made no difference.
As someone above mentioned input buffer, i think that is my issue.
I have none. As i stripped the amp from all the boards to feed it directly, i removed the Rs and Cs on the signal lines. Because i intend to remove the tone control PCB and only keep the basscon, i will probably have to redo the input and output Rs and Cs.
Is there any equation to figure out the values or should i just mirror the old ones? The main input for the amplifier board runs from the tone control, on the signal line there is a 47uF elec cap, then 18K resistor to ground and 330ohm in series with signal line. The amplifier board has 330ohm resistor too, then 100uF cap and 22K resistor to ground.
Ok. If you have modded it then don’t try to do any tests.
Return the amp yo exactly as it was before you worked on it. HK made good stuff and it would not have been noisy in its original form.
Return the amp yo exactly as it was before you worked on it. HK made good stuff and it would not have been noisy in its original form.
It had incredible amount of noises before, the tone boards add hiss and even more grounding noises.
Hmmm. Then there is a problem somewhere. Excessive Hiss is often an indication is oscillation. Note you will get hiss if the inputs are open. Did it hiss when they were shorted out?
I did not try that, i'll have to reflow the other boards to assemble it all back again.
Basically, there was a hiss present on all inputs, it changed with volume control and was affected by tone control (more treble - sharper hiss) and there was also a hum also affected by the tone control (more bass - more hum).
If i disconnected the basscon board from power, most of the volume adjusted hiss stopped. If i disconnected the tone control board from power, all of the hiss was gone. When i disconnected only the phono board from power, there was suddenly more hum (other 2 boards were connected).The hiss was there regardless of inputs being hooked up to something or not.
In the config i have the amplifier now, there is no hiss, only that hum and it goes away after shorting inputs.
this is how it looks now
Attachments
Last edited:
You need to restore the amp to its original condition (with your new caps in of course).
When the inputs are shorted, or fed from a typical source like a CD player or tuner, the hiss and most likely the hum will reduce greatly to the point that it is not obtrusive. The reason the hiss goes up with the treble turned up is because you are amplifying the noise from earlier stages. Typicaly they will be using 47 k input bias resistors and when not shunted with the input source resistance (max a few hundred ohms in modern gear) you will get the hiss.
When the inputs are shorted, or fed from a typical source like a CD player or tuner, the hiss and most likely the hum will reduce greatly to the point that it is not obtrusive. The reason the hiss goes up with the treble turned up is because you are amplifying the noise from earlier stages. Typicaly they will be using 47 k input bias resistors and when not shunted with the input source resistance (max a few hundred ohms in modern gear) you will get the hiss.
Both hum and hiss are audible from listening position at higher volume settings, even with source connected. I had the amp in original condition with new caps and all that stuff. It hissed and hummed. Then i started stripping it down to reduce the hums and hisses. If oscillation of the tone and loudness circuits is indeed my problem, what would be my option to fix it? The main output has some 150pF caps in parallel with the speakers, i guess that is for noise from the main amplifier(??).
Here are some schematics so you can roughly see how it's laid out. Thanks for all the help, i didn't know it could be oscillation.
Attachments
You need to be very clear on when its humming and hissing because although is might be oscillating, its something I would only assume after making quite a few more tests - this is more the kind of thing to be expected from a DIY amp.
Get the amp put together back to its original state with everything wired in and make sure its working (ignore the hum and hiss for the time being)
Get some phono jacks and short them out. Plug them into the inputs and then see what it does.
Does it hum and hiss when the inputs are shorted out and with nothing else connected?
Make sure your amp is not standing on top of another piece of equipment and that any nearby appliances are turned off for the test.
Get the amp put together back to its original state with everything wired in and make sure its working (ignore the hum and hiss for the time being)
Get some phono jacks and short them out. Plug them into the inputs and then see what it does.
Does it hum and hiss when the inputs are shorted out and with nothing else connected?
Make sure your amp is not standing on top of another piece of equipment and that any nearby appliances are turned off for the test.
I'm leaving for holiday today, so i'll update the thread at the end of the week when i assemble it together.
When i tested it, i had quite a lot of equipment around, my pc, some switching supplies and my networking gear, so that might add up to some of the oscillation.
By the way, it's not very loud hiss and hum. Although it was audible from the listening position, it can be overpowered by music.
One more thing i noticed before i rigged the amp apart.
I was feeding it from my DAC (Modi 2 Uber) into the CD input. The dac has output 1.5Vrms and the amplifier has input sensitivity of 0.150mVrms. On a track that goes very near 0dB (digital - so the dac is pushing full 1.5Vrms), one of the tone boards starts crackling.
Since i had loudness disabled, it must be bass/treble control transistors. Could it be that they're old and don't like such input voltage? If i fed the same track directly to the amp board through pre-outs on my headphone amp to achieve roughly the same volume as before, it was completely clean and without any crackles.
My initial idea was that the hiss is caused by bad transistors. What would be your take on this?
Clipping but where? The speakers played pretty quietly when it was crackling, when i fed that exact same track directly to the amplifier board i could get many times louder without any problems. I would have to be clipping the tone control board.
Is that possible?
The bjts on TC are 2SC2320 (+complementary pnp), afaik they're quite sensitive low noise transistors.
Last edited:
Ok, so i did some more testing with the amplifier hooked up "ghetto style". It should not be causing any issues different to the factory config with all the boards, i should have a 1K resistor on the input signal lines but it will work even without it. If i move my soldered on leads, the hums gets more intense or quieter. If i touch the amplifier chassis, there is a slight pop in the speakers and the hum gets quieter. What if i connected my amplifier to mains earth? Could that solve my noise issues?
- Status
- Not open for further replies.