Humdingers RULE!!

[Gtr0]

Hi Gang.

I needed to drop a note on my latest experience with high'ish gain amp noise.

Having built several high gain (and medium-high gain) amps over the last number of weeks, I have become accustomed to avoiding noise by following simple "rules of engagement" while not resorting to DC heaters or elevation (though I do have an interest in the latter). Besides keeping wires as short as possible, concentrating and prototyping component layout/placement, keeping "like" grounds together and connecting them upstream, using shielded wire on signal runs over an inch long, twisting wherever possible (i.e. heaters, or longer runs) etc, and on new/original designs, as of late, invoking the artificial CT with 2x100Ω resistors to ground. I have had great luck with these methods.

Backstory: I have an old'ish, if 1991 (the year I graduated High School!) were considered "old", Ampeg VL-502 - a Lee Jackson amp... basically an 800 with another tube gain stage with some other tricks. I have done some modifications to the circuit to give it some high end percussion - something that was sorely lacking in the amp. While I very much like the amp - it left some other characteristics to be desired. I decided to take a late 70's 2203 "clone" I built and convert it to a VL type amp, see if I could achieve what I wanted out of it.

Forward story: The 2203 clone was of standard early 70's Marshall layout in that it was on turret as apposed to the Marshall PCB of the day. Everything was setup as Marshall standard. It was pretty quiet... at least more so than my factory JCM800 studio thingy I have (the new mini JCM800 line).

Once I added that extra tube stage, however, it quickly became THE noisiest amp I have ever had. EVER.

So I quickly went through (okay, not SO quickly) and modernized it a bit. Meaning that I tightened and improved the grounding (though I did not convert it to star... maybe one of these days when I have nothing better to do), replaced all preamp/tube/control pot wires with their shortest equivalent, shielding all grid wires... in short I had made valiant efforts to improve it while not going absolutely crazy - which would be building it out from scratch ;-).

In the end it still had way too much noise. Even after swapping tubes several times. I hooked up my scope to it and could see crazy amounts of noise coming into the circuit at the plate of v2a. The grid and all circuit points before that seemed fine (though my digital Tek isn't the best with it's seemingly low monitor resolution). V2a was previously V1a... there was never an issue with noise to look at it on a scope, so I never did.

After replacing wires, cleaning and re-hitting solder points, I figured my last play was to add a humdinger, elevate heaters, or convert to DC heaters... in that order.

So I started with the humdinger... At this point you, the reader, if you made it this far without going into a coma, should know that I had left the heater CT connected to ground. At this point I replaced that with 2x100Ω resistors to a 100Ω piher mini pot with wiper to ground.

Started up the amp and it was STILL noisy... this was with the pot at about half way. Once I started twisting the pot however that noise almost disappeared completely. Went from the noisiest to one of the, if not THE, quietest amp of the lot.

I had to post this because I had been tearing my hair out literally for days on this, trying to figure out what the issue could be. A small handful of parts equating to about $1 saved the ... year!

Humdindys will go in EVERY amp from this point on!!!!!!!!!!!!!!!!!!!

[m1989jmp]

Hi Gang.

I needed to drop a note on my latest experience with high'ish gain amp noise.

Having built several high gain (and medium-high gain) amps over the last number of weeks, I have become accustomed to avoiding noise by following simple "rules of engagement" while not resorting to DC heaters or elevation (though I do have an interest in the latter). Besides keeping wires as short as possible, concentrating and prototyping component layout/placement, keeping "like" grounds together and connecting them upstream, using shielded wire on signal runs over an inch long, twisting wherever possible (i.e. heaters, or longer runs) etc, and on new/original designs, as of late, invoking the artificial CT with 2x100Ω resistors to ground. I have had great luck with these methods.

Backstory: I have an old'ish, if 1991 (the year I graduated High School!) were considered "old", Ampeg VL-502 - a Lee Jackson amp... basically an 800 with another tube gain stage with some other tricks. I have done some modifications to the circuit to give it some high end percussion - something that was sorely lacking in the amp. While I very much like the amp - it left some other characteristics to be desired. I decided to take a late 70's 2203 "clone" I built and convert it to a VL type amp, see if I could achieve what I wanted out of it.

Forward story: The 2203 clone was of standard early 70's Marshall layout in that it was on turret as apposed to the Marshall PCB of the day. Everything was setup as Marshall standard. It was pretty quiet... at least more so than my factory JCM800 studio thingy I have (the new mini JCM800 line).

Once I added that extra tube stage, however, it quickly became THE noisiest amp I have ever had. EVER.

So I quickly went through (okay, not SO quickly) and modernized it a bit. Meaning that I tightened and improved the grounding (though I did not convert it to star... maybe one of these days when I have nothing better to do), replaced all preamp/tube/control pot wires with their shortest equivalent, shielding all grid wires... in short I had made valiant efforts to improve it while not going absolutely crazy - which would be building it out from scratch ;-).

In the end it still had way too much noise. Even after swapping tubes several times. I hooked up my scope to it and could see crazy amounts of noise coming into the circuit at the plate of v2a. The grid and all circuit points before that seemed fine (though my digital Tek isn't the best with it's seemingly low monitor resolution). V2a was previously V1a... there was never an issue with noise to look at it on a scope, so I never did.

