How do I calculate current on the primary for spec'ing a gapped OT? (RCA BA-2C)

[Shattersignal]

This has nothing to do with voltages, everything to do with transformers and circuit design. 
...

Very interesting!  I really like what I hear so far, so I'm definitely not disappointed by the RCA   It sounds very very different from the Redd, which is a good thing, to be sure.  The sound is much more harmonically rich--owing to the no feedback topology, I'm sure.  It's a very exciting sound!  It's kind of like the difference between a 70s Fender and a Vox.

Doug, do you have a suggestion for an excellent input transformer for the build?  I am building two with A-10s on the input and one for a friend with a CM-75101a (which I haven't built yet, and haven't heard yet).  I don't mind pulling the UTCs and using them in DI boxes if there might be a better option out there for my stereo pair. 

One thing I might like to do with a pair of the BA-2Cs is re-processing stereo stems, and for processing program material.  I also imagine them very useful on drums and drum machines, as well, where low end extension and definition is of a particular interest. 

 

[Shattersignal]

UTC A-10; officially spec'd to be 20hz-20khz is +/-2dB.  LS-10 is +/-1dB.  This is significant when you add input and output curves together.

Thanks for the graph!  Yes, that plot looks very similar to what I'd expect from an A-10, as depicted in the UTC catalogs.  I wouldn't mind eventually dropping the coin on some expensive iron, if there's a specific recommendation anyone has.  I love Cinemag, generally.  I've also used Sowter many times and I've been extremely impressed.

I opted to try the UTCs since I happen to have a closely matched pair of early black-can metal-badged A-10s and I always though it would be fun to build some RCA amps around them.  But I won't be heartbroken if it tuns out that another transformer might fit my needs better. 

 

[Shattersignal]

I now have the two tube plates near the correct voltage, and in the correct voltage ratio.  The bandwidth issue has completely disappeared.  THIS MIC PRE SOUNDS AMAZING.  It's everything I wanted in a classic American tube amp design.

Please disregard my comments about a lacking low end--the low end is deep and full, and also has a nice *subtle* second harmonic sound that adds heft and excitement from the bottom octave all the way through the midrange.  I am EXTREMELY PLEASED with how this has turned out, and I imagine that it will sound killer on drums and all manner of transient-rich low frequency content, including on program material and even on a finished mix, with the proper padding applied. 

Thanks to all who have helped me!! 

 

[Shattersignal]

I'm installing DI jacks to the v1 grids in these amps.  The leak R will be 1M, in accordance with Fender amp custom.  What value should my grid stoppers be?  I gather that this value is determined by the spec of the tube.  The microphone input path doesn't even use a grid stopper--do I need to use one on the DI path?  Furthermore, could such a resistor be useful to prevent loading on the microphone path, when the switched DI jack is not in use, and therefore is shunted to ground?  I wonder if there is a special use for such a resistor when both and microphone input and a DI are sent to the same grid.  I currently have it wired up without a stopper R on the DI path and I'm having difficulty sorting out a grounding scheme that benefits both inputs while also not grounding out the microphone input path.   

If it would be useful, can anyone please guide me on how to determine the proper grid stopper value for these DIs?  The 6j7 data sheet indicates an input capacitance of 7uuf.  Is that the same as Miller Capacitance?  Another factor seems to be the tube's stated maximum series grid resistance.  I don't see that on the 6j7 data sheet.  I'm sure I'm overthinking things, but I'm looking to understand the principles involved here.  Has anyone installed direct-to-grid DIs in their RCA 6j7 or 1620 amps?  What value did you use and why, please?

Also, for a "stereo pair" of these amps, should I be concerned to carefully match either the leak or the stopper resistors between the two channels?  It's not clear to me how vital that might be in order to closely match the sound of the two DIs.  I realize that there is a limit to how closely matched any two tube amps can be.

I've familiarized myself with CJ's excellent post regarding grid stoppers in general.  Thanks, CJ!
https://groupdiy.com/index.php?topic=59139.msg751343#msg751343

I also see EMRR's post on the same thread, which might indicate that a grid stopper isn't needed in this amp:
https://groupdiy.com/index.php?topic=59139.msg751467#msg751467

 

[emrr]

No grid stopper needed

Tube matching in circuit and identical transformers is what you get for stereo matching in a non-feedback amp, assuming all resistors are within 5%. 

