SE 6V6 Filtering

[DG1911]

New here. I have a SE 6v6 amp I've been working on. It's kind of a mutt, loosely based on an old Valco. At any rate, it used a 6SL7 in the preamp, 6V6 and 5Y3. 2 different cathode resistors are selectable for bias. Beats the hell out of the 6V6 and part of the "charm" is in the tube cutoff. At any rate, AC filaments and it hums like hell. Tried elevating the heaters. Didn't help much. Humdinger didn't do much. Don't want to do DC filaments, PT secondard cant handle it and I don't want to replace it or use a separate supply. Here's the thing though, I was working on lowering the plate voltage, and skipped using diodes on the filament CT. Instead, I added another filter stage ahead of the filter section for the plates. I read about trying this on another forum with hum reduction also being a possible benefit. The additional stage (5uf and 500R) dropped my plate voltage about 30 volts. Played with bias some afterward and decided I don't like the result, tone wise but ... it killed most of the hum!!! That's the benefit I'd like to keep.

So, I am wondering if I can keep that added filter stage but drop the resistor value to something really small to get the voltage back up but keep the benefit of the hum reduction. Would there be any issue with just replacing the 500 ohm resistor with something lower ... 8 ohms, 1 ohm, something of real low value that I have on hand.

I appreciate any input in this. In laymans terms if possible.

Thanks kindly.

[Enzo]

Try a choke instead of the resistor.


You already discovered it, but keep in mind that there are MANY sources of hum in an amp, and each one is different. You have to solve each source independently. In other words all the filter caps in the world will do nothing to fix grounding problems, and elevating heaters or putting them on DC will do absolutely zero for power supply ripple hum. So hum is not generic.

SIngle ended amps naturally hum, for the reason you discovered. They usually take the B+ for the power tube from the first filter. So you add a filter stage. Fender has a service bulletin from years back, telling service centers not to try to"fix" the hum in Champ amps, it was normal. It would not be covered under warranty.

Look at Fender Champ drawings. The 5C1 has the power tube drawing from the second cap, like your solution. (is that where you got 500 ohms as a value?) The 5E1 uses a choke there instead, but still draws off that second node. Then the 5F1 does what everyone else does, takes the power tube supply off the first cap. Now the amp hums, but they save the cost of the extra cap and the choke or resistor.

The best approach would have been to find the source of the hum and just solve that, rather than guessing at several solutions for possible causes.

[DG1911]

Try a choke instead of the resistor.


You already discovered it, but keep in mind that there are MANY sources of hum in an amp, and each one is different. You have to solve each source independently. In other words all the filter caps in the world will do nothing to fix grounding problems, and elevating heaters or putting them on DC will do absolutely zero for power supply ripple hum. So hum is not generic.

SIngle ended amps naturally hum, for the reason you discovered. They usually take the B+ for the power tube from the first filter. So you add a filter stage. Fender has a service bulletin from years back, telling service centers not to try to"fix" the hum in Champ amps, it was normal. It would not be covered under warranty.

Look at Fender Champ drawings. The 5C1 has the power tube drawing from the second cap, like your solution. (is that where you got 500 ohms as a value?) The 5E1 uses a choke there instead, but still draws off that second node. Then the 5F1 does what everyone else does, takes the power tube supply off the first cap. Now the amp hums, but they save the cost of the extra cap and the choke or resistor.

The best approach would have been to find the source of the hum and just solve that, rather than guessing at several solutions for possible causes.

Very small chassis. I don't have the space for a choke. But the extra filter section wasn't primarily to solve the hum. I wanted the voltage drop and opted for adding the filter stage as a shot in the dark that it would also reduce the hum, which it did. The 500 ohm resistor was based on math and what values I had on hand in high enough wattage (It's actually 510 ohm I think). This amp has switchable cathode resistors. (Well, it did. Right now I'm using the switch to switch in and out a cathode bypass cap). In one position it runs the 6V6 at 12 watts dissipation, but the plate voltage is up around 420 volts. In the other it's at 21 watts (60ma+) !!!, plate voltage is lower at around 380 watts.

