SE EL34 verb/trem help

[goldstache]

I threw together this single ended amp and I’m really digging it.


The only trouble I’m having is when the trem intensity is at max with the slowest lfo setting via rate switch the oscillator imparts a little hash with regards to bias dipping.  I’m cathode biased at about 90% (74ma). So I’m wondering if the wide excursions are frying the EL34.


At max trem intensity I’m getting around 25VAC swing on the EL grid. When the intensity pot is lowered (to about 8, from 10, the VAC dips to about my EL’s cathode voltage (20vdc)
So I’m thinking it’s just swinging it into cutoff???????


My first time round not just copying a bias trem circuit and I’m really getting into the nuts and bolts of it.
I attached a .047uF cap signal to ground on the wiper of my intensity pot and it is swinging to about 15VAC and the hash is gone but is a little less intense obviously. Similar to the brown deluxe.


I’m wondering if I should cool off the EL34 bias to regain maximum chop at ten on intensity or........?????


Teach me stuff everybody!!!


Thanks as per usual

[goldstache]

Here is a slightly better schematic.

[shooter]

I’m cathode biased at about 90% (74ma).

with no signal measure Vk (top of cathode R)
hopefully you have a meter that has a min/max function otherwise watch close 

while the meter is hooked up monitoring, slam in your full volume guitar NO trem.
play 5 minutes and come up with your DC swing at the cathode.

now repeat WITH trem, post your high and low swings for both conditions

[goldstache]

Thanks for the steps Shooter!


Volume and Tone at max with hot tele bridge single coil and everything my paw can muster:


RK voltage:
Max-30.4V
Min-18.8V


RK voltage w/trem:
Max-48.7V
Min-2.8V


That's what I suspected, it's yanking the power tube too much. Right?


So now, what do I play with?  Trem oscillator strength, EL34 Bias, etc....??????

[shooter]

So now, what do I play with?

leave the trem OFF for now n re-calculate Pdiss
at 30.4vdc you're .1382 A (assumed 220ohm Rk, couldn't read it well)
at 18.8vdc you're .0855 A

measure your plate VDC, subtract the 30.4 then * .1382 A for Pdiss, compare with datasheet and post that #.

I've got the same kinda thing but just on paper and already asked your question, so you get to be the Canary in the mine 

[goldstache]

Thanks for being my beacon in the shaft Shooter!!


I have a 270 ohm cathode resistor in there.


30.4/270=.122A


(90ma PT) figured it was drawing more at max!!!!  Gets warm, but you can lay it hand on it for a period of time.


323v at plate
Minus 30.4
=292.6V (* .122A)
=35.6..........R??????????


What am I looking for here?  Sorry got lost in the maths.


I’m already smelling that I should cool the bias off and move to a 300 ohm resistor or more?

[shooter]

=35.6..........

Watts dissipated by tube

I have 2 datasheets one shows 25W max, the other shows 35W max so my math says aim for 30W
and yup, do that by bumping up Rk, 330ohm is a good start
keep trem OFF for now
enjoy the shaft work 

[goldstache]

Gotcha, thanks a bunch for walking me through that.  It's a nice sounding amp and slays a 5f2.
So rebias the power section than take another look at the LFO feed you say?  I can do that.  This is the first time I have realized why Leo went to optical. I still really love bias trem and am trying to wrap my head around the design centers for pairing a wiggler to the power tube(s) grid(s). In comparison to VK at idle, what kind of swing excursions should I look for in your opinion?  I should really just get the Min with trem on further away from 0 right?  If I bias colder the voltage at RK is higher in my mind.  So that low voltage (2.8V) minimum is the most dangerous culprit right???


Many thanks! 

[shooter]

I just asked my 2 ducks and they come up with a consensus of 25W max plate, so you might have to go to like 390, or maybe 420 

the math stays the same, only the variables change 

aside
I looked at my 34 build, used 270 for Rk, but I was at 402 plate so no real help

[shooter]

than take another look at the LFO feed you say

leave trem OFF til you are 90% - 100% of the 25W, OR someone else chimes with a different Pdiss for EL34's

[goldstache]

Sweet, 420 sounds perfect...... I can’t remember.


I’ll play around with it.  It’s got a used JJ 34 in it that was lying on the bench, so it could be giving up the ghost.  Though, upon completion of build the hash was not there and then a couple gigs later (trem above noon the whole time) it showed up so I suspected I was cooking her.


Thanks for the gracious help and please feel free to chime in if anyone has thoughts.


-joe

[shooter]

You know, once you've had SE there ain't no going back 
so keep working
keep looking under the pile for a PT that can crap out 400vdc @150mA MIN, while your looking snag a 6550, or KT88 before it gets stepped on then........

[goldstache]

Originally I was hoping to get 350 on the plate. 
I used a 260-0-260 @90ma stancor I had on hand. Did some light math and figured it’d pull about that power/pre at idle.
Went with it. Was hoping to get about 340 across the valve after cathode voltage deduction.


Do you think if I go to a 275-0-275 at 120ma I’d be closer to my original calculations under load?  That’s around 390v before cathode subtraction.  Or 300-0-300 @120ma????


