Reduced V1 Hum - New Trick (For Me that is)

[HaroldBrooks]

I made some changes to my old Supro Supreme (circa 1949) over the last year or so to "Hot Rod" it so I could play Hard Rock but still have a Chicago Blues amp as a basis. All the things I did made the amp into a screamer, particularly when I use a Parametric EQ / Boost pushing the front end...

However, in raising the gain and plate voltages, and picking the highest output tubes I could find, I raised the noise floor and hum a bit as well. The hum is louder and lower when I turned the volume pot up and down, so I reasoned it must be something before that pot causing the hum.

I was OK with the level of hum in the old hot-rodded Supro for a while, until I bought another Valco National amp that was nearly silent hum wise, even when dimed - ( A very Similar later model still using a field coil speaker and same tubes).

So now the hum of the Supro Supreme at an idle started to get on my nerves. I re-soldered all lf the grounds and made sure ALL the caps are new, and even put a 220uf cathode bypass cap in for the 6J7 tube. Some of this helped, but only a tiny bit.

I was just about to give up when I noticed a Cap on the schematic (Please see below) that came of the 6J7 Pentode Screen grid and went to ground. The value was .02uf, so I figured it may be in my best interest to make that a bit larger and see what happens. I took out the .02uf cap and installed two .047 caps in parallel (.094uf), and now the hum has been cut virtually in half, and is very acceptable. I've put the amp through the passes, and everything seems fine, but again the hum is much lower at idle ! I think the purpose of this cap is to dump any AC on the grid, but I don't quite know how to calculate the value for that cap, other than using a high pass filter calculation including the 1meg grid voltage resistor, but not sure if that is correct or not.

Any comments / insights are very welcome ! Thanks in advance for your help, I would like to understand what the heck I am doing !


https://el34world.com/charts/Schemat...supro_1600.pdf

[nickb]

That suggests to me that the hum was on the supply to the 6J7. Go back to 0.02uf and put a say 47uF in parallel with the 10uF on that supply node as see what happens.

[pdf64]

That stock value should be good but maybe the original 0.02uF cap is bad? Have you tried a known good one in there?

[HaroldBrooks]

I did replace the original .02uf cap early on. I also took part of nickb's advice and added two additional 10uf caps to the 6J7 DC power and 6SC7 DC lines, and that helped a tiny bit, but the biggest thing was adding the extra capacitance to the cap coming off of the screen DC feed. Thanks for your help.

[pdf64]

Ok fair enough.
Fyi merlin advises that the screen is best bypassed to the cathode, rather than to 0V (not sure why) see http://www.valvewizard.co.uk/pentode.html

[nickb]

...I think the purpose of this cap is to dump any AC on the grid, but I don't quite know how to calculate the value for that cap, other than using a high pass filter calculation including the 1meg grid voltage resistor, but not sure if that is correct or not.

Any comments / insights are very welcome ! Thanks in advance for your help, I would like to understand what the heck I am doing !


https://el34world.com/charts/Schemat...supro_1600.pdf

Yes, the cap does bypass unwanted AC to ground. In order to maximize the gain you want to hold the screen grid steady else the varying voltage on it caused by the screen grid current varying with the signal gives negative feedback and reduces the gain. The easy way to approximate this is to bypass the AC current to ground using a capacitor. So long as it does it over the range of frequencies of interest, it's good enough.

So, yes, the formula for the corner frequency is 1/2/3.14/R/C. The problem is R is hard to know since it consists of the screen grid resistor in parallel with the tube's internal screen grid resistance. I don't know what that value for a 6J7 is but I do know that for an EF86 it's around 84K so let's just go with that by way of example. So R is (1000 * 84)/(1000 + 84) K = 77K and with a 0.02uF that makes the frequency 103Hz. I can't tell you why they chose that as it seems a bit higher that I would have picked but it's not unreasonable.

Now we get to your hum. The screen grid acts very much like grid 1. You apply a varying voltage and it varies the plate current. It's not nearly as effective as grid 1 but it still has some gain. Thus if you were to inject hum on to the screen grid it will appear on the output. In this case the hum comes from the plate supply (with the 10uF cap) and that is why adding capacitance helped. The choice of 103Hz for the corner frequency also means that hum at 120Hz gets through the screen grid to the output without too much attenuation. So finally we get there, increasing the screen grid cap lowers the corner frequency and reduces that hum on the screen grid.

[trobbins]

Quite often with vintage amps, the power transformer HT secondary CT is taken directly to chassis, rather than more appropriately taking the CT directly to the first filter cap negative terminal (which is likely to be a combo electrolytic can base). The other high current connections of concern would be the output stage common cathode 0V point connection to chassis, and whether the field choke returns directly to the next main filter cap (rather than to the output transformer CT). How is your amp wired?

The other typical noise/hum issue with the first stage is that it may use various chassis ground connections, including at the input jacks. Noise and hum from the various chassis connections may not be easy to counteract unless you want to reconfigure all the chassis 0V connections.

