EL84 Mic Pre

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

DaveP

Well-known member
Joined
Nov 8, 2005
Messages
3,019
Location
France
I thought I'd share this with the community.
I've been talking with Doug at Electromagnetic about the virtues of big 1930's tubes and no feedback circuits.  He just about convinced me so I decided to modify my REDD47 clone to save on metal bashing.

With no feedback, you lose possible instability and phase shifts etc, but you also lose the reduction of noise, distortion and low output Z, so you need to correct for that.

My circuit is inspired by the RCA BA-2C.
Through Morgan Jones I found the sweet spot for the E88CC which minimises distortion even with a 100k load pot; the tube is doubled up to reduce noise.  I wanted to use the original Sowter output TX so I needed a low plate resistance to drive it, brainwave! use a triode wired EL84!

I found the sweetspot for that by using the ruler from  Radiotron on the Plate graph, turns out to be 4~5k for minimum distortion.

2q3446w.jpg


I used big HT caps to stabilise the HT and reduce any hum to a minimum and a DC heater supply.


Anyway, I got -83dB noise which is not an issue in practice and a clean output of ~10VRMS into 600 ohms and I love it.  I also decided to put a switch on it to adjust the input Z for various mikes (even 500 ohms for an SM57) and that works ok.  I did not bother with a zobel after taking PRR's advice and the TX sounds fine.

Thanks to Doug and everyone whose posts inspired this project.

best
DaveP
 
Very nice indeed,
would you mind posting pictures of your build? Did you do it P2P or on veroboard? You mentioned the REDD 47, which I like very much mainly because of its "clean" sound. Is there any similarity in sound in comparison to your EL84 preamp? Congrats!
regards
Bernd
 
There are a couple of odd things in your circuit. Firstly, your switched resistors on the input transformer secondary. These will alter the frequency response of the transformer depending on whether you are loading the transformer excessively. You can verify this by simple measurements of frequency response. Transformers are not perfect, and don't just reflect the secondary loading back to the primary.
Second, the value of coupling cap from the output valve to the output transformer may be a bit low. Again measurements will confirm it.
 
Hey, a nice one with power. If you concider to have a little feedback after all and...
to raise the input margain 6-12 dB + to decrease the THD one decimal...
make the C1-47u switchable.
Cheers Bo
 
I thought you meant an EL84 as a MIKE AMP.  ;D

Not a line driver.

Triode EL84 might be a dandy input stage. Specially with that low-ratio transformer. Bring it up to say 20mA into 10K load. Gain won't be huge.
 
Great project! I like the unusual approach to everyday stuff ...

One thing though, would it hurt to add a grid bias resistor to the 84? I'm thinking a 1M to ground after R10 to protect the tube if the wiper goes intermittent.

Just a thought.  :)

 
Thanks for your comments,

"Would you mind posting pictures of your build? Did you do it P2P or on veroboard? " Bernd

This is the REDD47 before the conversion, interior is similar now but its now fixed in a rack so please excuse no final pics

b6bp5e.jpg

2rw81sx.jpg

nn3ddx.jpg


I'm 61, I don't like veroboard or PCB's or transistors much.  With P2P you can make the thickness of wire suit the job and there is not a flaky bit of foil between every connection.  There is so much I prefer about the 50's and 60's, like manufacturing instead of service industries, but I'm fast becoming a grumpy old git, the downside was rationing and lack of information, the internet is this age's saving grace, this forum is wonderful.

"There are a couple of odd things in your circuit. Firstly, your switched resistors on the input transformer secondary. These will alter the frequency response of the transformer depending on whether you are loading the transformer excessively. You can verify this by simple measurements of frequency response. Transformers are not perfect, and don't just reflect the secondary loading back to the primary.
Second, the value of coupling cap from the output valve to the output transformer may be a bit low. Again measurements will confirm it."
Radardoug

The whole idea of the switchable input impedance is to change the frequency response to give a little colour if required.
The coupling cap is fine, its double the REDD47 value and I will never be recording anything below 42Hz anyway, I'm not into organ pipes or traffic rumble.

"If you concider to have a little feedback after all and...
to raise the input margain 6-12 dB + to decrease the THD one decimal... " Bovox

I was not trying to eliminate distortion completely, otherwise I would have stuck with the REDD47, I was trying to get a vintage sound from the pre feedback days, as it happens, I can't hear any distortion from it anyway!


"Triode EL84 might be a dandy input stage. Specially with that low-ratio transformer. Bring it up to say 20mA into 10K load. Gain won't be huge. "  PRR.

Good point about the EL84 as an input tube but as you say there would not have been enough gain and I needed 40~50dB.

Great project! I like the unusual approach to everyday stuff ...

