Valve mic preamp design incoherent rambling

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Replacing parafeed choke with SE Trafo. A resistor in series with the transfomer mimicks higher dcr. Right. If we look at the v74 it has a 45H choke, dcr ca. 1k in the line stage. If a add a 670ohm resistor i series with the LL1671 (total cdr 1k) it will drop 20v at 20ma.
But why do you want to drop 20V? 4V is enough to put the E81L in the 10mA area with a grounded grid resistor. You could run the output stage at 10mA with 400 ohms total DCr, which is just adding 60 ohms in the cathode.
 
But why do you want to drop 20V? 4V is enough to put the E81L in the 10mA area with a grounded grid resistor. You could run the output stage at 10mA with 400 ohms total DCr, which is just adding 60 ohms in the cathode.
I re-read your earlier post and I get what you mean, the choke can swing negative (y)
 
But why do you want to drop 20V? 4V is enough to put the E81L in the 10mA area with a grounded grid resistor. You could run the output stage at 10mA with 400 ohms total DCr, which is just adding 60 ohms in the
Yep I missread u I like your simple solution 👌
 

Attachments

  • 31B03542-7BDF-4162-AE99-48E28271D9BC.jpeg
    31B03542-7BDF-4162-AE99-48E28271D9BC.jpeg
    78.6 KB · Views: 97
Last edited:
I rebuilt a small guitar combo a while ago and used an output stage exactly like you have drawn right now, albeit with a russian 6E5P output tube. I actually had to apply some positive bias to the grid because the output transformer's DCR put it too far negative. Just wanted to affirm you in your plans, it should work out just fine (y).
 
I rebuilt a small guitar combo a while ago and used an output stage exactly like you have drawn right now, albeit with a russian 6E5P output tube. I actually had to apply some positive bias to the grid because the output transformer's DCR put it too far negative. Just wanted to affirm you in your plans, it should work out just fine (y).
great to hear (y)
 
Why? You don't need voltage across a xfmr for it to work. You need current. In what is it different than putting the same xfmr in the anode?
I agree you need some DCR for cathode-bias to work. The datasheet shows perfect adequacy.

Haha, of course. I think we were talking at cross purposes, a different topology. The circuit had morphed before my reply had time to get cold anyway so it was moot.
 
Theres a kind of a cross polination in several recent preamp threads going on , its no reason for appologies and its all good food for thought at the end of the day .

Disc seal RF triodes Ive long wondered about usage in audio , I see now the high gain naturally means 'Miller effect' becomes an issue at high frequencies .
Grounded grid /fet driven cathode looks good though.
Could the 6C45Pi be the native compatriot to a 6s17k-v front end ?
Im liking the power hungry and costly iron pre topologies that have come into the convo of late , you only need a channel or two for vocals and lead acoustic instruments in an overdub situation either way , its not the end of the world if you throw away a few extra watts off the anode ,
now were suckin' diesel :p
 
Theres a kind of a cross polination in several recent preamp threads going on , its no reason for appologies and its all good food for thought at the end of the day .

Disc seal RF triodes Ive long wondered about usage in audio , I see now the high gain naturally means 'Miller effect' becomes an issue at high frequencies .
Grounded grid /fet driven cathode looks good though.
Could the 6C45Pi be the native compatriot to a 6s17k-v front end ?
Im liking the power hungry and costly iron pre topologies that have come into the convo of late , you only need a channel or two for vocals and lead acoustic instruments in an overdub situation either way , its not the end of the world if you throw away a few extra watts off the anode ,
now were suckin' diesel :p
Yep this little planar tubes really got some super power and I think its the combination. Low noise and microphonics combined with very high gain and voltage headroom. They were build to survive the enviroment of rocket noise con... But Glen Miller made them turn down the treble 🤣

With 1:10 + maxed out cascade you could get 80dB of gain in one stage and that really is all the gain you would ever need.

The easiest way the use 6s17k-v is with grid bias of around 47K, which matches well with a 1:5 ITP. In this setup u got the shared cathode/filament connection grounded and u dont need to get super esoteric on the heather supply noise issue.

You want to cascade 6s17k-v and 6c45P?
 

Attachments

  • 194ADB31-9B36-499E-BF21-30376F339A64.jpeg
    194ADB31-9B36-499E-BF21-30376F339A64.jpeg
    61.9 KB · Views: 54
Last edited:
The main 'problem' with the cascode is that the top tube represents a much lower impedance load to the bottom tube than the usual plate resistor. The voltage gain of the bottom tube is therefore very low, which is why Mr. Miller is not a problem, but this lower than usual load is also what creates the distortion.

