Grounding vintage amplifier to mains

Old vintage 1976(?) Harman Kardon 900+.

Just got done doing a very comprehensive restoration - all the caps, nearly every transistor, etc. Plays wonderfully, but still has a bit of annoying mains hum that comes through the speakers. Not very loud - don't even notice it when music is playing - but it is considerably louder than just about any other vintage gear I have owned.

I can power this thing using my bench supply and it is dead-quiet. I been looking around and it seems these old HK receivers were prone to humming noise.

One guy - who seemed to be a former HK repair tech in the 70's and 80's had mentioned some of this was eliminated by re-routing a ground wire that runs by the tone board - but he has seemingly been offline since before the pandemic (gulp).

Another person (maybe it was here) somewhere had mentioned also having a 900+ that always had a bit of hum when the bottom cover is installed. I do notice the hum slightly more when the cover is on mine, too.

But the real reason for this discussion is: I notice that there is a 2.2M ohm resistor that connects one side of the AC power cord directly to the chassis ground.

For those who do not know... on this particular model, ground is ground is ground. Signal ground, power ground, chassis ground... is all one ground. Also seem to be a couple potential areas for ground loops from what I see and understand of the schematics.

Since the ac plug on this stereo is non-polarized... it seems to reason that the "ground" on this unit could be connected to either side of the AC line (nuetral or hot).

Would there be any harm in removing this 2.2M resistor bridge? If not, you think it is likely to help any of the hum feedback I am getting?

If I unplug this unit while powered on, the humming stops instantly. Keep in mind: it will continue to play music for 3-4 seconds after unplugged due to the large PSU caps. So the hum is certainly coming from the AC line.

Another thought: would it be of any advantage to install a grounded plug on thus unit? Or is connecting chassis ground to earth ground not a good idea?

-Dean
 
One other thing I notice is: the dial indicator lamps are all operated on AC voltage. These wires run through the chassis - next to signal wires. Even through these wires are shielded, my OCD makes me feel like having this low voltage AC could be a source of noise.
 
I would use a 3 wire cord and remove the resistor. Also remove anything else that is attached from the AC primary wires to the chassis.
Attach the Safety Ground/Protective Earthe to the chassis near where the AC cord enters the chassis.
Connect the circuit common and DC supply common to the chassis near the signal input jacks.
 
Putting it on the bench now.

There are really only 2 areas the noise can be coming from: inverter board or tone board.

When the clamp board is active (which grounds the amps inputs to earth), the hum is virtually non-existent. Once the system comes up, the clamp board removes the short (allowing audio signal). The only two active components in the signal path are the tone board and inverter.

I do know that: when the amps are disconnected from the inverter board, the hum is 100% gone. The amps themselves are quiet as can be.

I am going to take a closer look at the grounding paths and see if there is some grounding loop or something routed wrong. This thing had been repaired likely a few dozen times before I ever got it.

As mentioned - there was a member on a couple other forums who once worked at a HK service center. He mentioned on a thread somewhere that there was a service bulletin about something similar... here is his original quote (from Feb 2019)

"It can be eliminated with an easy modification by re-routing a ground wire near the tone board to a common ground point at the power supply. I have no documentation on it anymore and now I'd have to be looking at the unit to describe exactly what must be done but I don't have a 900+ handy right now"

I am going to review things and - if needed - re-route a ground wire or two. See if that solves the issue.

I am also going to remove this 2.2m resistor - leave the chassis/ground floating for now.

I will report back.
 
my rule of thumb, each positive or B+ supply wire mist have a negative ground returns wire, that the pair is twisted is better...
signal ground returns go to psu common in a separate wire and must never ride on the psu ground return wire...
if you follow this, then you should not have ground loops...
this is why balanced signals are superior....
 
So I lifted the resistor - hard to tell how much this may have helped.

I also got to poking around and found an extra (hidden) ground wire from tone board to the power amp common chassis ground point. This is in addition to the ground wire it already has running to the tone PSU board (as specified in the manual). Looks like something someone added at some point for some unknown reason - certainly not factory. But it was well hidden in the wire mess.

Helped a little.

I also re-routed the ground wire from the aux PSU so that it is directly to the chassis/common ground point. It originally routed THROUGH the inverter board. The idea: more "star-like" in the routing strategy.

Maybe that helped a little.. again, hard to tell w/ the poor quality speakers on my workbench.

But I noticed something else here... I been listening to this w/ the Bose 901's (I know, I know...). I dug out the Dahlquist DQ-10's and - while still present, the hum sound is about 1/4 as bad. I think it is just the sensitivity of the 901's in the mid-range. Combine that w/ the fact they are sitting in front of a solid brick wall and.. well..

But with normal speakers, the hum is at typical "old receiver design" levels.

It is NOT the 901 EQ, either. Already took that out of the loop. And the hum is present with volume turned all the way down.

Sounds like most any 1970's receiver with the Dahlquists. Go figure.

Strange
 
Last edited:
You could always try to AC couple the common point (on the power supply board) to the chassis.... no ill side-effect re buzz/ground loops - only the benefits from sinking the noise to the chassis potential. Of course, the chassis must be grounded to the mains ground.

Use the combination of 0.1 and 0.01uF caps. You could use the disk-ceramic caps - as short as possible should the leads be for the thing to actually work at high frequencies.

Good luck.