Cascode tube mic input with 12BH7 output driver

Shop deals on ebay
shop deals on amazon
Advertise with us
GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

chrispsound

Well-known member
Joined
Sep 30, 2010
Messages
159
Location
Sacramento
http://chrispsound.com/wp-content/uploads/2011/10/IMAGE0002.jpg

Just finished a NY Dave Mila and would like to tackle this pre from Jensen's Schematics AS054 and AS021.  I have a question about R9, I assume it is for feedback, but is it essential?  If omitted can the same output transformer(or similar) be used.  If it is essential, where would I incorporate it back into the circuit and would R9's value need to be changed?  Thanks, ChrisP
 
Hi Chrispsound.
R9 is for feedback. Is it essential? Well take a look at this thread where NYD expose his first try with it's line output gain using approximately same wcf out stage:

www.groupdiy.com/index.php?topic=724.0

The big differences with your idea is: the second stage of your schematic is a srpp design and in NYD this is different. In place it use 2 common cathode 12AV7 amplifier series (1 whole tube) with feedback (R9) returning to cathode of first 12AV7. Other meaningfull  differences:

. is the cathode resistor of the low half of WCF which is 220r in place of 330r (giving +3,2V in place of 5V);
. grid stoppers on 12bh7 are 1000r (on both grids)
. C4 (on your schematic) is 0.22uf in NYD it's 0.1uf
. C5 (on your schematic) is 22uf, in NYD it's 10uf
. R9 ("    "      "            ) is 30k , in NYD it's 56K and decoupled from 12BH7 high triode cathode/low plate triode junction using a 1uf cap.

Main difference is the srpp stage for my limited knowledge. This is not the place i'll use this kind of stage, i prefer the idea of NYD. Why? Because i don't know behavior of srpp in feedback loop and this is something i've never encountered in schematics i know.

Well try a search on froum about NYD tube output line driver for schematic as it's missing from original thread, probably someone as it in here...

>For the out transformer, use a 2:1 and it should be ok.
(by the way you have mixed the cascode design with the mod Jensen recommended for the LA2A out  isn't it?) 
 
Thanks for responding (I was feeling lonely and unloved), I actually have the NYD line amp schematic deep in my download folder.  So I will insert feedback with 1.0uf decoupling capacitor above the 12AU7's (a) 1k2 cathode resistor.  I was wondering if a 50K or 25K  2watt linear or audio pot could replace  feedback resistor R9 for testing and/or flavor.  The output was taken from the JT-10K61-1M datasheet.  I almost have this built on turret board so I guess there is only one way to figure out if this thing works (the fun way!).  Any other tips or recommendations from smarter(which isn't hard, I took Algebra 2 twice because that's what I thought the 2 stood for) and more experienced members is always welcomed.  Help from the mathematically and conceptually illiterate is welcomed as well.  Thanks ChrisP
 
IMG_3352.jpg


Sounds great, easy and relatively cheap to build with a low parts count.  The Jensen input xformer is amazing, when compared with another with the same windings there is no comparison.  Definitely recommend this one especially since most tube diy'ers have these tubes and most parts on hand.  I will take it out of the test box after some use and give it a proper home and power supply after putting it through its paces.  Thanks Kr for your sole response.  ChrisP
 
Looks good, first time I've seen anyone actually tackle this circuit.  Look forward to some bench tests, comparison of FB, etc. 
 
I have the free version of rightmark audio analyzer, a software scope, a DAW, a multimeter and Sonar.  How would I go about getting bench tests and comparison's of FB etc.?  I usually just hook up a 57, then a AT4047(w/external phantom) and a P bass DI'd, to see if it passes decent audio.  Then I throw it into rotation during recording sessions(haven't done this yet with this one) and if a producer(aka drunk/ stoned friend) likes it and wants to use it more than once I assume its worth moving into a permanent box and making a faceplate for it.  So far everyone in my studio likes Colins EZ1073 over anything else I got (1. Classic API VP26 and VP312, 2. NYD MILA, 3. Bruno's SSL 9K,  4. Great River MP-500NV 1 being the second most popular).  Thanks  ChrisP
 
chrispsound said:
I am doing it with no feedback, almost done, will post pics and results.
Click to expand...
The overall actual gain must be around 80-90dB (including 20dB from the input tx).
You need to control gain in a different way than just attenuating the input signal, which increases operational noise to an almost unusable level..
 
