600:600 output trx

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
kambo said:
Ra 5k Rk 890r ,  bypassed...
i have another 6sn7 in front as micpre...

OK.  Your bottom triode is getting way more signal with a 5K load on the top valve.  If you like it, more power to ya 😊
I would plop a 385 ohm resistor at the top myself.  Shooting from the hip and an adled & dodgy memory, I'd also start with a quiescent current from about a 400 ohm-ish cathode resistor on a 6SN7.   
But each to each my friend 🙂
 
Winston O'Boogie said:
OK.  Your bottom triode is getting way more signal with a 5K load on the top valve.  If you like it, more power to ya 😊
I would plop a 385 ohm resistor at the top myself.  Shooting from the hip and an adled & dodgy memory, I'd also start with a quiescent current from about a 400 ohm-ish cathode resistor on a 6SN7.   
But each to each my friend 🙂

i actually started with 400 ohm.  but didnt like the sound at all!
i usually record in to same cubase file, and have instant access to my other micpre takes,,. ( ie: miked speakers playing drums, music,
and some talking etc! )

this setting sounding really good to my ears. very fat and very lo distortion! around 8mA is more hairy sounding!
btw, i have something like 130-150vdc(need to measure) coming from first stage to wcf input!

edit: top and bottom is 3.4mA 100th digit is only 1 off... 3.45 top - 3.44 bottom

 
kambo said:
i actually started with 400 ohm.  but didnt like the sound at all!
i usually record in to same cubase file, and have instant access to my other micpre takes,,. ( ie: miked speakers playing drums, music,
and some talking etc! )

this setting sounding really good to my ears. very fat and very lo distortion! around 8mA is more hairy sounding!
btw, i have something like 130-150vdc(need to measure) coming from first stage to wcf input!

edit: top and bottom is 3.4mA 100th digit is only 1 off... 3.45 top - 3.44 bottom

Like I said, each to each ☺️.    Both triodes will indeed be DC biased the same, the balance required is one of signal.  You have something rather more a push-push than push-pull stage as is. 
With a 2:1 transformer loaded with 600 on the secondary, we have about a 2K4 primary plus dcr.  I don't see how an imbalanced White with 3.4mA could drive that and be happy myself which was why I suggested going up in current a bit, but if you like it 👍
Peace.
 
ohh... i am not driving 600 ohm, nor trying to, i have no such equipment either!
this extra stage is about mic pre sound!
i recorded square/sine wave at full gain too. looking exactly the same as original! i c no problem! 
i can load it with 600 ohm in the morning, probably what u say would happen  ;D


btw: ltspice simulation and realmodel is not working the same  ... i only get 70-80vdc on top grid in  to wcf - not 140-170 as ltscpice would tell u!
 
schematic

edit: u2 - u4 grid 82vdc

edit 2: that gain pot section has leftover wire/res  from ltspice ... its 100k pot wiper to U2

edit 3  :  new schematic uploaded  : readings from bench, not ltscpice

edit 4 : might as well add feedback for extra gain  ;)
 

Attachments

  • 6sn7 rev-1.png
    6sn7 rev-1.png
    49.5 KB · Views: 19
Winston O'Boogie said:
So:

Ra = rp/mu.

Ra for a 12BH7 is 5100.  Mu is 16.5.

So Ra = 310 ohms.

Unfortunately those values of mu and Ra are true only at one operating point. They are not fixed or constant values. At lower plate voltages (such as in a White) u increases. Ra is very dependent on standing current. So exactly what is the design standing current in the Jensen design? Here is a datasheet for the 12BH7:

http://www.r-type.org/pdfs/12bh7.pdf

Are we sure Jensen got it wrong?

Cheers

Ian
 
NYD line amp is it at 17mA too ...
edit : Jensen might get it wrong but, NYD wouldnt  ;D
 
ruffrecords said:
Unfortunately those values of mu and Ra are true only at one operating point. They are not fixed or constant values. At lower plate voltages (such as in a White) u increases. Ra is very dependent on standing current. So exactly what is the design standing current in the Jensen design? Here is a datasheet for the 12BH7:

http://www.r-type.org/pdfs/12bh7.pdf

Are we sure Jensen got it wrong?

Cheers

Ian

Unless I completely lost my marbles again 😜  I'm sure Jensen got it wrong.  Not as wrong as some have and, I admit, it's not as complete a disaster as elsewhere,  but it's off.  Plug the values into the equations I gave from any point on the 12BH7 datesheet curves and it will hover around the value I gave. 

Edit: Having said that, anyone following this, I'm interested in how other folks determine the optimum value for Ra.  We can do it by building, measuring and adjusting but, we ought to be able to get in the ballpark just as we can with other parameters.  I'm open to
hearing suggestions for rules of thumb and napkin scribbles.


 
To avoid confusion, the pdf attached shows the values I chose for Ra based on 1/gm or Rp/u.
PRR would be the person to know this stuff backwards.  Hopefully he can advise.

Otherwise, that's ma story and ah'm-a stickin' to it until ah knows diff 'runt. 
 