At any rate, I appreciate the input but still wondering whether I can keep that extra filter stage but just replace the 500 ohm dropper with a very small value resistor. Just to keep the hum reduction benefit of the additional stage without dropping any/much voltage. I lost some of the nasty "charm" of the amp by dropping the voltage but I really appreciate the hum reduction.

[Leo_Gnardo]

Very small chassis. I don't have the space for a choke. - - - -

still wondering whether I can keep that extra filter stage but just replace the 500 ohm dropper with a very small value resistor. Just to keep the hum reduction benefit of the additional stage without dropping any/much voltage. I lost some of the nasty "charm" of the amp by dropping the voltage but I really appreciate the hum reduction.

The lower the value of your resistor the less efficient your hum filtering will be, hence the charm of using a choke. I've done it on several Champs used for recording & it works a charm, employing a choke sold for use in a Fender Twin. You could fit it in the cab somewhere, doesn't have to be on the chassis, as long as you adequately insulate the hi voltage wires running to it.

The layman's terms you're looking for (I'll try), the choke is an inductor, simply a coil of wire wrapped around a core of steel plates which help improve its efficiency. It has minimum resistance at very low frequencies, the lowest of which is zero, aka D.C. - direct current. So it will pass - most - of your high voltage while reducint it only a small fraction. That's what we'd like to have, right? OK so far. At higher frequencies, the choke acts as a resistor to current trying to pass thru it at those frequencies, in this case 120 Hz - cycles per second - the hum from your raw power supply. That's why it makes a good filter component, it stops a lot of the hum, but passes most of the power. Hope this helps.

[DG1911]

The lower the value of your resistor the less efficient your hum filtering will be, hence the charm of using a choke. I've done it on several Champs used for recording & it works a charm, employing a choke sold for use in a Fender Twin. You could fit it in the cab somewhere, doesn't have to be on the chassis, as long as you adequately insulate the hi voltage wires running to it.

The layman's terms you're looking for (I'll try), the choke is an inductor, simply a coil of wire wrapped around a core of steel plates which help improve its efficiency. It has minimum resistance at very low frequencies, the lowest of which is zero, aka D.C. - direct current. So it will pass - most - of your high voltage while reducint it only a small fraction. That's what we'd like to have, right? OK so far. At higher frequencies, the choke acts as a resistor to current trying to pass thru it at those frequencies, in this case 120 Hz - cycles per second - the hum from your raw power supply. That's why it makes a good filter component, it stops a lot of the hum, but passes most of the power. Hope this helps.

Yes, that helps. Thanks. It's not a combo so no cabinet to fit it into. It's a cage style amp. So a 90ma, 4hy choke. I'll check the chassis for real estate. Height isn't a problem but finding a space for the 3" X 2" footprint isn't going to be easy. I may be making a decision on whether the reduction of the hum is worth the change in tone. Probably not. If I cant fit the choke, I may have to see how low I can go on the R and see if I can find a sweet spot.

Thanks

[Tom Phillips]

Another approach you can consider to recover your lost voltage is to change the 5Y3 rectifier to a 5AR4 or even convert to SS rectifiers. With your class A power amp, which has a pretty steady current draw over the full range of output power, you are not using tube rectifier induced power sag as part of your amp's response characteristic. This will allow you to stick with your extra filter node with the added 500 Ohm resistor. I have used as much as a 1k Ohm resistor to get Chanps into a hum free sweet spot.

[printer2]

I played around with a Champ type amp a long time ago and found I got the most bang for my buck increasing the capacitance on the screen. This assumes you have a dropping resistor between your screen and plate.

[DG1911]

Yes, there is a dropping resistor between the plate and screen.

And I should have been more clear. The 6SL7 has it's own node. So picture the supply for a 5F1. I stuck the additional filter in front of it all (between rectifier and plate). Maybe a no-no but it did greatly reduce the hum.