KrIVIUM2323 said:
Main difference is the srpp stage for my limited knowledge. This is not the place i'll use this kind of stage, i prefer the idea of NYD. Why? Because i don't know behavior of srpp in feedback loop
Click to expand...
Why would you think the WCF would have an unpredictable influence? The WCF has a well-defined and well-behaved phase response, in fact probably better than any single-ended stage that no one would hesitate to include in a NFB loop.
and this is something i've never encountered in schematics i know.
Click to expand...
Probably because it's a tad over-engineered.
>For the out transformer, use a 2:1 and it should be ok.
Click to expand...
This is probably ok if the intended load impedance is a modern 10k line input, which would reflect at 40k to the output stage, but if you want to have some extra output capability, you may want to have a lower ratio, like a 4:1. And anyway you don't need the +30dBu output that comes with a 2:1.
 
I am using an Edcor XSM15K/600 5:1 output xformer(I figured I may need extra attenuation).  I need to double check my work because I was just thinking the gain was perfect and the noise is relatively low,  lower than the noise from some of my other pre's.  I probably need to test the pre or record some talent with it and then post the results.  Any quick and easy tests anyone know of that can produce semi-shady results.  The right angle tube sockets were a local surplus electronics store find.  Any recommendations on how to tame volume other than input attenuation(I have a thick skull thats slowly receding)?  Thanks  ChrisP
 
abbey road d enfer said:
chrispsound said:
Any recommendations on how to tame volume other than input attenuation
Click to expand...
NFB
Click to expand...
The original Jensen schematic as054 and the output driver schematic from the JT-10K61-1M that it is DC coupled too shows no NFB provisions so I thought to look at some examples of NFB that has been used in a similar circuit but could not find any after googling.  I will probably try sending feedback willy nilly but I want to use it like the schematics intended before I start modding(especially since I have little knowledge of properly implementing NFB).  Any other tips for the newb on how to analyze basic functions of a freshly built pre would be nice.  ChrisP
 
abbey road d enfer said:
chrispsound said:
I am doing it with no feedback, almost done, will post pics and results.
Click to expand...
The overall actual gain must be around 80-90dB (including 20dB from the input tx).
You need to control gain in a different way than just attenuating the input signal, which increases operational noise to an almost unusable level..
Click to expand...
You are right, I am an idiot.  I unintentionally put 47k resistors were the 1M to grounds go.  Also my power supply is not working right,  the trimpot is very sensitive so I was probably only getting 180V(but I can't be sure), If I adjust it just slightly it jumps to 260 with no adjustment in between.  It works quiet enough with my mistakes and passes some decent audio, albeit slightly distorted.  When I get everything to how it should be, I basically have a noise box with a lot of gain.  Oh well, back to the drawing board.  ChrisP
 
Congrat Chrispsound !

Why would you think the WCF would have an unpredictable influence?
Click to expand...

Hello Abbeyroad, i am not concerned about the WCF but the SRPP.

Mainly because SRPP must be calculated for input Z of next stage for best behavior and as the wcf don't have grid resistor and as i don't know how to define input Z for this stage... But it doesn't seem to be a problem as Chrisp seems to like it this way..
 

Similar threads

C
Passive DI / Reamp Box design queries
Replies
5
Views
3K
crduff
C
Duizels
Making some mics - Beginner suggestions
Replies
12
Views
2K
Duizels
Duizels
C
MILA-1 with 6CG7's Help
Replies
16
Views
6K
dirtyhanfri
dirtyhanfri
Matador
Two-Tube 'MILA-like' 6SN7 mic pre thoughts
2 3
Replies
52
Views
23K
johnheath
johnheath

Latest posts

Back
Top