HUMMMMMM

[lowell]

Well, I'm back to messing around with my preamp. I've built two now, and both have quite a bit of hum. I managed, using alligator clips, to move some grounding around and get it to an acceptable LIVE gig level of hum. I run it into a LM3886 chip amp that I installed in my 1x12 cab. However I would like to make them quiet enough to record with. And would also like to build some mic pres as well.

It's basically a Fender 5F1 Champ circuit, with a 5k Ra for the 6v6 and 1uf output cap. I have the output attenuated to ~2v at max.

I am using an Antek 05t350 torroidal transformer, and it's inside the chassis, very close to the boards and components. I know that this is most likely my main source of hum. Curious if anyone has any insight on this. Just how "quiet" are torroidal transformers?

My supply is a full wave rectifier - 100uf - 100ohm - 30uf (6v6)- 10k - 47uf (12ax7). There's no center tap on the secondary, should I use a bridge? Will that help with ripple hum?

I am aware that single-ended doesn't cancel psu hum, thus the larger filter caps.

The 6v6 is in triode mode, 5k Ra, and 5k Rscreen. 500 Rk. It's biased cold so the plate doesn't sit at 1/2HT, which I found sounds better anyways.

I've tried more filtering, moving grounds around etc... Even tried a DC filament supply, didn't help AT ALL. I initially built it with star grounding. Each triode of the 12ax7 was grounded with it's related filter, grid leak, cathode etc.. and same goes for the 6v6. I then strung the grounds together: res cap to IEC ground lug, 6v6 related components to res cap neg lead, 2nd preamp stage to 6v6 star, 1st preamp stage to it's star then to 2nd stage star etc.. I know this is confusing without a drawing. But I basically star grounded each stage, then ran that star to the following stages star, with the resevoir cap star being the last one and connects to the IEC earth lug.

I found that I had to move the 1st stage star ground to the speaker output jack for the least hum.

Where should the filament virtual ground go? I connected it to the 1st stage filter as it didn't seem to matter where it was.

I do not have the input jack isolated from the chassis, as it didn't make a difference when I tried it. I didn't try the output jack, because the input jack didn't make a difference, but maybe I should. Maybe they BOTH need to be for it to make any difference. Anyway I spent many hours moving things around and got it a little better. Just looking for ANY input on the subject y'all may have.

When I pull the 12ax7 MOST of the hum is gone. If I ground the 2nd 12ax7 triode's grid the hum goes away too. So it is in the first stage, and I'm guessing is straight EMI from the transformer.

[Zozobra]

Can't really comment much more than saying that the last few things I've built have used torroidal transformers (Antek and toroidy.pl) and they have been unsettlingly quiet to the point that on first live run into a cab I thought something was wrong when I switched it on. I recall reading somewhere (maybe here) that someone had hum issues with a SE build and in the end they fixed it with a choke.

[trobbins]

Nothing should go to the IEC mains earth connection point to the chassis.

To avoid confusion, just choose and use (and confirm by test), that only one link goes from circuit 0V to chassis.

What happens if you use a short direct link from 12AX7 input triode grid terminal to that triode's cathode circuit 0V?

Do you use a 12AX7 tube socket that has a central spigot for use as a 0V / screen point?

Hum may be ingressing the cathode or plate circuit layout of the first stage, or the coupling to the 2nd stage grid. Tests such as above should identify if inherent circuit layout can provide low hum.

[lowell]

I am considering a choke... but trying to avoid using one for reasons of real estate and cost.

trobbins... thanks for the input.

I did notice that the tube socket, which is mounted on my turret board, not the chassis, and is not grounded, does make a clicking noise of I touch a grounded clip to it. Hum doesn't change, but do hear a "click."

No central "spigot". What is this for? Curious, I've seen em, but wasn't aware they provided anything other than a convenient ground.

Ok, so connect chassis to earth ground with only one wire. All other 0v commons go to chassis?