[Tony Bones]

The hum is louder and lower when I turned the volume pot up and down, ...

https://el34world.com/charts/Schemat...supro_1600.pdf

If you mean that the hum is low with the volume at min and max, but the hum is loud with the volume at mid position, then that suggests that the hum is being picked up "through the air" by the circuit that connects the wiper of the volume pot to the grid of the second tube. Well, at least some of the hum; there are probably many sources. I don't know what the inside of the amp looks like, but pay close attention to that wire and what it passes near on its way from the volume pot to the tube socket.

[Tony Bones]

Fyi merlin advises that the screen is best bypassed to the cathode, rather than to 0V (not sure why) see http://www.valvewizard.co.uk/pentode.html

It's because the AC current path is from the cathode to the screen and then back to the cathode. If there's a cathode resistor, then connecting the cap to ground includes the resistor in the loop. Even if there's a bypass cap on the cathode resistor, you still end up with that cap unnecessarily in the loop. Probably not very important in most cases but, connecting directly to the cathode is the shortest path and includes the fewest opportunities for unwanted interactions.

EDIT: Sometimes in high gain amps a small bypass cap is added across a plate resistor to reduce high frequency fizz. Dumble installed that cap from plate to ground instead of plate to B+. That eliminated the PS smoothing cap from the AC loop. If he had been just a little more clever he would have connected the cap from plate to cathode and eliminated the cathode bypass cap too.

[HaroldBrooks]

The hum level followed the volume pot in a perfect linear fashion, but thanks just the same for adding that behavior with regards to the hum being loudest in the mid position.

That happens on one of my guitars. I back off the volume a little and the hum increases, but then starts to decrease as the volume is turned down more, and I always thought that quite peculiar.

Moving wires around in a vintage amp is my specialty , as it takes very little knowledge, but sometimes yields very good results ! I purchased a National 1210 about a year ago, and finally diagnosed the buzzing issue as a capacitor that was connected to the rectifier as a ground, but was practically laying on-top of one of the rectifier HV plate pins. Once I bent the cap out of the way, the hum went down 75%. Boy was that one easy !

Please keep any suggestions coming, you guys are helping me out a lot, as this is my new hobby and it's great to find a board like this one with so many talented practitioners all in one spot.

[HaroldBrooks]

I may connect mine to the cathode to see if it makes more of a difference, once again thanks pdf67 and Tony Bones.

[Gingertube]

Not applicable according to your post just above

BUT for future reference:

Hum or noise being max at mid setting on a volume pot or guitar volume control is a big clue. Actually with a log pot it will be a bit off center position (when resistance of top and bottom arms of the pot are equal).
This is where the SOURCE Impedance to drive the following stage is highest.
As Tony says, this is when pickup from capacitive or magnetic coupling from adjacent AC circuits (like heater runs) is worst.
In extreme cases it can also be a hint that the following stage is going into parasitic oscillation and a grid stop resistor is needed.

Cheers,
Ian

[antrecords]

I went rounds on all the same battles mentioned above, on the same amp (47). HUMMMMM. Raising the grid capacitance did not help, nor did adding capacitance to the B+ node really. It helped a little moving grounds: the filter caps (new) all to one point, at the power tube cathode resistor's 0v. The hum was still not right. I spiraled V1 filament wires for good measure, toyed with different shunting caps, shielding, moving lead dress. Nothing seemed to totally work.

"I was just about to give up when"

The single thing that did: The inputs were grounded to a dual lug strip on the front panel. I don't know if this was stock or not, I assume it is. I was really about to give up but moved both input grounds to the preamp's common ground (cathode R/C). Like magic the hum was greatly, greatly reduced. The single best cure in this instance.

Hope someone finds this useful.

Cheers.

[HaroldBrooks]

I will find it usefull ! You are correct, the dual lug strip is stock, and I ran another ground wire from the same location, but left the other groudings attached, and that reduced the hum before I added the bigger 6J7 grid cap.

Now I will have to see if running the actual ground wires to the common ground directly will help further, but to be honest at this point the hum in my particular amp is already pretty low, considering how much I jacked the gain up from stock.

Great you stuck with it and went even further. The early Valcos are great sounding amps, aren't they ?
0
My favorite Valcos are the old field coil amps, because of the raw but simultaneously warm and smooth distortion they mostly have, and big response changes to using an EQ up front. If you haven't already done this, run an EQ at the last stage in your rig into your amp, and dump the lowest of lows (below 80hz or so), and bump the mids in a couple of ranges of your choice. The amp comes alive and sounds very different than the stock sound. I use a Boss PQ-3b because it has a huge boost and cut range, and is very quite. The best amp "Mod" out there is running and EQ up front IHMO, then you can dip into the chassis for further changes as needed.

Good luck with your amp, and thanks for the extra info regarding hum reduction.

[antrecords]

Thanks for chiming in. I'll be curious if moving your input grounds has a good effect, let us know if you do. I like your EQ bump idea, should yield some great tones. Interestingly I have a Gibson BR6 of the same era here, extremely similar amp but different sound vs. the Valco. Same speaker/coil, gotta love the response.