"One thing though, would it hurt to add a grid bias resistor to the 84? I'm thinking a 1M to ground after R10 to protect the tube if the wiper goes intermittent.

Just a thought. " Magnetosound

Good point I clean forgot that, I'll fit one when I next open it up.

thanks to all for positive comments
DaveP
 
Hi Dave,

Good to see another grumpy old git building stuff the 'old fashioned way' (I turned 60 a couple of months ago).

I have been toying with the idea of using a triode strapped EL84 as an output stage myself so I was very interested to see your project.

I have a couple of questions.

1. What distortion did you get at 10V rms into 600 ohms??

2. What is the measured gain, end to end?

3. What are the type numbers of the Sowter transformers??

4. Was the noise measured at max gain or with the pot down??

OK, that 4 questions but us old buggers are like that ;-)

Cheers

Ian
 
Great project! For some years there were almost no new tube experiments on this forum, and now it seems there's a new one coming every few weeks. What a textbook clean p2p build, just like your previous projects have been as well. Thanks for posting these Dave.

This kind of wasteful spending, using power tubes for mere line level drivers is exactly the direction I've also been interested in. The project "upgrade" PRR suggested with EL84 as the mic amp seems very interesting. It's especially interesting since the shape of distortion available from these FAT power tubes is not easily achieved with little novals. These big bastards don't so much clip, instead they gently bend over the sound like an umbrella.
 
Hi Kingston,

Thanks for the encouragement.

As usual PRR is on the right track, I can't try his suggestion due to lack of Heater supply in this build and also time, I am using this gear for recording, so must get on with that now.

Hi Ian,

1.  I don't have a distortion meter so I measured at just before onset of clipping with a pure waveform on the scope.  I have found that if I can't see it, then I can't hear it either.

2.  The transformers cancel so gain V1 is 28 and gain V2 is 12.7, overall gain is 355 or 51dB.

3.    Sowter 8540 for both.

4.    Max gain, but in practice I rarely turn it up past a quarter on.

The PSRR is only about -10.2dB on the EL84, hence the smoothing effort.

best
DaveP
 
Hey Dave, I´m glad your happy.
I did some measurements with the same input idea with NYD´s MILA project.
With the bypass cap  you have 1%THD at the output,with -14dBu applied at the input.Before Tx.
Without the cap you have 6 dB less gain but 1% THD at the output with -2 dBu at the input.
What I´m saying is that if you record something that does not need 51 dB gain,
of course padded down with the 100k pot. You have a little more control.
Read the heavy drummer with your hot condenser.
Maybe you can do with 46 dB and pleasure amount of THD?
Cheers Bo



 
bovox said:
With the bypass cap  you have 1%THD at the output,with -14dBu applied at the input.Before Tx.
Without the cap you have 6 dB less gain but 1% THD at the output with -2 dBu at the input.

This isn't such a clear cut case. THD is a highly inaccurate number when describing tube distortion. I mean, what is the dominant harmonic, or what harmonic spread is there in general?

Bypassed cathode has no feedback, and is less linear at least on paper. Unbypassed cathode has a moderate amount of local feedback and THD goes down. But the distribution of harmonics changes! Which sounds better depends on personal taste. And actually psychoacoustics. 3% THD that consists mostly 2nd harmonic is invisible to most people, while 3% THD that spreads randomly above maybe 3rd to 5th harmonic sounds absolutely nasty.
 
Hi Dave,

Nice job!  Love the p to p work.  One thing I'm unclear on is R15.  Would you mind explaining the reasoning behind loading the OT pri with 47K? 


Thanks
 
The reason I used a bypass cap is that it lowers the output resistance of the tube (or it doesn't raise it).  It was very hard to find a plate load that had decent gain and could drive 100k load.  I am actually looking for a degree of 2nd harmonic distortion in this project, that's supposed to be a tubes virtue after all.

R15 helps to stabilise the working point of the tube on the scope for looking at noise etc.

best DaveP
 
Hey, ..Kingston I did use my ears as well.It was more of an exampel over what you ev. can control when you have fixed gain 51 dB.. I did not like that distortsion ,2nd or not ,when it was a bit over those -14dBu.
Cheers Bo
 
DaveP said:
2.   The transformers cancel so gain V1 is 28 and gain V2 is 12.7, overall gain is 355 or 51dB.

4.    Max gain, but in practice I rarely turn it up past a quarter on.

So noise was measured at max gain and in your first post you said noise was -83dB. So with gain at 51dB I make that an EIN of 134dB!!!

Cheers

Ian
 
Sorry Ian my error, it was -63dB at max gain, not -83dB and so E.I.N was -114dB about the same as a V72.
It was about -83dB at the gain I use for my condensor mic.  Sorry for the confusion.

best
Dave
 
Back
Top