Can you post a photo of the input wiring, starting from the socket all the way to the grid?
Can you also try the same test on the cap with a shorted input?
Last edited:
I have also moved the high voltage secondary transformer off the pcb so it cant put hum into the circuit.
Also the high VAC lines are now on flying leads held away from the circuit.
Also the high VAC lines are now on flying leads held away from the circuit.
An externally hosted image should be here but it was not working when we last tested it.
Is the 100k pot a metal body type? If so, can you ground its bushing/body for shielding?
Are the transformer leads twisted? The trace to the input grid is very long and meandering.
It's best to have the input tube very close to the 100k pot to minimize the parasitics in its grid area.
Large loop areas will pick up noise easily, especially in high impedance circuits such as this.
If you use a rat's nest, move the parts around to make most of the flying leads as short as possible,
especially for the input stage, before routing.
Last edited:
I have covered everything it could be like ground loops.
The power supply wasn't optimal so I rewired it and it made no difference.
I cut the grid track from the valve to the grid stopper and put in a wire instead away from interference but that didn't help. However just putting my hand near the wire increased the hum.
There is no detectable hum on the B+ to the valves.
Star grounding didn't help but it was almost optimal as it was.
Looks like its going to be another pcb layout and minimise input circuit.
A copper pour around the valve might be sensible too, connected to ground.
I have previous valve mixer I designed and that had very little hum.
But the input circuit was as short as possible and it had a copper pour around the sensitive circuit connected to ground.
Keep the input trace short, direct, and over a solid ground plane (or at least right next
to a ground trace) for its entire length. This should fix the problem.
Keep the proper HV clearance from the plate pads and related traces to the ground plane.
Last edited:
I think I rushed this one too much and it wasn't as good as it could have been.
Just started on another layout.
Fixed the smoothing capacitors star grounding.
Kept mic input as short as possible.
Will add copper pours when its fully laid out.
Just started on another layout.
Fixed the smoothing capacitors star grounding.
Kept mic input as short as possible.
Will add copper pours when its fully laid out.
An externally hosted image should be here but it was not working when we last tested it.
I'd be really worried about laying out a new PCB without understanding why the original doesn't work. I'd spend a little more time investigating. Nowhere close to having exhausted the possible causes. Any badly placed trace can be cut and jumpered to test a hypothesis.
How are you shielding the input tube? In a pinch tin foil grounded via a few turns of wire could suffice.
Have you tried elevating the filaments above ground by say 25V or so?
With the grid directly grounded do you still get hum?
Many other things to try.
How are you shielding the input tube? In a pinch tin foil grounded via a few turns of wire could suffice.
Have you tried elevating the filaments above ground by say 25V or so?
With the grid directly grounded do you still get hum?
Many other things to try.
I have tried lots of things to improve matters but very little improves things.
The filament is already elevated to 40v.
With the grid grounded the hum goes down to about 40mV on the output.
I have designed a previous valve mixer with good results so I think this one is just poor pcb layout. I have used a star ground on the power supply.
The mic input valve has its components about 4 inches away so I am guessing the problem is just hum pick up from the surrounding area.
That's some atrocious grounding. Looks like you may have confused yourself with bipolar power supplies somewere along the lines.
Last edited:
I put a screened wire from the grid stopper to the valve grid and the noise remained the same.
I think I got to the bottom of my problem.
I had accidentally picked up a 12ax7 instead of a 12au7.
The 12ax7 has a gain of 100 which means my pre amp has a gain of 10,000 !
The 12au7 has a gain of 17 which gives a more sensible 289.
However there is still quite a bit of residual hum about 40mV on the output.
This means the first stage is getting a couple of millivolts hum into it from somewhere.
I think I got to the bottom of my problem.
I had accidentally picked up a 12ax7 instead of a 12au7.
The 12ax7 has a gain of 100 which means my pre amp has a gain of 10,000 !
The 12au7 has a gain of 17 which gives a more sensible 289.
However there is still quite a bit of residual hum about 40mV on the output.
This means the first stage is getting a couple of millivolts hum into it from somewhere.
I once drove myself completely nuts attempting to eradicate what appeared to be motorboating . It worked fine on the bench but when connected in-situ , there was a random quiet throbbing noise which at first I thought was caused by my phone . After ripping stuff apart adding extra decoupling and even dusting down my scope then poking around , it turned out to be something else . Turned out to be the plug-in wifi access point in the workshop . Moving it to another socket some distance away from the amp restored my sanity . You may find it may not be the amp but something else . Lamp dimmers can do nasty things for example . Have you ruled out such factors yet ?
316a
316a
I spotted a mistake tonight.
While I am working on the mixer I have flexi-lamp that hangs over the mixer to see what I am doing. If I turn it off hum goes down by 25% ! I can understand CFL lamps causing interference as they have oscillators in them but hadn't realised a normal incandescent bulb could interfere too. The interference is 50Hz so perhaps it should have been obvious.
I have been able to connect back up the cathode bypass capacitors now I am using a 12AU7.
I had hoped I could get away with a cheap plastic box but I need a metal box to act as screening.
While I am working on the mixer I have flexi-lamp that hangs over the mixer to see what I am doing. If I turn it off hum goes down by 25% ! I can understand CFL lamps causing interference as they have oscillators in them but hadn't realised a normal incandescent bulb could interfere too. The interference is 50Hz so perhaps it should have been obvious.
I have been able to connect back up the cathode bypass capacitors now I am using a 12AU7.
I had hoped I could get away with a cheap plastic box but I need a metal box to act as screening.
Last edited:
Another one to look at is if you have a pot on the input that has a metal shell . If the metal shell is left floating , that can also cause issues
316a
If I leave the input floating then touching the pot can causes loads of hum.
This is why I am going to need a metal enclosure and I can ground the enclosure.
I found removing the grid stopper and moving it right up next to the valve grid caused the hum to have much less harmonics in it. The hum is just about bearable now.
I stole the circuit from a previous 4 channel stereo mixer schematic that worked.
I made the mistake of changing the layout.
The grid side of the valves is very poorly laid out.
The tracks are long and the grid stoppers miles from the grid.
Just asking for trouble.
The pcb design was rushed hence the rookie mistakes.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.