Another RCA BA-2 Build

[dmp]

Did a few more tests. First, flipped the secondary on the input transformer. Similar response but a little worse when flipped.
(green is flipped)

 

[dmp]

Then put 75 ohm resistors on the input in series with the source - something I read in another thread to appx a mic's source impedance.
This changed the response a lot on both the high end and the low end.
(Whit is without source resistors, green is with)

 

[dmp]

And finally, with a pad, a high end bump like emrr expected. I used 1100/220/1100 since I had the values on hand.
I'm finishing up the front panel now and final wiring.

 

[emrr]

So the sec as was is best capacitance to ground relationship. 

Series R alone is usually a bad thing for commonly encountered equipment, most sound cards and medium priced converters seem to be fairly close to mic Z anyway, only the minority is truly lower (like 0-50 R), and benefit from series R alone. 

Pad as expected.  Now if you built those series R out to 40 dB line pad range you might find what I see here, increasing high end loss.  I suspect the loss of power through the higher series R is more noticeable in the highs, but I can't back that up outside of common field observation.  You don't see similar loss connecting to a modern transformerless input. 

 

[beatnik]

hey dmp just a question what software are you using to make those frequency response graphics? thank you

 

[dmp]

what software are you using to make those frequency response graphics?

It's called RMAA

 

[dmp]

Here are final pictures of my build

 

[dmp]

And the back

 

[dmp]

The insides

 

[letterbeacon]

Looks great!  I really like the lone knob on the front, with I had left off the attenuator on mine.

 

[buildafriend]

I second that, I call things like that "the go knob." Simplicity is beautiful.

It looks great, great work!

 

[dmp]

Thanks guys!
This preamp really sounds great. I played recordings of acoustic guitar from this BA-2, my 1290 clone, and a Jensen 990, and in a blind test a colleague found it noticeably better - more real.
I'm wondering if I should have put in a instrument DI. Opinions?

 

[emrr]

Easiest DI ever.  Pop off the grid cap and have another grid cap on hand with a 1/4" jack and an alligator clip for grounding.

 

[buildafriend]

Has anyone had any luck with using any modern OP xfrmrs? I just ordered a CMMI-10C and I want to stick with new iron for the sake of making sure that my unit has a nice long life ahead of it. It's also nice to be closer to sure that what I get will be to spec.

 

[letterbeacon]

I believe 'gary o' has built one using modern Sowter transformers...

 

[buildafriend]

Thanks!
http://www.groupdiy.com/index.php?topic=5306.msg63842#msg63842

1:10 step up on the input, it's a sower 9045

And his output iron is a sowter push pull 9050.

Useable inputs are:
CMMI 10C
Sowter 9045
...?

Useable outputs are:
Sowter 9050
Would any cinemags fit the output application well?
...?

I wonder if UTC A-11 and A-25 would work. I know that they can be used on the BC-2B..

Push-Pull..  now I have to do some homework.

I would sincerely appreciate it if you gurus helped me with a few questions here, before I end up in a hole while building this.

Push pull.. this means that the PSU go's to the op iron CT? That's bizarre. Why?

Is the BA-2 originally a push pull circuit? Would I have to alter anything in my BA-2 to use a push pull OP xfrmr? It does not look like a push pull circuit when I look at the schemo, unless there is something coming from that switch at the bottom that I do not understand.

Schematic:

http://i98.photobucket.com/albums/l265/studiodave/ba2c13.gif

 

[dmp]

It is definitely not push-pull. Output stages with two tubes following a phase inverter is a  push pull circuit.
( one output tube is driven by the signal and the other by the inverted signal - one pushes / one pulls, allowing both tubes to be biased lower since they don't have to fully swing both ways).

These RCA circuits are single ended and are called transformer coupled. In principle, the plate current for both push pull or SE designs go through the transformer, eliminating the need for a DC blocking cap on the output, as in a resistor coupled stage.

The transformer gurus can ID what DC current spec the output transformer needs to handle. I believe EMRR posted a value based on either a spec or measurement. Check the other BA threads.

In other news I've compared this BA2 preamp to several other preamps and it is the king of the hill. It is particularly good on acoustic instruments - makes them sound a little more real! I brought it to a studio last night and compared it to a Neve and a Jensen 990 preamp with several mics. Big differences to hear. I was surprised how the RCA BA2 made tube condensors sound a little bit better. The only mic that we preferred the 990 with over the BA2 was with a RCA 44. The only thing I would want is to have a little more gain available for some mics/sources. Would it be as simple as to add a third tube (like the first stage) in between the first and second to allow the gain to go up to 70dB? Or would that raise the noise floor ~20dB?

 

[emrr]

No free lunch on noise.  Lassoharp essentially built a 3 stage, it's a little different but worth looking at for ideas.  I'm always tracking with safety compression, and use gain in the compressor if more is needed. 

 

[dmp]

So, a question for you guys with more experience with tube preamps (somewhat off topic).
This BA2 has a sound I really like - transparent, real, etc...
It is WAY preferable to the other tube pre I built, NYD's one bottle (JE115, 12av7 w/ adj nfb)

I want to build another tube preamp - but wonder what is the most important factor(s) to keep to have the same great sound? Is it the 1620 tubes? No nfb design? The input Tx? Or the transformer coupled output / Tx? Advice?

Ian's tube mixer project looks pretty awesome and is a no nfb design. But would it have the sound qualities I'm after using a 12ax7 & 6922 w/ the cap coupled output? I also have a LS10 / LS50 that I want to build a preamp around but the LS50 would not be a transformer coupled design either...
Seems this might be crucial looking at other 'respected' designs... UA610, v76, v72...
Any help appreciated.

 

[emrr]

The iron is the biggest part of the sound.  I had a BA-2 once that had EF86's in it, no other parts changes at all, just new sockets dropped in.  I can't split hairs on the sound differences other than to say it wasn't the biggest part of the picture by any stretch. 

No NFB design is a big part of it, after the iron sound. 

The 2 versions of Gates SA-70 / SA-70A, difference is transformer types, sound is clearly different.  The Gates/WE mashup I did sounded even different, predominately in lacking the output transformer. 

LS10/LS50 is a great platform for doing a no feedback build.  You can try the LS10 easily with clip leads in the BA-2 build for comparison.