relationship of cathode voltage to noise?

[Phil_S]

balance pot artificial center tap, bottom plate firmly attached to chassis. Express presence control, francesca layout, shielded volume pot to v1b, 33k grid resistor on V1a. Oh, I almost forgot 1.8k Rk V1a/b. This depends on your K voltage. I only had 1.02 on A and .89 B. This bumped both cathode voltages to about 1.33 with my 174 volt plates on V1. Cut the noise a bunch, I may pull the 33K resistor.

sorry to keep beating the same drum but my TW is silent
from thread: Noisy Express

You really got my attention. Can we pursue this for a bit? I feel like this one put me in a fog. I've got hiss beyond what I'd like. Are you suggesting that there is a direct relationship between noise and Vk? This would be new territory for me.

I can't measure anything in my TWX right now because I'm waiting for a replacement OT (egad, don't ask; it'll be here in a day or two).

My records indicate that V1A Va=193 Vk=1.26 and V1B Va=276 Vk=4.42. I used the "Kelly 90" schematic, so the R/C for V1A is 1k5/22mf and for V1B it is 2k7/22mf.

I didn't think much of it, but I'm really disliking that Vk=4.42. Don't know if that's significant or not, but your post (from another thread on noise) makes me want to explore this.

Is there an ideal target for voltages on V1? What do you think they are? Suggestions on how to get there?

The amp has shielded in to both grids on V1. That helped some, but not enough. 5751's helped too, but I keep thinking this could be better and that 5751's aren't the "right" answer. (Cheating!)

If I can find something to use for a bias pot without spending an arm and leg, I'll do that, too.

[skyboltone]

Phil, this is a question for one of our brighter bulbs than me. Larry?? Dana?? Anyway, we set the cathode voltage positive with respect to ground based on the idle current flow through the cathode resistor. Meanwhile the filament is boiling off negative electrons, some of which leak across to the cathode and pass from there to the plate and make hum and noise. Now, the reason we set the cathode positive is because that makes the grid negative with respect to the cathode thereby blocking electron flow from cathode to plate through the tube. That is until we wiggle the grid.

So, with my limited background in such things, if we raise the cathode voltage we cut the gain of the tube slightly (more wiggle required) and have less likelihood of electrons leaking across to the cathode from the filament. We keep the electrons on the filment where they belong (by making them higher potential than the cathode they rather than having them attracted to the cathode. Works for me.
Take with grain of salt.
Dan H

[gearhead]

SB.

How did you implement the balance pot artifical tap? Found your other thread (100 or 250ohm pot), but am not much up on heater mods.

[Phil_S]

I'm having some difficulty getting my head around the bias pot trick. I'm working with hiss. I thought the heater string injeccts hum. Aren't these tow different animals? Just trying to sort this out.

BTW, if I had the right pot, I wouldn't hesitate to put it in. If figure, it couldn't hurt. As it stands, that won't happen until I have another parts order and that's likely to be months away.

[skyboltone]

| 6.3VAC |
| |
| |
|\/\/\/\/\/\/\/\/|/\/\/\/\/\/\/\/\/\/\| 250ohm 2 watt pot
| | |
| | |
|
to filaments | to filaments
\/
Ground wiper
Here you go, one of them old fashioned text drawings. Novosibir,(Larry) recommends putting 45-50VDC on the wiper instead of the ground. If it's humming still, you got problems with cross talk between the Pt and the Ot. Well these wonderful folks at the software convention don't like text drawings. its ineligible or illegible or plane unreadable. Let me try something else. Do this make sense?

[img:504:150]http://www.lynx.bc.ca/~jc/568humAdjust.gif[/img]


Here's the rest of that link: http://www.lynx.bc.ca/~jc/hum.html

Dan

[skyboltone]

I'm having some difficulty getting my head around the bias pot trick. I'm working with hiss. I thought the heater string injeccts hum. Aren't these tow different animals? Just trying to sort this out.

BTW, if I had the right pot, I wouldn't hesitate to put it in. If figure, it couldn't hurt. As it stands, that won't happen until I have another parts order and that's likely to be months away.

You got a 5k? anything just as an experiment. There were so many steps to the noise killin' effort that I didn't document but it was a high pitched scream added to the primary that corresponded to a mixing of the note, it's harmonics and a 120hz carrier.

[Phil_S]

Dan:
I've got a big old 1K locking pot, probably mil spec. It will take a hole in the chassis to make it work and some relocation of the artificial CT. I'll think about this. It's the old carpenter's rule: measure twice, cut once. I've got a hole to drill...

AFA your statement about the increased cathode voltage...are you able to clarify the V1 info. Are you saying V1A went from 1.02v to 1.33v and V1B went from .089v to 1.33v? Are you using identical R/C on each section? So, you have 1k8/22mf...?