Got some more questions coming. Y'all have been great. Thanks.

[Enzo]

The lower the resistor value, the closer the caps come to just being in parallel. SHort across the resistor, putting the two caps you have in parallel. Is that less humy than the original single cap was?

[eschertron]

And if you go with SS rectification, you can make the filter caps as big as you want. Change the 20u or 40u reservoir cap for 80u or 100u and that'd help any other filter circuitry out too. That's basically the flip side of Enzo's experiment: with a bigger reservoir and no extra resistor, how much does it help? If you still want more, then put another RC stage to clean it up more. Be advised - watch the size of the first cap if you are still playing with a tube rectifier in place.

[Jazz P Bass]

Yup!

[DG1911]

The lower the resistor value, the closer the caps come to just being in parallel. SHort across the resistor, putting the two caps you have in parallel. Is that less humy than the original single cap was?

I'd have to try it but 10uf is max cap value for 5y3, I believe. I see where you are going about possibly just upping the cap values. I have tried that for the 6sl7. It doesn't seem to make much difference. haven't tried it on the screen. But if adding that node dropped the hum, more capacitance will/should help. But, could it also have been a result of the lower dissipation when I dropped the voltages about 30 volts? Without changing the cathode bias resistor, that dropped dissipation to 18 watts from 21 watts. In fact, hum is much lower when switching to the "cleaner" mode which is just a higher value cathode resistor.
That puts dissipation at 12 watts although plate voltage jumps up to close to 420v.

Don't really want to go to SS rectification. But, It turns out that I probably can squeeze a choke up top. Possible problem is that I can't be too picky about where I place it. So the question would be whether there are any rules as to orientation and positioning of a choke in relation to the PT or OT. The choke would have to sit fairly close to the PT with the brackets front to back on the chassis, so no options in orienting the laminations. Are chokes sensitive to that like the positioning between OT's and PT's?
If not, I'll spend the 20 bucks to try it.

Thanks

[Tom Phillips]

I'd have to try it but 10uf is max cap value for 5y3, I believe....

It is 20 uF but that only applies to the first node. If you use your 510 Ohm resistor in between the new first node and the second node then the second cap can be larger. Even 220 uF or more.

But if adding that node dropped the hum, more capacitance will/should help. But, could it also have been a result of the lower dissipation when I dropped the voltages about 30 volts? Without changing the cathode bias resistor, that dropped dissipation to 18 watts from 21 watts. In fact, hum is much lower when switching to the "cleaner" mode which is just a higher value cathode resistor.
That puts dissipation at 12 watts although plate voltage jumps up to close to 420v....

The reason the hum changes is because the load on the power supply changes and that affect the ripple on the supply voltage. Lower dissipation = Lower Load = Less ripple = Less hum.


You are exceeding the max dissipation spec of 14 W for the 6V6. I'm surprised that it isn't red plating at 21W. I assume that you are prepared to burn through 6V6s quickly to get the sound you like.

[Zozobra]

Perhaps a job for a capacitor multiplier for the 2nd filter stage if you don't mind adding some silicon?

EDIT: actually, that's probably not a good idea with a tube rectifier :/

[DG1911]

It is 20 uF but that only applies to the first node. If you use your 510 Ohm resistor in between the new first node and the second node then the second cap can be larger. Even 220 uF or more.


The reason the hum changes is because the load on the power supply changes and that affect the ripple on the supply voltage. Lower dissipation = Lower Load = Less ripple = Less hum.


You are exceeding the max dissipation spec of 14 W for the 6V6. I'm surprised that it isn't red plating at 21W. I assume that you are prepared to burn through 6V6s quickly to get the sound you like.

Yep, I know it's well exceeding the max dissipation and I do expect to go through tubes. But, no red plating. Also understood that it's only the first filter cap that needs to be limited in capacitance for the 5Y3. But, all the data I'm finding shows 10uf. I'll experiment with upping capacitance after the first node. Also will probably spend the 20 bucks on a choke. Seems worth experimenting with.

Thanks.