After replacing wires, cleaning and re-hitting solder points, I figured my last play was to add a humdinger, elevate heaters, or convert to DC heaters... in that order.

So I started with the humdinger... At this point you, the reader, if you made it this far without going into a coma, should know that I had left the heater CT connected to ground. At this point I replaced that with 2x100Ω resistors to a 100Ω piher mini pot with wiper to ground.

Started up the amp and it was STILL noisy... this was with the pot at about half way. Once I started twisting the pot however that noise almost disappeared completely. Went from the noisiest to one of the, if not THE, quietest amp of the lot.

I had to post this because I had been tearing my hair out literally for days on this, trying to figure out what the issue could be. A small handful of parts equating to about $1 saved the ... year!

Humdindys will go in EVERY amp from this point on!!!!!!!!!!!!!!!!!!!

Yeah, had a 70s Twin Reverb come in with dead power tubes; after replacing filter and bias caps, still had ridiculous amount of hum. It went completely away when I tweaked humdinger pot on the back panel.

[Chuck H]

You are only saying the amp was "noisy". And now it's not "noisy". You didn't specify hum. Did you experience noise reduction other than hum reduction by adding the humdinger circuit?

[Helmholtz]

100Ω piher mini pot...

With 100R it should have a power rating of 0.5W or more.

[Gtr0]

You are only saying the amp was "noisy". And now it's not "noisy". You didn't specify hum. Did you experience noise reduction other than hum reduction by adding the humdinger circuit?

Sorry, I didn't receive the usual notification on this post... Well, it was hum with buzzy noise on top, I'd do better to remember to separate those in description in the future. So the hum went away and the buzz was significantly reduced to more than acceptable levels (for me)

With 100R it should have a power rating of 0.5W or more.

Is that also the case when there are two 100R resistors - one on each leg - feeding the trim pot? So from each leg of the heater supply I have 1x100Ω 1/4 watt each tied to the in/out of the trim pot with wiper to ground.

[Dave H]

Is that also the case when there are two 100R resistors - one on each leg - feeding the trim pot? So from each leg of the heater supply I have 1x100Ω 1/4 watt each tied to the in/out of the trim pot with wiper to ground.

If you have 3 x 100R in series across 6.3V then each resistor sees 2.1V and the power is 2.1^2 /100 = 0.044W

[Helmholtz]

Is that also the case when there are two 100R resistors - one on each leg - feeding the trim pot? So from each leg of the heater supply I have 1x100Ω 1/4 watt each tied to the in/out of the trim pot with wiper to ground.

No, I misread your circuit description (language barrier). See reply from Dave H.
Apologies for confusion.

[pdf64]

Bear in mind that the rest of the tubes will be liable to their heater-cathode insulation being damaged if a power tube undergoes the common failure mode that shorts the tube and pops any standard heater balancing resistors.
To mitigate that, I suggest much beefier but higher value back up balancing resistors (eg 1k 5W 500V),or a beefy ~50V SIDAC / series opposing zener pair, between a heater line or CT and 0V. And a fast blow HT fuse, ie after the reservoir.

[Chuck H]

Bear in mind that the rest of the tubes will be liable to their heater-cathode insulation being damaged if a power tube undergoes the common failure mode that shorts the tube and pops any standard heater balancing resistors.
To mitigate that, I suggest much beefier but higher value back up balancing resistors (eg 1k 5W 500V),or a beefy ~50V SIDAC / series opposing zener pair, between a heater line or CT and 0V. And a fast blow HT fuse, ie after the reservoir.

Maybe this should be added to R.G.'s "immortal" mods

[Helmholtz]

if a power tube undergoes the common failure mode that shorts the tube

You mean the power tube's plate shorting to its heater?

[Chuck H]

You mean the power tube's plate shorting to its heater?

Of the few times I've had power tubes acutely fail, that is POP rather than just demonstrate poor performance, two have been shorts to the heater. So in my experience this would be a common failure mode.

[pdf64]

You mean the power tube's plate shorting to its heater?

I don't know whether it's the plate, g2 or something else to sends HT current to the heaters; I suppose g2 seems the most likely contender? Perhaps that would explain why the 1W 470ohm g2 current limiters of Fender type 6L6 amps often get blown along with humdingers/balancing resistors.

Maybe this should be added to R.G.'s "immortal" mods

Wow, that would be my career high! I don't think RG updates them often though, so I'm not holding my breath

[Helmholtz]

suppose g2 seems the most likely contender? Perhaps that would explain why the 1W 470ohm g2 current limiters of Fender type 6L6 amps often get blown along with humdingers/balancing resistors.

Makes sense as even the g2 supply might be able to inject something like 200mA DC.
Seems a grounded CT heater winding is a safer solution.

[g1]

Crate uses a SIDAC in the BV120

[pdf64]

Crate uses a SIDAC in the BV120...

Yes, thanks, I knew I remembered it from somewhere
The 120V (breakover voltage?) seems somewhat high, better than nothing but may not eliminate damage to some tube types. Something lower would be better surely? Though it seems tricky to find much under 120V.
Any idea of a part number for it, or any other spec? It's sketchy on detail to say the least.