5F2A with 2 - 6V6's - HUMMM

[rockybottom16]

I just finished an experiment of sorts - a 5F2A running 2 - 6V6's in parallel. The amp sounds wonderful, but it has a pretty significant hum. Most of the amps I've built over the years have been push-pull. So I know that SE amps can be a little hummy, but I think this is excessive. The volume control has no affect on it. In fact I seem to remember the hum is there even with V1 pulled. So I'm thinking either a ground issue or maybe I need to move the OT around a little to see if that helps. But now that I think about it, the hum doesn't show up until the tubes warm up, so maybe the OT position is OK. I'm using the brass plate with everything except the Mains ground going to it. Filament CT goes to 6v6 cathode. Now I'm thinking I should chuck the brass plate and do a buss across the pot backs for the preamp stuff to the input jack ground, and ground the power amp and power supply stuff over by the PT. Incidentally, the OT I'm using is a Classictone 18031. It is currently mounted on the opposite end of the chassis from the PT. I used the 5K primary and have a 4 ohm speaker load hooked to the 8 ohm secondary tap.

I know grounding issues have been beaten to death, but I'd still appreciate any advice on chasing down this issue.

[mozz]

Is it 60hz hum or 120hz? Do both output tubes have the same filament wiring, 2 to 2, 7 to 7?

[eschertron]

Schematic?
The ripple in an SE design does not cancel the way it does in PP. You may want to look at the PSU component values. If your design copies the values from the 5F2a for a single 6V6, I'd wager that it's not filtered enough for 2x 6V6. In other words, it won't be twice as hummy as a 5F2a, but much more.

[rockybottom16]

mozz - I'm not sure if it's 60 or 120hz. Yes, 2 to 2 and 7 to 7, but I'm not sure why that is important since this is not a push pull arrangement.

eschertron - Yeah - The main filter caps are 16-16-10-10, which is pretty close to stock 5F2A values. Should I bump it up to 22-22-10-10? I think I have some 22's on hand.

[Dave H]

The main filter caps are 16-16-10-10, which is pretty close to stock 5F2A values. Should I bump it up to 22-22-10-10? I think I have some 22's on hand.

As it isn't affected by the volume control it's probably SE power supply hum. Try tacking two 22u caps across the two 16u filter caps to see if that helps or you could try a choke between the 16u caps like the 5F2.

[rockybottom16]

OK. I tacked a 22uf cap across each of the two 16's and also one across the first 10uf (screen). That seemed to reduce the hum somewhat. It still seems louder than it should be, but then again it's been a while since I spent much time with an SE amp. I have some questions:

1) Since the hum does not appear immediately upon turn on, can I assume that my OT location is not the issue? Hum doesn't start until tubes warm up.

2) Since the volume control has no effect on the hum, can I assume my grounding is OK?

3) The rectifier is a USA-made 5Y3GT. Is all this extra filtering bad for the rectifier tube?

I tried to add an image of the 5F2A schematic, but had no luck. It shows 2 - 16uf caps in parallel for the first filter section for a total of 32uf (OT). With my added capacitance there is now what amounts to 2 - 38uf caps in parallel for a total of 76uf. That seems a bit much.

Thanks!

[Chuck H]

I think I see where mozz is going with this...

What if you connected the filament wiring for one of the power tubes inverse? Shouldn't that cancel any filament hum in the power tubes relative to the matching of the tubes?

76uf is too much for a 5y3. You can use a Pi filter after the rectifier. Something like a 22uf right after the rectifier followed by a 100 ohm resistor and then a 100uf for your main filter to the power tube plates.

Can we assume you are using a bypass cap across the 6v6 cathode resistor?

I would expect some hum from ripple or filaments in your case, but excessive hum could be a ground loop? Make double sure you don't have anything grounded in two places. You should reexamine the tube pull test. What happens to the hum if you pull the second preamp tube? Do the panel controls affect the hum? That sort of thing.

[rockybottom16]

Hi Chuck - I was snooping around the web some more and found some folks that were dealing with the 5f2a (Hoffman's site). There I found what you called a "Pi" filter. I just finished trying that: 30uf attached to the rectifier to ground, a 10 watt 120 ohm cement resistor connecting the rectifier to the standard 5f2a set of caps (16-16-10-10, where the 16's are in parallel). The hum was reduced quite a bit, and now I'm not concerned about the 5y3.