Can you describe your SE EL34 in sound?  Be sure to use colorful guitar player adjectives, that’s all my eyes can read.

[PRR]

> at max with the slowest lfo setting ...imparts a little hash

Reduce C21. This will reduce the amplitude of the LFO signal at the lowest rate.

But it is allowed to adjust Amplitude when you adjust Rate. The controls interact, in the modulator details and also in the listener's mind.

[shooter]

Can you describe your SE EL34 in sound

I describe guitar strings by ...the fat E that I tune to D
my build is oranges to your apples, I did a PP self biased, interstage transformer for the PI.
will say the owner is a REAL player, told me he was looking for "that tone" for like 20yrs  musicians are weird 

get your dissipation dialed, bring in the trem n re-measure,
Big bottles in SE eat current n volts for fun 
haven't looked at you PS yet, but 90mA gets you to "sag", I typically like very stiff PS, personal choice

[goldstache]

Right on, here is the PS.

[goldstache]

Ended up with 360 ohm cathode resistor. 
About 30 watts at full bore.


Trem section got a 470k where 220k feed the intensity pot. Then a .047 from wiper to ground.  This seems to be teetering on destruction at mass volume and trem intensity. The “hash” noise is gone.  Swing is wide still, but has better balance of intensity at both the rate switch setting. I didn’t try PRR’s advice.  Though I certainly will after I adjust to the sound of it now. I may still be wiggling a bit aggressive. We’ll see.
Thanks for all the assistance. 
-j

[goldstache]

After much cooking and playing, I’m really happy with where this amp is operating, and it sounds good to my ears.


2 questions:


Calculating current across the power tube at idle and full bore, via cathode resistor voltage monitoring, how do I account for screen current? I’m ASSuming that the calculations just include it. What’s the best way to separate it?
Vdrop across screen resistor (to current) minus the idle bias current for the whole tube?
Just go for 10-15ma general rule? (Seems lacking since I want to know it’s swing at different play conditions)


Any others????


2nd question:
Adapt to push pull.
What would you do.


Already wanting a 300-0-0300v @250ma (GZ34) get the power tubes to be doing their thing a bit more than the 260-0-260 @90ma (SS) I used for the single ended proto.
Then just droppers to get the preamp a little closer to where it’s operating in the SE???
 Maybe cathodyne PI?

Thoughts?

[shooter]

how do I account for screen current

your Vk Rk math includes both plate n screen, n yup, do the Vdrop math at G2, subtract from cathode current and you're left with plate I.

glad you like the sound.
as too PP 34's, look at what the famous have done, copy/paste, tweak to taste
or invent a new wheel 

[PRR]

> how do I account for screen current?

6V6, 6L6, 6550, etc: Ig2 is about 5%-7% of Ik.

EL34, EL84: Ig2 is about 15% of Ik.

[goldstache]

After jamming this amp in many cabs in different rooms, I’ve come to the conclusion that it’s just too bright. Especially with open back cabs.


I’ve tried raising coupling caps to obscene values .47
A little bandwidth extension but almost imperceivable difference. 
Got any tips on reigning in the freq response.  Amp has no NFB and cathodes are bypassed.


Do I ditch the tone control and try a cut control? Tweak values in tone control?


Is there something I’m missing???  Why is this thing so bright?  I’m playing with the tone control off mostly and that’s after tweaking the guitars controls.


Any pointers welcomed.
Thanks.
Schemo up top

[VMS]

First i would remove the 500pf C8 cap that is parallel with R7 470k resistor.

[goldstache]

I’m sorry, that part is already off the schemo.  The only other stuff I’ve changed are in the trem section and oscillation mixer to 34’ grid.
 
Thanks for the reply

[PRR]

.

[goldstache]

Classic cut and bypass increase coming right up!


Thanks

[goldstache]

So after trying out the proposed mods, which did seem to improve the low end response, I finally did the math on the EL34 cathode decoupler.


My 100uF is starting to roll off at 70hz or so. 
So maybe a 50uF to move it down to 10hz roll would be better???


I never realized that was a cheap place to shave bass in a cathode biased design.

[PRR]

> I finally did the math on the EL34 cathode decoupler.

Bypass.

> My 100uF is starting to roll off at 70hz or so.
So maybe a 50uF to move it down to 10hz roll would be better???

I don't follow the math.

Or why you might want to "shave bass" in an amp which is "too bright".

[sluckey]

I would experiment with a capacitor between pins 5 and 8 on the EL34 socket. Easy to do right on the socket. Start with .002, then .001, 500pf, 100pf, etc. Should find something that is pleasing.

[goldstache]

Thanks for calling me on the verbiage.
Bypass cap calc:


I was using
1/2*pi*hz*R


Times 1,000,000 for the uF value. 
I have a 330 bias resistor.
For a hz freq of 10hz I was getting a 48uF.


Maybe I got that all wrong.

[PRR]

> For a hz freq of 10hz I was getting a 48uF.

That's right. But above you said you had 100uFd and 70Hz.

BTW, we don't normally design the power stage of a guitar amp for sub-sub-sonic response. We don't have signal below 82Hz, out speaker won't reproduce it, and when we overload hard a low-frequency cap takes longer to recover.