Screen Grid Connected to Plate ? Caused Big Hum.

[HaroldBrooks]

I read that if you connect the screen grid to the plate, it reduces the power of your output pentode, and changes the harmonic character to act and sound more like a triode (amp uses a single 6L6 power tube).

Tried this last night on my single ended 6L6 amp, and while it indeed dropped power and sounded very different and useful, it also created a very pronounced Hum.

I'm not sure if I am doing something wrong, I disconnected the screen grid resistor to remove it's static positive voltage, and soldered a short wire lead between plates 3 and 4 (plate and screen) of the 6L6.

Any thoughts as to what might be wrong ? I would like to make this a switchable feature in this amp, but the Hum is just too much at this point. Maybe a resistor ? Or a Cap ?

As usual, Thanks for any help !

[Old Tele man]

Try inserting a 100 ohm resistor between the screen and plate and verify the hum isn't coming through a bad/weak filter capacitor.

[HaroldBrooks]

Thank Old Tele man.

Would a 2 watt resistor do ? It's a low watt amp, so I'm (only) guessing it would be ok.

[Old Tele man]

2W is fine...plate should be catching 90-95% of total (ie: cathode) current flow, so unless something else is wrong, screen current should be roughly 5-10% or less than plate current.

[eschertron]

How big was the filter node that you bypassed for the change? On a SE amp, power supply hum is always a problem and the screen node may have been filtered more than just a bit? Can you post a schematic of at least the power supply filter section? A quick sketch would be enough.

[pdf64]

What were the plate, screen grid and cathode Vdc before and after the change?
It may just be that the screen grid voltage was increased significantly, resulting in the tube drawing more current, causing more hum.

[HaroldBrooks]

This was my first amp to start "Modding" and I've changed a few things just to see what improvements I could make...

That being said, I also dropped the size of the two resistors between the filter caps a bit (trying to boost the plate voltage to the preamp tube). In other words I was just looking for trouble back then !

I think I might have dropped the resistors to half there original value so I may have upset the later stage filter including the screen grid, and the can type Mallory filter cap is original, so that might also be another issue now.

I am using a 6L6 instead of the 6V6 seen on the schematic, as I think the transformer has an extra capacity in amps in the filament feed that allows me to do this. I looked up the power transformer specs and it seems so, and have been using the amp with the 6L6 (like the sound better) for over 1 year, without any overheating of the transformer (about 118 deg F with an infrared gauge) after running flat out for more than an hour.

So I will see if adding some more filter capacitance helps. Here's the original schematic : https://www.google.com/search?q=6150...xiIMsxfhVBrJM:

REVISION : Ok, I am fairly sure I increased the hum when I dropped the 47k resistor at the filter caps all the way down to 15k. I got away with it when the screen grid was independent, but the hum got much larger when I tied it to the plate. I used two 47uf caps running in parallel (94uf) and that made the amp dead quite. I think the original can cap 10uf leg has seen better days, so I will put in a new filter cap (perhaps 20uf), regardless if I make this change or not.

The amp does sound different and "nice" with the 6L6 screen grid attached to the plate, but for now I will just leave it stock as a pentode, and add that new filter capacitor. I think I learned my lesson regarding mucking with the filter cap resistors too much. It was a good experiment and I might add a switch down the road if I think it's something I want to use consistently.

You don't know until you try !

Thanks !!!

[J M Fahey]

Single ended amplifiers have poor/nonexistent supply ripple rejection.

Not sinewave mains induced hum but sawtooth ripple if you scope it.

Plate supply is usually poorly filtered, screen supply often has an extra RC filter but if you disconnect it from that relatively cleaner place and connect it to hummier plate supply ... even if through the OT primary ... instant hum injection.

[HaroldBrooks]

I've got to get a bit better at record keeping when I make changes (used to do that), but I believe the plate was 282 vdc and the screen was 265vdc or there abouts'. Not sure what it changed to when I connected them.

Adding a huge amount of capacitance killed the hum virtually 100%, but somehow I must have screwed with the filter cap circuit enough that this was required to kill the hum. Such is the case from doing mods with little theory under my belt.

I just added resistance back to the 47k (now 15k) between the last two filter cap nodes and that helped, and I also replace the last 10uf cap with a 47uf one, and that also helped quite a bit.

So I might get back to adding the switch to change the pentode to a triode, as soon as I order a suitable switch. I can also switch in additional resistance back to the last stages of the filter cap if need be.

Actually I am glad this happened, as I learned more the purpose of the AC filtering at different stages. Up to this point I have only been considering proper grounding and the level of capacitance, but there are the resistors that add a critical element...

I have to take a course in electronics !

Thanks pdf64 for the input.

[HaroldBrooks]

Thanks for your comments J M Fahey. I've had more trouble getting the sound I like from my single ended amps, for sure and hum seems like more of an issue.

The push pull pairs also seem to have a smoother tonal structure for the most part. Here's the paradox though, sometimes you don't want smooth ! That's why I love all my amps !

There's only one amp I've neglected for quite some time now that uses the old Radio tubes. Sounds ok through it's tiny speaker, but once you start to drive it up front, it falls apart sonically.
It's a Gregory 007 with a super tiny 4" radio pm speaker. It was my first tube amp after I returned to playing after a long interlude.

I put some grid stopper resistors on the preamp and power tubes and that helped a bit, but it seems like that whole structure is out of it's league for any high gain rock playing, so it mostly collects dust.

[J M Fahey]

Oh, I LOVE single ended amps.
Way more complex crunch/dirt in a good way.
Sadly not strong enough for stage use but you can always pick that wonky but tasty signal and reamplify with some dumb but powerful SS amp, and the the sky is the limit.

[Old Tele man]

...or, parallel another output tube with the original SE tube, but remember input drive (due to lower Z.in) will change.

[tubeswell]

here's some options for SE filtering (with pentode mode for comparison)

As others have pointed out, the problem is likely to be insufficient filtering, with lack of ripple rejection in the power rail due to tying g2 to the plate without considering the power supply filtering. So you can either add a 'whole of supply' filter between the rectifier and the plate supply node (CLC is best to eliminate ripple), or try brute force B+ filtering by upping the B+ capacitor (but if there is no smoothing cap in the filter, it will be more difficult to eliminate hum in triode mode). Other suggestions include: organising your ground returns, elevating the heater, and grid resistor snubbing to reduce RF/oscillation). You could also try shielding signal grid wires (with short cable runs)

[HaroldBrooks]

All good ! Thanks for being so comprehensive, I got an education with that post for sure ! As in many things in life, it's never usually one thing causing a bad effect.

I added a big 47uf filter cap to the screen grid supply node and that helped quite a bit in reducing the hum. For now, I reverted the power tube to a pentode. At some time in the future I will add a switchable change from Pentode to triode, perhaps in one of my other amps as well and I now will consider that plate ripple may be of concern when connecting the screen grid. I have to also see about elevating the heater(s) voltage to the cathode, as I don't think this amp takes advantage of that.

Thanks for the fantastic schematic on power supply filters, and for Your help !

[Iron Works]

I had a Traynor Guitar Mate YGM-1 from the 1967 and the dropping resistor between plates and screens was 12 ohms as per the schematic. There was a bad hum that many had tried to resolve in the past. Ultimately a 470 ohm dropping resistor cured the hum. I assume Traynor made many of these that didn't hum so I'm not certain what total configuration of components made this one different but all dropping resistors were in spec and dropping as per the math. Later versions of the YGM had larger resistors isolating the screens from the plates.