 

[Shattersignal]

No grid stopper needed

Tube matching in circuit and identical transformers is what you get for stereo matching in a non-feedback amp, assuming all resistors are within 5%.

Excellent.  Thanks!  I'll put my attention into the tubes when dialing in the stereo pair.

I still need to figure out what's up with my DI circuit.  It keeps injecting noise when the DI jack isn't in use, and if I get that noise to ground out, the mic input is also sent to ground.  I'll need to spend some time with the logic of my switching jack.  The funny thing is, it's wired identically to the DI I put in my Redd47, which works perfectly.  The only difference it that the Redd uses a grid stopper between the DI and v1 grid.  Could the grid stopper be the missing ingredient when a DI path and a microphone input path both enter the same grid??  A grid stopper on the DI path would prevent the entire signal from dumping to ground when the jack is silenced.  Anyone have thoughts on the best way to wire a switching DI 1/4" jack in a tube mic preamp?

 

[Shattersignal]

...Double check V2 cathode R, and bypass cap.  Break the path and insert a current meter, see what you get.

In order to test current flow to the cathode, do I need to break BOTH the R and the bypass cap?  Or just the R?  I suppose once the cap has drained, it won't affect current flow, either way?

I'm studying up on cathode bypass cap functionality right now--fascinating stuff.  The cap acts as a HPF for a sort of local NFB network, and therefore higher cap values will increase low frequency harmonic content in the amp.  Am I right in assuming, therefore, that higher value cathode bypass caps can increase low frequency content of the amp, albeit at the expense of linearity?  Or do I have it backwards--i.e. higher value INCREASES NFB?  I'm not clear on this point.

Has anyone experimented with higher/different value cathode bypass caps in the RCA circuit?  I see references to folks upping the cathode bypass capacitance in the Redd47 circuit, with desirable results--which I'm planning on trying in my Redd47.  Is this worth trying in the BA-2C?  I just ordered nicer Nichicon Muse KZ caps for the RCA, which will replace the ones I currently have installed.  I also used Amtrans AMRG resistors on the cathode, which seems to have been a good choice, from what I gather, based on the assumption that this resistor actually has a significant bearing on the sound of the amp.  Anyone have thoughts on how much these parts actually contribute to the sound of the amp, given the same parts values?  How about different value caps?

EDIT: I'm using 47uf cathode bypass caps on both tubes.  Schemo indicates 40uf, which I interpret to mean between 40uf and 80uf, given 1930s cap value tolerances.  That seems like a fairly wide range of values that would have a significant bearing on the tone of the amp.

 

[emrr]

Bypass cap size has a practical limit.  2x-5x larger is a good range of experimentation.  sometimes the transformers are already dictating.  Sometimes larger caps yield less low end, and the original value is a specific EQ recipe.  Really large caps increase the amount of leakage DC I understand, which has it's own effect.  Break only the R for current. 

 

[Matt C]

I'm studying up on cathode bypass cap functionality right now--fascinating stuff.  The cap acts as a HPF for a sort of local NFB network, and therefore higher cap values will increase low frequency harmonic content in the amp.  Am I right in assuming, therefore, that higher value cathode bypass caps can increase low frequency content of the amp, albeit at the expense of linearity?  Or do I have it backwards--i.e. higher value INCREASES NFB?  I'm not clear on this point.

bypassing the cathode increases gain by decreasing NFB. The capacitor and the cathode resistor form a filter that decides for which frequencies the cathode is bypassed.  Usually for clean pro-audio type stuff like this the cathode is bypassed for the entire audible frequency range.  Reducing the value of the cap would raise the filter's corner frequency. Frequencies above that would still see a bypassed cathode and have high gain.  Frequencies below that would not see a bypassed cathode, which means you get NFB at those frequencies, meaning less gain.

 

[Shattersignal]

Bypass cap size has a practical limit.  2x-5x larger is a good range of experimentation.  sometimes the transformers are already dictating.  Sometimes larger caps yield less low end, and the original value is a specific EQ recipe.  Really large caps increase the amount of leakage DC I understand, which has it's own effect.  Break only the R for current.

Excellent knowledge, here!  Thanks again, Doug!!

