• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Preamp Grounding and Hum

BRSHiFi

Previously known as kingden
2008-08-28 4:27 am
Evanston, IL
Hi all,

I am building a simple preamp circuit on a breadboard. This is mainly to optimize the circuit conveniently.

All the circuit is is a 12BH7 differential stage with a current source tying each cathode together. The output is picked off of the phase+ portion through a cap straight to the output jack. The power supply is just a pi filter with a resistor and two caps (22uf each).

I am using a DC filament supply with a 7812 coming off of a separate transformer winding.

The preamp grounds are:

- CT of B+ winding.
- Main filter cap.
- Filter cap after resistor.
- Input RCA jack.
- Input resistor (no volume added yet).
- LM334 ground and current set resistor.
- Output RCA jack.

The filament supply grounds:

- CT of filament winding.
- Filter cap.
- Ground of 7812.
- One end of filament.

I have been messing around by putting the grounds in different places but I always have hum and buzz. What is the best way to ground a small signal system to stop the humming? I have long mastered this exercise on power amplifiers but preamps seem to be a different story. I know the circuit is on a breadboard but even with that it still should be possible to make it work.
 
Hi kingneb, a couple of thoughts.

First, does the amplifier hum all by itself with nothing connected to input(s) and headphones connected to the output?

Second, does the system hum when you connect the preamp to your power amp (if you do), still with nothing connected to the input(s)?

It's possible you have a ground loop when the preamp is connected to another component of your system - if so, the solution will be different than if the hum is present in the preamp all by itself.

Also, can you post a schematic of your design?

Best luck. :cool:

~ Sam
 

BRSHiFi

Previously known as kingden
2008-08-28 4:27 am
Evanston, IL
My headphones are 32 ohms. Nevertheless I hooked them up and a faint buzz came through when I hot switched the input. The residual hum level with my headphones cannot really be determined as the audio is too soft due to the low impedance.

It hums (buzzes more than hums) when a power amp is connected, with an input or without an input, and even shorted. The level changes with a short, no source, or a source connected.

I took out the power supply and substituted a an old 1960 Heathkit 6L6 regulated bench supply. Still hums/buzzes.

The schematic is attached.
 

Attachments

  • Untitled.png
    Untitled.png
    44.1 KB · Views: 570
OK. Not easy to be sure from the picture, but you seem to have avoided the usual problem of poor CT and reservoir cap grounding.

You might have some induction. Try reducing the size of circuit loops. I realise this is not always easy with breadboards. However, you could put the rectifier diodes closer together so you have a smaller loop from the transformer secondary.
 
Hi kingneb - OK, for some reason I though you were taking the output off the cathode, my bad. Yes, the headphones will load the plate and the audio will be quiet.

Still, the experiment seems to indicate that the hum is in your pre-amp if you can hear any hum (even at a low level) with the preamp input disconnected and output just going to the headphones.

The 47k/22uF B+ filter should do a pretty good job of power supply filtering, so I wouldn't suspect the B+ filter for now.

One thing I would do is to float the filament supply from your common ground. The power transformer might be injecting B+ rectifier pulses into the filaments by connecting the 12.6VCT to your amplifier common ground.

Since the filaments are not part of the signal path in your circuit, you can just float the whole filament circuit from your common ground.

Nice breadboard, BTW - I love those prototype boards, they certainly make for easy circuit changes!

Best luck. :cool:

~ Sam

EDIT: Looked at your photo again, it looks like you may have already separated the filament ground from the amplifier common... Gotta get me some new glasses :)

Still thinking about the problem. I also see where you've disconnected the B+ rectifier from your protoboard and it looks like the alligator clips are going off to your external bench supply (?)

2nd EDIT: The longish-leads from your 7812 regulator might need some extra bypassing. Something between 1uF-10uF right at teh 7812 output before the twisted-pair filament wires. The longish-leads can lead to less effective regulation in the 7812, might cause some noise. Just a thought.

3rd EDIT: You might also try using just a fixed resistor cathode bias on the LTP. See if the LM334Z is contributing to the problem.
 
Last edited:
Kingneb, are you sure you have >12VDC at the input to the 7812? You need 14.8VDC minimum at the LM7812 input to give you 12VDC at the output.

The transformer wires (green with green/yellow CT) look like a 6.3VAC CT winding - which probably won't give you enough DC volts at the regulator input to produce a 12VDC regulated output.

But sometimes color codes aren't the best way to determine transformer voltage...
 

BRSHiFi

Previously known as kingden
2008-08-28 4:27 am
Evanston, IL
Just be clear, I am at present using the Heathkit bench power supply for B+. I restored it a few years ago, all new parts in it. As I said before, it has no affect on the buzz.

I did try to float the heaters on the filament supply. This makes the buzz extremely horrible.

The winding on the transformer is 24 volts so 12 volts regulation is not the issue.

Also, there is a non-zero potential between my preamp input ground and my PC case. I got a pretty noticeable tickle from this potential difference. Also, this non-zero potential causes AM 1400 to be picked up to varying degrees.

