Why would my screens be higher than my plates?

[rhinson]

no. a trannie doesn't have an imp.----just a turns ratio.

This much I understand.

So...take two different OTs for example. We have a 6L6 OT that has a 4k:8R turns ratio and a 6V6 OT with a 8K:8R turns ratio. Wouldn't the 8k:8R OT have more turns...i.e. more wire...i.e. more DC resistance?

So then we decide to use the 6V6 OT with 6L6s instead. So we 'halve' our speaker Z to reflect half the primary Z...so we are at 4K for the primary. Great. But we STILL have all that wire that's causing the DC resistance.

See what I'm saying? Yes...the reflected Z is the same (and we can do things to keep it the same...like changing the load/speaker Z)...but one has way more wire and thus way more DC resistance. And we can't change that...the wire/DC resistance is fixed. This is what I'm saying may actually BE the problem. But I'm just guessing...trying to learn.

the 4k/4ohms (8k/8ohms) trannie does have a little bigger turns ratio than the 4k/8ohm example. but that doesn't mean more resistance/wire necessarily. this depends on the trannie design and the internal winding geometry/pattern and it's complexity and style. if all things were equal then it would always be more resistance, but since there's the physical design involved this will vary.
as far as the problem in this thread, i'm not sure there is one, but there could be. 10v is a lot of diff between plate and screen at idle in a push/pull class ab amp. but i'm not sure there's enough info here to diagnose. i was just answering the general comment about a higher screen than plate voltage and how the output trannie can play a part depending on the other variables (choke, tubes, circuit). rh

[JimiB]

FWIW I have that same OT In My Dlite and it drops the normal 5-10V

['67_Plexi]

For what it's worth, you can get away with a small differential on 6L6's with the screens a little higher than the plates without any issues at all.
It is absolutely happening because of the high DC resistance of your primaries. THe values I gave you were from a 3k transformer. Your's being so much higher and with the hot bias current draw will increase the PD between the plates and screens. EL34's can be a bit more tempremental about this, but usually because you use a larger choke and a lower primary impedance the problem doesn't occur.

[rhinson]

For what it's worth, you can get away with a small differential on 6L6's with the screens a little higher than the plates without any issues at all.
It is absolutely happening because of the high DC resistance of your primaries. THe values I gave you were from a 3k transformer. Your's being so much higher and with the hot bias current draw will increase the PD between the plates and screens. EL34's can be a bit more tempremental about this, but usually because you use a larger choke and a lower primary impedance the problem doesn't occur.

yep el34's are always in the OTHER direction---ie--screens a little to a lot lower than the plates at idle. the screens in true el34's draw a lot more current at idle than 6l6/6v6 species. 6l6's/6v6's usually only draw one or 2ma at idle. el34 screens can be as much as 7 or 8 ma at idle, with 4 or 5ma being fairly typical. rh

['67_Plexi]

no. a trannie doesn't have an imp.----just a turns ratio.

This much I understand.

So...take two different OTs for example. We have a 6L6 OT that has a 4k:8R turns ratio and a 6V6 OT with a 8K:8R turns ratio. Wouldn't the 8k:8R OT have more turns...i.e. more wire...i.e. more DC resistance?

So then we decide to use the 6V6 OT with 6L6s instead. So we 'halve' our speaker Z to reflect half the primary Z...so we are at 4K for the primary. Great. But we STILL have all that wire that's causing the DC resistance.

See what I'm saying? Yes...the reflected Z is the same (and we can do things to keep it the same...like changing the load/speaker Z)...but one has way more wire and thus way more DC resistance. And we can't change that...the wire/DC resistance is fixed. This is what I'm saying may actually BE the problem. But I'm just guessing...trying to learn.

the 4k/4ohms (8k/8ohms) trannie does have a little bigger turns ratio than the 4k/8ohm example. but that doesn't mean more resistance/wire necessarily. this depends on the trannie design and the internal winding geometry/pattern and it's complexity and style. if all things were equal then it would always be more resistance, but since there's the physical design involved this will vary.
as far as the problem in this thread, i'm not sure there is one, but there could be. 10v is a lot of diff between plate and screen at idle in a push/pull class ab amp. but i'm not sure there's enough info here to diagnose. i was just answering the general comment about a higher screen than plate voltage and how the output trannie can play a part depending on the other variables (choke, tubes, circuit). rh

It's important that people don't look too deep and get things confused. What we are talking about in this thread is simply DC resistance. AC impedance is irrelevant. We are talking about DC voltage drops. Though DC resistance is one of the three elements that make up the impedance, at DC levels, capacitive reactance and inductive reactance are non-existant (if you do the math you will get a divide by zero error), therefore DC resistance is the only resitance that matters.
It's pretty much a given that a transformer with a higher impedance will have a higher DC resistance, but depending on the transformer , I guess there could be designs where it wasn't strictly the case, though I'm no deep expert on transformer design to those levels.