Attachments

  • 1:gm Rp:u Followers.pdf
    133.7 KB · Views: 12
Winston O'Boogie said:
Unless I completely lost my marbles again 😜  I'm sure Jensen got it wrong.  Not as wrong as some have and, I admit, it's not as complete a disaster as elsewhere,  but it's off.  Plug the values into the equations I gave from any point on the 12BH7 datesheet curves and it will hover around the value I gave. 

I can understand setting the plate resistor at 1/gm simply because the stage gain is gm.Ra which will be unity (as required) if Ra = 1/gm.

What I do not understand is your contention that one value will work for most anywhere on the 12BH7 curves. gm varies significantly with plate current which means Ra depends critically on standing current.

Cheers

ian
 
ruffrecords said:
I can understand setting the plate resistor at 1/gm simply because the stage gain is gm.Ra which will be unity (as required) if Ra = 1/gm.

What I do not understand is your contention that one value will work for most anywhere on the 12BH7 curves. gm varies significantly with plate current which means Ra depends critically on standing current.

Cheers

ian

Ah I see.  Yes I did say that, which is not correct.  Sorry.  Had been up all night... have had a nap now ☺️
I meant, we have a given point on the curves which we have defined based on our standing current, hence gm is reasonably known.
We set our value for Ra based on this and, within reasonable boundaries around this point, where Rp for the 12BH7 may be 5000 or 5300, and Mu may be 17.5 or 16.5, then the Ra will be quite reasonably within a few percentage points of being exact.

So, given the operating point of the 12BH7 as used in the Jensen schematic, I say that my value of 310 ohms is a better choice than the 680 ohms they chose.  There is nowhere around that point at which the reciprocal of gm is approaching 680. 
At low anode voltage use (as defined in the datasheet you provided), the 1/gm is lower, in which case Jensen's Ra value is even further off.

Again, sorry for the sloppy and incorrect statement.  My intentions were in the right place, brain not so much!




 
Winston O'Boogie said:
Ah I see.  Yes I did say that, which is not correct.  Sorry.  Had been up all night... have had a nap now ☺️
I meant, we have a given point on the curves which we have defined based on our standing current, hence gm is reasonably known.
We set our value for Ra based on this and, within reasonable boundaries around this point, where Rp for the 12BH7 may be 5000 or 5300, and Mu may be 17.5 or 16.5, then the Ra will be quite reasonably within a few percentage points of being exact.

So, given the operating point of the 12BH7 as used in the Jensen schematic, I say that my value of 310 ohms is a better choice than the 680 ohms they chose.  There is nowhere around that point at which the reciprocal of gm is approaching 680. 
At low anode voltage use (as defined in the datasheet you provided), the 1/gm is lower, in which case Jensen's Ra value is even further off.

Again, sorry for the sloppy and incorrect statement.  My intentions were in the right place, brain not so much!

OK, understood. So I drew a load line from plate volts =0 and Ia =30mA through the -5V Vg and Ia = 15mA point. From load line I estimate a 10V swing in Vg gives a 16mA swing in Ia which gives gm=1.6mA/V so Ra = 625 ohms. Have I gone wrong somewhere?

Cheers

Ian

Edit: I used the data sheet found here:

https://www.alldatasheet.com/datasheet-pdf/pdf/230561/GEC/12BH7.html
 
ruffrecords said:
k
From load line I estimate a 10V swing in Vg gives a 16mA swing in Ia which gives gm=1.6mA/V so Ra = 625 ohms. Have I gone wrong somewhere?

Cheers

Ian

Edit: I used the data sheet found here:

https://www.alldatasheet.com/datasheet-pdf/pdf/230561/GEC/12BH7.html

Thanks Ian.  I'll check the datasheet in a while, not able to print and no computer at the moment so it's a little hard on a small screen but I'll give it a go.
There's every possibility you have NOT gone wrong.  Could very well be me who is wrong.  I've had to try and relearn stuff I knew quite well a decade+ ago.  Not using information for a long while, along with seizures and brain damage catches me out sometimes 😉

Again, PRR would be the one to nail this.

Back at ya as soon as...
Cheers.
 
Winston O'Boogie said:
Thanks Ian.  I'll check the datasheet in a while, not able to print and no computer at the moment so it's a little hard on a small screen but I'll give it a go.
There's every possibility you have NOT gone wrong.  Could very well be me who is wrong.  I've had to try and relearn stuff I knew quite well a decade+ ago.  Not using information for a long while, along with seizures and brain damage catches me out sometimes 😉

Again, PRR would be the one to nail this.

Back at ya as soon as...
Cheers.

No problem mate. Just trying to help. Take it easy and get back when you can.

Cheers

Ian
 
so i loaded with 600 ohm, it was terrible as u guessed!
with 10k its awesome sounding pre, THD 0.03%

with Ra 400 ohm and 100ohm Rk its set at 8.4 mA and with  600 ohm it doesnt even sweat :)
different sound. THD 0.3%
(all at 20dbu max gain )
i think i am gonna have a switch somewhere in front/back for 600/10k for different sounding option
 
Back
Top