Single tube mic pre with continuous gain control

Help Support GroupDIY:

ruffrecords

Well-known member
Joined
Nov 10, 2006
Messages
13,755
Location
Norfolk - UK
One problem with tube mic pres is varying the gain over a reasonably large range. This is usually a problem caused by the difficulties in closing the NFB loop at dc in tube circuits. That is why you see lots of tube mic pres with gains variable only over a small range  (REDD47, V72 etc). I solved this problem in the EZ Tube Mixer project but it took one and a half tubes to do it. So I wondered if I could come up with a simpler variation on the same theme that would allow you to build a one tube mic pre with single REV log pot to vary the gain over a wide range. Here is the result:

onetubemicpre.png
#

It is only a sim at the moment but it seems to work. It is basically a modded version of the EZ Tube Mixer mic pre stage with a single ended output instead of the SRPP and the gain of the amp can be taken right down to 0dB. With a 1:10 input transformer and 4:1 output transformer it will have up to 48dB of gain and drive +20dBu into a 600 ohm load with less than 1% distortion. Gain of the amp is set by R8 and can be varied from 0dB to 40dB so the pre can have a gain of 8dB to 48dB.

Cheers

Ian
 

bluebird

Well-known member
Joined
Jun 11, 2004
Messages
1,070
Location
Los Angeles
Is it the distortion or the amount of gain you have a problem with in "Ye Old" NYD one bottle design?

I wonder if you used a 12DW7 instead of the 6SN7 with the NYD design. 12AX7 half for the front end and 12AU7 half driving the output.
 

Attachments

  • Simple6SN7Preamp.png
    Simple6SN7Preamp.png
    5.6 KB · Views: 79

bluebird

Well-known member
Joined
Jun 11, 2004
Messages
1,070
Location
Los Angeles
bruno2000 said:
Nice design!  Love to try it.
Best,
Bruno2000

If you do, you might have to float the heaters up a little. I'm not sure where that first cathode is sitting DC wise but you don't want more than a 50v difference between the cathode and the heater on a 6DJ8.
 

ruffrecords

Well-known member
Joined
Nov 10, 2006
Messages
13,755
Location
Norfolk - UK
bluebird said:
Is it the distortion or the amount of gain you have a problem with in "Ye Old" NYD one bottle design?
It is the inability to smoothly vary the gain.
I wonder if you used a 12DW7 instead of the 6SN7 with the NYD design. 12AX7 half for the front end and 12AU7 half driving the output.
Not a bad idea. It would probably be about the same. Less gain and more distortion in the 12AU7 stage but more open loop gain due to the 12AX7. Pity it uses such an oddball tube.

Cheers

Ian
 

ruffrecords

Well-known member
Joined
Nov 10, 2006
Messages
13,755
Location
Norfolk - UK
bluebird said:
If you do, you might have to float the heaters up a little. I'm not sure where that first cathode is sitting DC wise but you don't want more than a 50v difference between the cathode and the heater on a 6DJ8.

Yes, the heaters probably need floating. The first cathode is at about 130V. Most 6DJ8s have one half with a 50V Vhk and the other about 150V IIRC so  by using the right halves you might get away with it - after all, they were designed for use in cascode circuits.

Cheers

Ian
 

bluebird

Well-known member
Joined
Jun 11, 2004
Messages
1,070
Location
Los Angeles
ruffrecords said:
Pity it uses such an oddball tube.

Looks like  JJ now makes them.
https://tubedepot.com/products/jj-ecc832-12dw7-preamp-vacuum-tube

Question: You seem to use the ECC88/6DJ8 a lot in your designs. I was playing around with some transformer-less mic pre tube designs a while back using vintage 6DJ8's I had lying around.  Most of them had noise problems or really low gain. Just very inconsistent. More so than any other tube type I've ever worked with.  Yes they were old but I had probably 20 of them and only 1 or 2 were usable.
Do you find this problem with the new ones you get? Do they wear worse or quicker than other types of tubes?
 

ruffrecords

Well-known member
Joined
Nov 10, 2006
Messages
13,755
Location
Norfolk - UK
bluebird said:
Looks like  JJ now makes them.
https://tubedepot.com/products/jj-ecc832-12dw7-preamp-vacuum-tube
Mmmmm, interesting!
Question: You seem to use the ECC88/6DJ8 a lot in your designs. I was playing around with some transformer-less mic pre tube designs a while back using vintage 6DJ8's I had lying around.  Most of them had noise problems or really low gain. Just very inconsistent. More so than any other tube type I've ever worked with.  Yes they were old but I had probably 20 of them and only 1 or 2 were usable.
Do you find this problem with the new ones you get? Do they wear worse or quicker than other types of tubes?
Although the schematic shows ECC88, what I mostly use is the 6922, mainly for its higher Vhk. Initially I used old tubes bought cheaply on eBay but I soon discovered they are very variable so I soon moved over to current production tubes. I still have a drawer full of old ECC88s and PCC88s that I will probably never use.