And yes, there is a 20uf/50v cathode bypass cap on the 6v6s. This is bascally a tweed princeton build, but with 2 - 6V6's running in parallel and a 250 ohm cathode resistor. And of course a bigger SE OT.
There is only one preamp tube. Pulling it has no effect on the hum.

I think I'm gonna leave her alone for now. What hum is left is not really objectionable, and the amp sounds wonderful. Just like a Tweed Princeton, only more of it!

Thanks folks for your input.

[mozz]

I wasn't actually thinking whether parallel filament wires or swapped would make any difference, i was thinking push pull rule of thumb(old habit), but you now have me wondering if reversing them would cancel anything. It should be easy enough to measure the hum difference with any true rms meter across the speaker load.

Also, you can try the trick of taking the output transformer bolts out while running, rotate the trans and see if any hum changes.

[tubeswell]

Yes reversing the heater pins in a parallel SE output stage can provide some phase cancellation in the signal, but its not much, because the gain of the output tubes is relatively low anyway, so the S:N ratio from stuff like heater noise is not as significant as it would be in the 1st pre-amp gain stage. Having the heaters hum-balanced through a heater winding CT with a humdinger or a 2-resistor artificial ground reference will be more effective in minimising heater hum. A bit of elevation can also help with pre-amp hum.

I think it is worth tackling the ground returns*, lead dress, RF noise suppression (with grid stoppers and input/grid lead shielding), and any rectifier diode switching transient noise suppression (in addition to searching for heater hum). Also a dud tube can cause significant hum. Filtering the high tension supply with a HT choke is also a good idea with SE amps

*See Merlin Blencowe's grounding article

[Justin Thomas]

Pics of the actual build layout would help too...

Just an experience I had a few years back... Rebuilding a trashed Gibson Thor (2EL34, 12AX7, EF86). Sounded great, but loud hum I couldn't shake. It was there in the original build, and after I turned it ibto my own design. I actually gave up on it for a year because I couldn't shake this one last hum - definitely hum, not buzz. Came up as the tubes warmed up. I COULD drown it out if I played LOUD. I followed EVERY grounding rule & layout trick I could possibly find. I swapped ground points everywhere. I tried everything.

One day I was about to throw up my hands & was exceptionally lazy, and I was trying a mod & had to run another ground wire. I left it extra long and loose, hanging outside of the chassis. Because who cares if it's noisy? The @$#%ing amp is already incorrigibly noisy... When I fired it up, the hum was about 90% gone. I'll be danged, that wire lying in a route halfway around the chassis, going by the PT, killed the hum! The fricking PT was throwing off such a field that it made the amp hum. By permanently routing that one extra ground wire halfway around the chassis and at a length five times greater than "necessary" according to convention, it bucked the LT hum, as best as I can figure out.


Normally we suggest rotating the tranny in a situation like this, but that's not always an option...
Maybe you just have a crappy extra-noisy PT... Or your layout is less than ideal...

Justin

[Gingertube]

Also remember that the screen supply is far more critical than the anode supply so make sure it has additional filtering.
Cheers,
Ian

[alexradium]

I would try elevating the heater center tap.
One trick that was used in the old single ended radios,was the output transformer,that had the primary with a tap around 90%,to which the plate went to,and the screen went to the top,so the two currents were out of phase and cancel out the hum,.

[booj]

The schematic shows one side of the filament voltage going to ground. this will cause some 60HZ hum. Placement of the output transformer can make a difference, but its the same as the original chassis it shouldn't. The input signal wire is going to pick up noise. If routed the same it should be ok.

[loudthud]

The schematic shows one side of the filament voltage going to ground. this will cause some 60HZ hum.

Nobody would use the 50's style wiring from the original schematic. Rockybottom16 stated that the CT of the heater supply was connected to the output tubes cathodes.

Placement of the output transformer can make a difference, but its the same as the original chassis it shouldn't.

Rockybottom16 states that the hum doesn't occur until the tubes warm up. That eliminates OT placement as the source.

The input signal wire is going to pick up noise. If routed the same it should be ok.

Rockybottom16 states that the Volume pot does not effect the hum. That means that the input signal wire is not picking up hum.