I looked for any distortion data I'd taken regarding tube cascodes I'd played with in the past and found the attached in amongst an old email .
I didn't note all the parameters unfortunately, but it would have been a typical stage as used on the front end of a mic amp. Peak to peak voltage of 75V. Distortion doesn't look any worse than a single triode to me, and the H2 vs H3 etc. is triode like too.

Edit: no global feedback or local feedback, just a "naked" cascode into a buffer.
 

Attachments

  • Cascode @ 75V P2P.png
    Cascode @ 75V P2P.png
    53.3 KB · Views: 62
Last edited:
Hi Winston, I agree those distortion figures are not bad at al and as you say very triode like. Can you remember which tube this was? Was this unloaded?

Hey Ian,
yes I completely forgot to mention that the tube was an E88CC. Most likely Tesla 1970's production as those were the best I was finding at affordable prices back then.
Things have changed and, as you've indicated, your research shows current E88CC's from EH are good these days.

As for AC load, I'm 95% certain this will've been DC coupled to another E88CC as cathode follower with a solid state load for constant current. So not any appreciable external loading on the cascode.
It's possible there was some cancellation of distortion happening between the cascode and follower - I certainly remember that happening with various similar schemes, but still...

My point really was to encourage you to give your idea for a similar cascode a try :)
 
Last edited:
My circuit was something like this. I'm guessing on some of the R values as I can't find a damn drawing. I did realize that the cathode follower wasn't loaded with a CCS, I'd remembered that wrong earlier. Anyway:

Basically, I was looking to replace an EF86 in terms of gain and lack of Miller, but be quieter.
The bottom of the cascode is operating at a higher current (about 7mA in total) to raise Gm for gain, and also for lower noise. The top of the cascode is operating lower current so as to use a bigger Ra.
In my actual proto, what's shown as R8 on this circuit was probably a couple of Mosfets as a cascoded 5mA or so CCS. But you get the idea.

The voltage for the cascoding grid comes from the cathode follower output as DC feedback. A "driven" cascode I guess you'd call it.
With higher source impedances (such as the 10K I got from my 1:7 input transformer), high frequency distortion and higher order H was lower when I set it up like this.
I've seen pentode screen grids driven this way too (the screen grid voltage tracking the plate voltage) so nothing new about it.

Nothing new in any of it actually, but then... When is there ever anything new under this sun? :D.

Edit: Removed my earlier attachment showing 150V peak to peak distortion because I can't be sure that one came from using a 320V supply, my old email is a bit confusing on that aspect so...
 

Attachments

  • Hinson's Faux EF86.png
    Hinson's Faux EF86.png
    108.3 KB · Views: 73
Last edited:
Interesting circuit. Reminds me a lot of Rupert's tube pre with an EF86 first stage followed by a CF which drive the EF86 screen grid in almost exactly the way you derive to cascode top grid voltage.

I am pleased you were trying to encourage me too try a cascode. I admit I have not built any but I have simulated one to find out how good the distortion is and it was not very good. But I will check it out again.

Cheers

Ian
 
Interesting circuit. Reminds me a lot of Rupert's tube pre with an EF86 first stage followed by a CF which drive the EF86 screen grid in almost exactly the way you derive to cascode top grid voltage.

I've peeked at that Neve desk circuit on and off for 15 years or so and never even noticed that detail. Good catch!
Like I said, nothing new under the sun 😃
I think I first noticed the trick in an old Hi-Fi valve amp and it's obviously been done a lot in solid-state cascode circuits.


I am pleased you were trying to encourage me too try a cascode. I admit I have not built any but I have simulated one to find out how good the distortion is and it was not very good. But I will check it out again.

Well, 6922 valve cascode operation was used in other systems such as TV, communication, and oscilloscope amplifiers, and lack of distortion (albeit up to higher frequencies) is just as important there as it is to us working closer to DC so, surely it can't be all that bad.

I don't know that I'd approach things the same way now as that circuit I posted, and I certainly wouldn't advise anyone using it as a model as I'm probably off on remembering a couple of points (that value cathode R was a complete guess). Just tweak and twidde the bits as necessary.
 

Attachments

  • Neve tube input.png
    Neve tube input.png
    94.9 KB · Views: 55

Latest posts

Back
Top