Maybe your amp is different from mine, but I've got each section on a separate R/C for the cathode. V1A on mine is 1k5/22mf and V1B is 2k7/22mf.

Thanks.

Phil

[skyboltone]

Dan:
I've got a big old 1K locking pot, probably mil spec. It will take a hole in the chassis to make it work and some relocation of the artificial CT. I'll think about this. It's the old carpenter's rule: measure twice, cut once. I've got a hole to drill...

AFA your statement about the increased cathode voltage...are you able to clarify the V1 info. Are you saying V1A went from 1.02v to 1.33v and V1B went from .089v to 1.33v? Are you using identical R/C on each section? So, you have 1k8/22mf...?

Maybe your amp is different from mine, but I've got each section on a separate R/C for the cathode. V1A on mine is 1k5/22mf and V1B is 2k7/22mf.

Thanks.

Phil

Mine's a Liverpool Phil. Schematic shows 1K5/22 on both V1A and V1B. You can do the balance pot test with the pot in your hand. Just clip the two outside leads one each to pins 4+5 and 9 of any pre amp tube and the wiper to ground back at the PT. Roll the pot back and forth for minimum noise. Just make dang sure you lift the filament winding CT and the artifical CT or you'll defeat the purpose. You are building an adjustable artificial CT. Any way, what are your cathode voltages??

[peesinstew]

I got a 250 Ohm 2W pot from digikey and installed it.

For some reason it seems to dial in/out more buzz than hum. When I have it dialed in to the quietest setting, I think I'm getting more hum than I was getting without the pot. Luckily, I didn't have much hum to begin with. I do have it mounted in a hole that is right next to the PT, so maybe that has something to do with it. The hole was already there, so I used it out of convenience.

[gearhead]

Novosibir,(Larry) recommends putting 45-50VDC on the wiper instead of the ground.
Dan

Thanks for both drawings!

Where would you pull that voltage from? Voltage divider at the end of the power filter chain?

Five 9V batteries? LOL.

[jjman]

I'm not understanding Vk=4.42. And the plate voltage is higher than v1a? It sounds like v1b is running way hot? Double check the K resistor and cap on it. Maybe the cap is shorted? You could remove it while diagnosing at idle. I would think a shorted cap or 270 ohm k resistor could cause excess hiss from excess idle current.

But then again, I would think the plate voltage would be low. I don't have access to the schematic at the moment. ;(

Hiss would normally not be related to the heaters since they are 60hz=hum. Hiss is normally from DC flowing thru something that's not a "clean path" for it, like a noisy plate resistor or bad tube.

[Phil_S]

jjman: Sometimes all you need is for someone to hold up a mirror. My TWX is down right now with a blown OT, replacement on the way. Don't ask...it was a foolish error on my part. The point is that I can't power it up right now.

If my record is correct with V1B Va=276 and Vk= 4.42, Va is 2v under the B+5 reading. Given fluctuations possible for line voltage, it suggests the plate load resistor is open (it's not) or there is a short somewhere back to the B+ rail, which I will now need to hunt for. I think you are right that V1B is running way too hot.

Thanks for pointing this out. As soon as the new OT arrives, I'll look into it.

Phil

[jjman]

Now that I'm home I can see the scheme. Both v1 plate resistors come off the same PS node and are the same value. The higher v1b K resistor means an expected cooler idle which mean less voltage drop on it's plate resistor. So the v1b plate should have a higher voltage than v1a's plate to some extent.

v1a: 1.26/1500=0.8ma
v1b: 4.42/2700=1.6ma

So v1b is idling at a higher current when it should be a lower current that v1a. But this assumes the k resistor on v1b is truly 2700ohms. If it is, I would suspect a somewhat shorted k cap or plate resistor.

Of course it's all hypothetical until you're ready to fire it up.

[Phil_S]

I think I may have found a weak solder where the 100K meets the .022mf Mid & Bass coupling caps in the tone stack. Continuity is iffy reading ohms, sometimes good and sometimes not. I should see 100K. The R itself reads fine. Probe from the tube pin to the coupling caps says it's open. I'm not sure this should account for what I've reported. We'll see what happens. I'll save it for when I fire up the iron.

This is one reason groups like this work. It can be difficult working alone, even when you seem to know what you're doing.

Thanks.

[skyboltone]

Novosibir,(Larry) recommends putting 45-50VDC on the wiper instead of the ground.
Dan

Thanks for both drawings!

Where would you pull that voltage from? Voltage divider at the end of the power filter chain?

Five 9V batteries? LOL.

Voltage divider off of B+. Take two resistors, say a 270K and a 33K. series them, attach the 270K to B+, the 33K to ground at same place as first caps then pull voltage off of junction of the two resistors. It's straight ratio and proportion. Check total current squared times 303K and make sure you don't exceed power handleing of resistors. Be above your power tube bias voltage.