[trobbins]

Many good threads and articles on grounding - chase a few up with google. Chassis to 0V circuit through only one link. Mains AC protective earth to chassis through only one link.

Socket spigot is mainly for hum and parasitic feedback screening. Valve datasheets like EF86 often include a comment on using a base with spigot for best hum reduction.

But only testing with specific shorting will indicate where the hum is mainly being coupled in, and then some remedial changes can be made to manage the coupling.

[Enzo]

Isolate the problem, where is the hum coming from? Is it 60Hz or 120Hz?

You say the PT secondary has no CT and you have full wave rectification, but then you suggest a bridge. How can you have FW rectification on a non-center tapped winding without a bridge? if you mean you have four diodes instead of a square thing, well the square thing is just four diodes, so that won't change anything.

Do ANY of the controls affect the hum in ANY way?

Pull all the tubes except the power tube, still hum?

Pull the power tube and monitor the 12AX7 with a scope or signal tracer, is the hum still there?

have you actually read the ripple levels on the B+ nodes?

There are a million things that cause hum, and each is separate. You can try changing a million things, but none will help until you do the one thing to solve the particular hum you have. So find it first, THEN due the cures.


Here is the schematic of the 5F1:
http://bmamps.com/Schematics/fender/..._Schematic.pdf

You did not provide a schematic of what you built. You DESCRIBED the power supply as three caps and two resistors, with a cryptic "6V6" at one point. The 6V6 has TWO B+ connections, one from the first node through the OT, and the second from the second node. Larger caps may help a little, but this single ended circuit with the OT from the first filter is going to hum. To eliminate hum you would need to take the power tube B+ from the second node, or add an extra RC stage before that node.

COmpare the 5E1, which by design ought to have less hum from B+ ripple.
http://bmamps.com/Schematics/fender/..._Schematic.pdf

You don't need the extra filter to be a choke, just a resistor there will work too.

[lowell]

You say the PT secondary has no CT and you have full wave rectification, but then you suggest a bridge. How can you have FW rectification on a non-center tapped winding without a bridge? if you mean you have four diodes instead of a square thing, well the square thing is just four diodes, so that won't change anything.

Maybe this is why?? I have a 2-diode full-wave rectifier. Both diode cathodes to the reservoir cap + terminal...- terminal is grounded. There is no ground connection in regards to the rectifier or PT secondary. I'm thinking I need to review rectification...and that maybe I need a bridge with the ground reference?

[Malcolm Irving]

Maybe this is why?? I have a 2-diode full-wave rectifier. Both diode cathodes to the reservoir cap + terminal...- terminal is grounded. There is no ground connection in regards to the rectifier or PT secondary. I'm thinking I need to review rectification...and that maybe I need a bridge with the ground reference?

Enzo is right. For full wave rectification you either need a CT and two diodes or four diodes and no CT. Can you take another look - the way you describe it no energy would be supplied to the reservoir cap.

[mozz]

Steel or aluminum chassis?

[lowell]

Wow, crazy. Here's the transformer - http://cdn2.bigcommerce.com/server60...0.1280.jpg?c=2

I have it wired exactly as I described. Aluminum chassis.

Also, I'm using ONE primary, should I use both in parallel?

[Justin Thomas]

Look at some old mid 60s Ampeg gut shots -they make a neat bridge out of four bundled diodes. Rather than buy another part, use 2 extra diodes if you got them laying around.

Justin

[loudthud]

Also, I'm using ONE primary, should I use both in parallel?

You should be using both. Do you have a virtual center tap on the heater windings?

[lowell]

I have some bridges laying around.

So begs the question....how the hell is this thing working? Other than the hum...the signal rocks! I have plate voltages etc...

[lowell]

Ok thanks. Yes I have a virtual CT for the heaters. That was my first go-to for the hum. I added it and the hum improved greatly. Before that, the hum was insane.

[Enzo]

That is cause you have multiple hum sources. As you cure each one, you are still left with the remaining ones.