Different grounding experiments change AM 1400's audibility but it becomes clearly audible when an input is added that grounds to my PC, whether that be the sound card or my USB function generator.
 
Just be clear, I am at present using the Heathkit bench power supply for B+. I restored it a few years ago, all new parts in it. As I said before, it has no affect on the buzz.

I did try to float the heaters on the filament supply. This makes the buzz extremely horrible.

The winding on the transformer is 24 volts so 12 volts regulation is not the issue.

Also, there is a non-zero potential between my preamp input ground and my PC case. I got a pretty noticeable tickle from this potential difference. Also, this non-zero potential causes AM 1400 to be picked up to varying degrees.

Different grounding experiments change AM 1400's audibility but it becomes clearly audible when an input is added that grounds to my PC, whether that be the sound card or my USB function generator.

OK, very good.

Well, if you're getting bitten by the preamp input ground then it seems pretty clear you have a ground-loop problem with the system (source + preamp).

I'm still unclear (to me, at least) whether you have the hum (using earphones) with the input completely disconnected and floating, and whether your hum appears only when you hook up your source to the preamp input.

From the photo, the "suicide plug" doesn't have safety ground connected - so I wouldn't think that is a path (for the ground loop). When your Griefkit ( :) - I have one, too...) HV supply is used, it may or may not have a safety ground-plug (the original design had 2-wire plug, possibly you upgraded to a 3-wire cord when you serviced the unit?)

Perhaps using the back-to-back diode trick to connect your preamp ground to the safety (earth) ground on the plug might do the trick. :cool:

~ Sam
 

BRSHiFi

Previously known as kingden
2008-08-28 4:27 am
Evanston, IL
Heathkit uses original 2 wire power cord. Power amplifier, earth ground not hooked up, filament transformer, 2 prong.

PC, three prong.

The unit buzzes pretty bad with no input (floating). The buzz drops when a source is connected. The suicide plug reduces the buzz the most. AM1400 audible on all conditions.

I have a scope. The buzz not obviously detectable on it. It is clearly a ground loop because the buzz disappears immediately when preamp power is killed.
 
Last edited:
Heathkit uses original 2 wire power cord. Power amplifier, earth ground not hooked up, filament transformer, 2 prong.

PC, three prong.

The unit buzzes pretty bad with no input (floating). The buzz drops when a source is connected. The suicide plug reduces the buzz the most. AM1400 audible on all conditions.

I have a scope. The buzz not obviously detectable on it. It is clearly a ground loop because the buzz disappears immediately when preamp power is killed.

A 100 ohm grid stopper might kill the AM1400. If you have a strong transmitter in the vicinity, there might be enough pickup to get into the first triode and get rectified someplace - all it takes is something that looks like a diode to the radio signal.

I think I would also get the protoboard layout common back to earth-ground somehow - probably through a couple of paralleled back-to-back diodes. Bypass the diodes with a 0.01uF cap so the RF signal sees a path to earth.

The whole thing floating is probably causing both your tingle and the AM radio reception.
 

BRSHiFi

Previously known as kingden
2008-08-28 4:27 am
Evanston, IL
Those diodes are a huge game changer.

The buzz is gone. Here is the situation now:

- I have a thermionic hiss with an intermittent paper tearing sound. Tapping the tube reveals pretty bad microphonics. Both 12BH7s I have on hand are that way.

- A floating and shorted input yields a faint buzz type hum (the type of hum from a turntable ground not being connected).

- With a source connected, I have an audible virgin 60Hz hum. It is louder than the buzz sound but still on the faint side.
 
Last edited:

BRSHiFi

Previously known as kingden
2008-08-28 4:27 am
Evanston, IL
I did some more tinkering around. I added an opamp buffer on the output to drive headphones and changed the tube to a 7025 and added a feedback network. gain is about 12dB. Attached is the new schematic.

With headphones there is hiss (not as bad as 12BH7) with faint hum. The hum stops immediately when power is removed but the hiss lingers until the tube cools down and the power supply caps discharge.

Apparently there is some ground loop stuff still going on. As in the previous post, the diodes to earth really helped. Now how should we ground to eliminate the humming?

I still am using the Heathkit. I will hook up the original power supply as soon as we can eliminate the hum (and hopefully hiss) with the existing configuration.
 

Attachments

  • Untitled.png
    Untitled.png
    48.6 KB · Views: 264
Excellent news, kingneb!

You might try bypassing the cathode with a small cap - 0.01uF or so. That silicon constant-current sink might be injecting noise into the cathode, which will stay hot (and emitting) for a while after the filament supply turns off (and until the B+ drops).

I'm assuming the filament supply will drain the cap quickly - although you might also try bypassing the 7812 with a 0.1uF cap as well (could be injecting noise there, too). :cool:

~ Sam

EDIT: You've also got some feedback on the output grid - did you try leaving the output grid grounded and just going with the unity-gain op-amp buffer in the output?
 
Last edited: