5F2A using EL84 Harsh Clipping

[shaun]

Hi All.  I just built a 5F2A and used an EL84 power tube. And of course, now I'm having trouble diagnosing a problem.
The amp powers up and sounds quite nice, until I hit the guitar with any velocity and a very harsh clipping sound blats through the speaker. Almost sounds like the speaker is clipping badly, but it happens at quite low volumes, so it seems to be internal. I'm guessing there's a mismatch between impedance between stages perhaps, but that's really just speculation. (I'm using new transformers for Champ from tubesandmore, and I have the options of 3.2, 8, and 16 ohm outputs - just mentioning this for reference).

Thanks for any help. Much appreciated.

[sluckey]

EL84 requires less signal to overdrive than a 6V6. What happens if you disconnect that 22K resistor from the speaker jack?

[shooter]

If the 22k doesn't get you there, drop the 25uF to something between 1 n 4.7uf

[nandrewjackson]

Wouldn't Shaun want to lower the 22K to something like 10K to reduce the gain there? Would cutting it out make the amp super duper noisy?


 

[shooter]

there might be a parasitic oscillation feeding back on itself, by opening up the NFB you break the "loop".

[shaun]

Wouldn't Shaun want to lower the 22K to something like 10K to reduce the gain there? Would cutting it out make the amp super duper noisy?

There are two 22k resistors in the build, so assuming we're discussing the negative feedback resistor, when the NFB is removed, there is supposed to be a little more noise - i.e. NFB is a sort of noise cancelling system, as far as I know. When I removed the NFB, there was no obvious increase in noise, but it didn't solve my problem either.

The sound I'm getting is an extremely harsh clipping noise, not typical tube distortion from an overdriven section. However, I might reduce the plate voltage at the preamp by increasing the plate resistors; I imagine that taming the voltage is a good place to start when clipping occurs - if it is actually clipping. But increasing the plate resistors to reduce plate voltage could increase plate current, so I'll learn something at least.

I'm also guessing that the 220k grid leak resistor on the EL84 could be adjusted to aid in balancing the impedance to the EL84. But that's guesswork. Still, I'll try it to see what happens.

[shooter]

I would encourage you to simply lift the 25uF on V1, this will drop gain enough to see if you're barking up the right tree

if all things are held equal, an 84 can be overdriven by ~ 1/2 the drive of a 6V6 and when they are overdriven there's a great chance they sound like 

[66Strat]

Just a thought, maybe the EL84 is operated at too high of a voltage for single end operation? I would expect to see something less than 300 volts at the plate and around 250 volts at the screen grid.

[shaun]

I would encourage you to simply lift the 25uF on V1, this will drop gain enough to see if you're barking up the right tree

if all things are held equal, an 84 can be overdriven by ~ 1/2 the drive of a 6V6 and when they are overdriven there's a great chance they sound like 

Thanks, Shooter. I will try that first then. Much appreciated. It sure does sound like !

Regarding the plate and screen voltages, I understand EL84s are sometimes run around 400v, and they can easily handle 300. But erring on the side of caution is always good for an inexperience builder like me:).  As for screens, I've read a lot and became confused, but the rule of thumb seems to be to keep them a little lower than PV; 50 volts lower sounds like too much, but I have a lot to learn in that area. Cheers!

[shooter]

maybe the EL84 is operated at too high of a voltage for single end operation?

I missed that!
I never bonded well with 84's.  they are like a pestilent teen-ager 
and 400vdc will NOT make it easier.

EDIT:
337vdc on the schematic is close enough, what is the VDC at the cathode?

[shaun]

VDC at cathode of V1 is 1.06, and cathode of EL84 is 12.7v.

I lifted the 25uF on the V1 - no change.

I'm starting to wonder if it is actually the speaker clipping, but if so, it's doing it at not even half volume. Yet there does seem to be quite a bit of travel in the cone, and it occurs with any speaker I use. I'm using the 5k winding on the primary of the OT, rather than the 8k option. I wonder if there's an issue there.

Thanks.

[shooter]

time to throw spaghetti n see what sticks 
put a AY or AU7 in V1 n check
double your speaker impedance n check - is this a head or combo?

bump Rk from 330 to 370-430 (any mechanically sound - gator-clipped kludge is fine:) .  Re-measure plate and cathode VDC

EDIT: remove the cap from the 84s cathode n check

[shaun]

Yeah. Thanks Shooter. I'm thinking I might trade the EL84 out for the usual 6V6. I've build two of these using 6V6 with no trouble. This is the second I've tried with an EL84, and the first one gave me big trouble as well. So I conclude you're right about EL84s. From my experience, they do not fit snuggly into a Champ.

But it's a good idea to try the ay and au first!

[shooter]

 

that's how us slow-lane'rs gain wisdom 

just a fwiw
I did a 5C16V6 n it was well liked.  The one below is my 84 version next, wound up as a one trick pony, it could "imitate" the Marshall crunch at reasonable volume, re-purposed to the next iteration, a quad 84.  That was built for audio and worked well.  Even fooled a couple cork-sniffers 
to be fair though, we ALL inhaled 

[shaun]

Wow. I'd like to see the schematic for the audio build - sounds very interesting, as I am thinking of attempting an audio build rather than a guitar amp next.

[shooter]

this is the WOW it works schematic, lots of tweaks undocumented latter 
I did find a note and a unlabeled chart of Freq vs VACrms at speaker, out to 13K it was ~~ flat-ish, pretty significant knee at ~17k
but I got busy building these cabs cuz I had the speakers trapped in a box and needed a home.  straight build of the speaker makers plans including ALOT of quilt stapled "stuffing".

where I got the cork sniffers, they "liked" the channel separation 
It did take some tweaking n moving to get the "sweet spot" centered between speakers n ~~ 2 FT' behind

you could only run it ~8-10W very clean but it was a loud 10W

so now you're qualified, go build something 

[DummyLoad]

some suggestions - see attached for reference designations.


try lifting C6 - EL84 bypass cap & add the divider network R8 & R9.


suspect that you have a wiring error. if it's a NFB issue, it can be fixed by reversing the OT primary -OR- secondary leads. as stated lift one end of the 22K (R11 in attached drawing) and use a clip/test lead to reinsert - amp hum/buzz loudness should DECREASE when you clip the 22K (R11) back to the cathode of V1b - if amp gets louder, NFB is positive feedback and you will need to reverse the OT secondary leads. EL84 is a higher mu tube, try increasing the value of 22K NFB (R11) loop resistor - something between 27K-47K may work better for you.   

lastly, is one of the leads of the OT secondary grounded? one lead needs to be, obviously, not the one connected to the 22K NFB resistor.

--pete 

[shaun]

Thanks Shooter - very interesting schematic, and probably way over my head. But I can learn . Great looking speaker cabinets, too.

Thanks Dummyload. That's a great 5F2a schematic which I will add to my collection. I'll compare it closely to the original and see what is different. I really needed something like that as a reference point, otherwise who knows where I'd end up.

Electricity - what a weird thing; can't see it (unless something catches fire or starts smoking), can't smell it (unless something catches fire or starts smoking), can't feel it (unless you get a hundred volts up your arm). Can't really guess what's gonna happen, even when the schematic is followed to the tee. But I suppose it's the mystery keeps me soldering late into the night, just to see what'll happen next.

Cheers.