[Fischerman]

yep el34's are always in the OTHER direction

I was thinking that going to EL34s would improve the screen vs. plate voltage problem for the exact reason you state...or you could go the other way and use KT66s and probably make it worse!

I dunno the ins and outs of trannies and I noticed you seem to have some experience there rhinson so I wasn't gonna let the opportunity pass to hear your input. I was just thinking that usually an OT designed for 6V6s would have more DC resistance than one designed for 6L6s (even if you get the reflected Z right) and that by then running more current through it than 'usual' it would just exacerbate the problem.

[frischmann]

This has been a reallhy good thread.

thanks guys!

[Structo]

Hey frischmann
I had my D'Lite on the bench last night and took some voltage readings.
I have a 330R choke resistor and the 220K/150K plates on V1 & V2.
I have 110K/120K plates on the PI. I forgot to measure wall voltage but it is usually around 120vac.
My B+ voltages were as followed:

A) 433
B) 426
C) 420.5
D) 330.5
E) 324

V1) pin 1 194, pin 3 1.8 pin 6 205, pin 8 1.7

V2) pin 1 211.5, pin 3 1.7, pin 6 213, pin 8 1.65

V3) pin 1 292, pin 3 56, pin 6 288, pin 7 56

V4) (6L6) pin 1 34.8mv, pin 3 424, pin 4 427, pin 5 -46, pin 8 34.8mv

V5) (6L6) pin 1 34.3mv, pin 3 424, pin 4 427, pin 5 -46, pin 8 34.3mv

Then I changed the FET sim resistor to 150K from 220K and measured again.
My dropping string is stock other than the 330R choke and 150K FET resistor.

B+

A) 439
B) 433
C) 420.7
D) 318.6
E) 311.6

V1) pin 1 190.6, pin 3 1.74, pin 6 195.6, pin 8 1.65

V2) pin 1 204.5, pin 3 1.64, pin 6 205.8, pin 8 1.61

V3) pin1 294.6, pin 3 57, pin 6 291.3

V4) pin 1 36, pin 3 434.6, pin 4 431.8, pin 5 -47.3, pin 8 36

V5) pin 1 35.3, pin 3 435, pin 4 431, pin 5 -46.7, pin 8 35.3

I believe my PT is 325-0-45-325 200ma

So as you can see my screens are roughly 3-5 volts lower than my plates.

[frischmann]

Thanks Structo,
I have a new set of matched tubes and a new OPT (bassman) coming from Ruby.
We'll see how it goes from there.

I haven't changed the FET resistor yet.
Do you like it better as 150K?


Thanks,
PAul

[JimiB]

I'm not Structo but I liked the lower preamp voltages better!

[frischmann]

Thanks Jimib
That seems to be the consensus.

Hey Structo...
One more.

What are you using for you dropping string?

[Structo]

Thanks Structo,
I have a new set of matched tubes and a new OPT (bassman) coming from Ruby.
We'll see how it goes from there.

I haven't changed the FET resistor yet.
Do you like it better as 150K?


Thanks,
PAul

Hehehheh, I don't know yet what it sounds like yet.
I was up late last night messing with it but I didn't dare turn it on becuase Mrs. Structo was sleeping. And she is home today so maybe later I can put it through it's paces.

But I remember Scott saying that changing the FET is a way to drop the voltage across the whole string.

According to my notes, you want V1 in the 190-200v range.
V2 in the 200-205v range.

And I believe the PI should be around 290-300v.

So I am gettting closer, I hope these last tweaks help it.

I also put a treble bleed with a .001uf and 150K to ground.

And I changed the feedback resistore from 6K2 to 4K7 on the 4 ohm tap.

I need to keep better notes on this thing, months later it is hard to remember what you did inside......

[Structo]

Thanks Jimib
That seems to be the consensus.

Hey Structo...
One more.

What are you using for you dropping string?

I beleive I still have the 1K, 22K, 2K2 and 150K for the FET.

[frischmann]

I installed a new OT from ruby tubes today and what do you know, my voltages evened out as they should....

I seem to have some hum issues now...