So nowadays I uses mostly Electro Harmonix 6922EH or JJ E88CC types. I have found inconsistencies in the JJ version but mostly in the distortion spectra. The most consistent type is the 6922EH.  At first I used them  just for output stages (for their low rp) but with the Classic I started using them for mic pres and I was pleasantly surprised to find their noise is consistently low. You get the occasional one that has poor micriphonics or higher noise than usual bit I have used well over 100 of them and only had problems with a couple.

Cheers

Ian
 

ruffrecords

Well-known member
Joined
Nov 10, 2006
Messages
13,755
Location
Norfolk - UK
emrr said:
Interesting Ian, can't say I've seen anything like it.

Thank you. It is inspired by the V76 which modifies the first stage gain as the NFB is varied to ensure the amount of NFB is reasonably constant and hence ensures stability. Its the same idea as the Transamp. In fact, initially the design was meant as a virtual earth amplifier.

Cheers

Ian
 

mjrippe

Well-known member
Joined
Dec 27, 2011
Messages
2,564
Location
Brooklyn, NY
ruffrecords said:
It is the inability to smoothly vary the gain.Not a bad idea. It would probably be about the same. Less gain and more distortion in the 12AU7 stage but more open loop gain due to the 12AX7. Pity it uses such an oddball tube.

Cheers

Ian

If it were just the number of tubes that was important, you could use a triple-triode Compactron like the 6U10 - it is a 12AU7 and one half of a 12AX7 in one package  :)  Talk about an oddball tube...
 

ruffrecords

Well-known member
Joined
Nov 10, 2006
Messages
13,755
Location
Norfolk - UK
mjrippe said:
If it were just the number of tubes that was important, you could use a triple-triode Compactron like the 6U10 - it is a 12AU7 and one half of a 12AX7 in one package  :)  Talk about an oddball tube...

Is that one of those that comes in a 12 pin envelope?

Cheers

Ian
 

ruffrecords

Well-known member
Joined
Nov 10, 2006
Messages
13,755
Location
Norfolk - UK
abbey road d enfer said:
10! Not that it makes things simpler...
Ian, can you post the .asc file? Either as a .txt here or on the LTspice group?

No problem. Do you want the ECC88 model I use or do you have one?

Cheers

Ian
 

Attachments

  • ECC88onetubemicpre.txt
    3.1 KB · Views: 17

ruffrecords

Well-known member
Joined
Nov 10, 2006
Messages
13,755
Location
Norfolk - UK
bluebird said:
Thanks Ian, If I go down that road again I'll know what to use.

You can also get Russian 6HN23P types very cheaply and these are a good equivalent to the 6922. They can be variable but they are cheap enough to be worth weeding out for experimentation purposes anyway.

Cheers

Ian

Edit: well they used to be cheap. I just checked on eBay and people seem to want silly prices for them now.
 

ruffrecords

Well-known member
Joined
Nov 10, 2006
Messages
13,755
Location
Norfolk - UK
abbey road d enfer said:
In fact it goes down to -0.5dB since the OLG is not infinite. Well done Ian!
Now the hunt may begin for a suitable RevLog pot... :eek:

Thank you.

47K REV LOG  is fairly common. That + 390R will range down to 6dB. One with a switch could cover the last 6dB

The reason the feedback resistor is 47K is that I originally designed it as a tube VE mixer with 47K feed resistors. I am toying with the idea of reducing the 47K feedback resistor to 22K or so. Cannot go too low else the output gets loaded and distortion rises but the noise might be a bit less at lower gains and I might be able to raise the second plate a little to extend the headroom. Usual trade offs. With a 22K, a 47K REV LOG gets close to 3dB.

Cheers

Ian
 

abbey road d enfer

Well-known member
Moderator
Joined
Jan 22, 2008
Messages
12,996
Location
Marcelland
ruffrecords said:
Thank you.

47K REV LOG  is fairly common. That + 390R will range down to 6dB. One with a switch could cover the last 6dB
Typically for a 34 dB gain range, the ideal taper should be 7%. Most commonly available RevLog are 15 or 20%. They make the CW part of the rotation quite clustered.
With 15% taper, you have 12dB on the left and 22 on the right.
You have to place a special order for the more desirable 10 or 5% types, unless you can buy spares from a manufacturer.
 
Top