 

[Shattersignal]

bypassing the cathode increases gain by decreasing NFB. The capacitor and the cathode resistor form a filter that decides for which frequencies the cathode is bypassed.  Usually for clean pro-audio type stuff like this the cathode is bypassed for the entire audible frequency range.  Reducing the value of the cap would raise the filter's corner frequency. Frequencies above that would still see a bypassed cathode and have high gain.  Frequencies below that would not see a bypassed cathode, which means you get NFB at those frequencies, meaning less gain.

Thanks, Matt!  That's what I thought.  I'll run some calculations and see if there might be a benefit to increasing the bypass caps, i.e. if there's a chance that I can increase the bandwidth.  Thanks!!

 

[CJ]

be sure to post your final schematic so we can try and duplicate the results! 

 

[Shattersignal]

be sure to post your final schematic so we can try and duplicate the results! 

Sure thing

 

[Shattersignal]

I'm planning on making some audio demos, too.

 

[emrr]

Don't remember, did we mention having an output load resistor will cause a greater degree of response and gain change than in almost anything else, because of the lack of NFB?  Factor it.  Choices to make. 

 

[Shattersignal]

Don't remember, did we mention having an output load resistor will cause a greater degree of response and gain change than in almost anything else, because of the lack of NFB?  Factor it.  Choices to make.

nice!  No, I haven't tried that.  You mean a resistive 600ohm termination across the output?

 

[emrr]

right

 

[Shattersignal]

Yes, there seems to be some ringing in the OT.  After adding the termination, the level drops slightly (maybe 2dB?) and the tone darkens and maybe fills in a little bit in the low end.  I think I prefer the un-terminated sound right now--it seems to give a little more 10k+ and I enjoy the extra sizzle (i.e. subtle distortion!).  It definitely sound more different than the Redd47 when lightly loaded, that's for sure, and I think that's a large part of what I'm really enjoying about it right now.  I have some Neutrik XLR barrels with 800ohm resistor terminations already made up, so I think I'll leave the units un-terminated for the time being and use the barrels when I want a little more control of the high end.

Great suggestion, EMRR!

FWIW, I'm tracking the BA-2C straight into my Lavry Black AD10, which has an extremely high input impedance, i.e. extremely light loading.  I LOVE the way it sounds with my Manley Ref. C on vocals.  It's maybe one of the best vocal sounds I've ever heard.  The analytic quality of that mic is greatly enhanced by the rich harmonics of the preamp.  Bass guitar and synthesizer DI, same thing--some of the best I've ever heard.  I can even DI my Jazzmaster and it sounds like a close-mic'd clean Fender amp--a truly incredible sound that's extremely detailed but still rich with even harmonics.  It strikes a unique balance between placing the instrument forward, while also retaining a realistic sense of space.  My Neves 1073s don't do that--they tend to bring the instrument forward, while smearing the sense of space and flattening everything into one plane.  The Redd47 has TONS of depth and holographic spaciousness, but doesn't necessarily bring the instrument forward--also an excellent quality, but definitely different than the RCA. 

 

[Shattersignal]

Next question for the RCA BA Gurus: how do these amps sound with other brands of tubes?  I see 6j7 tubes from GE and others, besides RCA.  Who has tried other brands in these RCA mic preamps and what did you think about the change in sound?

In my experience, the GE house sound is very nice, tending to have a tiny bit more "air" and a little less gooey midrange enhancement, when compared against RCA tubes in the same circuit.  I tend to prefer RCAs in my guitar amps, but I've enjoyed GE tubes in my LDC microphones, because they seem to offer a little more detail at the expense of flattery.  Does this description carry over to the 6j7 type tubes?  It wouldn't be too expensive to just buy a few and try them out, but I thought I might benefit from a first-hand account, if there's anyone here who's already experimented with different tubes in these mic preamps and who has thoughts to share.

Thanks again!

 

[emrr]

I plug in whatever I have, and observe noise after burn in.  Rarely do I take  a variety and compare.  But, I'm also running more than a dozen channels, so not much space for cork sniffing, it's impractical with scale.  I don't hear obvious differences with all channels running on the same tracking sesdion.  1st stage tube is all you should worry about, 2nd hardly matters so long as it works correctly.