WE 407A Push-Pull Mic Preamp

[Larrchild]

Schematic

Picture of what we are talking About
Feedback is from a second (tertiary) winding. So Far=)

Whaddya say? Chokes in the 1st stage too? :razz:

 

[rafafredd]

Choke in the first stage? Wouldn´t it qualify as tottaly overkill? Well if you can do it, just let us know how it goes...

Just had an idea... What if you used just one balanced choke, so you don´t need a gap? Would it work?

 

[Larrchild]

What if you used just one balanced choke, so you don´t need a gap? Would it work?

Winston suggested something I tried a while back, which is use a cheaper CT transformer as a paraparafeed chokes (PP) and dont worry about the secondary, just let it float.

Might be better than 2 chokes.

BTW, I started a PCB in Eagle for this:

 

[CJ]

I was thinking of putting the feedback windings in between the pri and sec.
This way, if you wanted to run it without feedback, you could use a switch to disconnect theFBb from the circuit, and ground one end of the FB winding, therby creating a shield, which will lower capacitance and and thus increase bandwidth.

Another aspect of the xfmr being in the loop is that it needs to have greater bandwidth, by a factor of something I forgot.
Something like an extra octave for FB, and an extra octave for posterity.

So, tough assigment, a 40K:600 with feedback.
No wonder nobody makes one!

You can take the FB of the pri, thus avoiding most of the xfmr phase shift, but I do not know if thwe circuit would like the dc on thie cathodes.
Maybe use caps.

Hey, doesen't the FB winding need a ground ref or will it work as drawn?

 

[Larrchild]

I was thinking of putting the feedback windings in between the pri and sec.
This way, if you wanted to run it without feedback, you could use a switch to disconnect theFBb from the circuit, and ground one end of the FB winding, therby creating a shield, which will lower capacitance and and thus increase bandwidth.

Gotcha. That sounds best. Something we can "bypass" later without wishing we hadnt added it.

Another aspect of the xfmr being in the loop is that it needs to have greater bandwidth, by a factor of something I forgot.
Something like an extra octave for FB, and an extra octave for posterity.

I don't think we have 40khz here. And in the descent, there is phase rotation.
Hmm.

Hey, doesen't the FB winding need a ground ref or will it work as drawn?

Dang, I just looked at the schematic. (Thats the problem with projects that start in march) and yes, I show a center-tap on the tertiary winding. So pleaseeese dad, can we add one?

Each leg of the tertiary winding seems to be driving about 1500 ohms load thru the cathode resistors, if it matters. I will have series buildout resistors coming from the winding to set FB amount.
Also the buildout has to be high enough R so that the tertiary winding does not load the cathode voltage or a cap will be required.

 

[Winston OBoogie]

[quote author="Larrchild"]
Winston suggested something I tried a while back, which is use a cheaper CT transformer as a paraparafeed chokes (PP) and dont worry about the secondary, just let it float.
[/quote]

Not that your circuit needs it, but the secondary could also be used for 'transforming' additional L and C back to the anodes for tone or emphasis or ? etc. An OPT stepdown might be a better choice in that case though. I recall someone here posting an old Altec comp schematic that utilized this technique.

Anyway, the ciruit in hand: Nothing useful to add. I like it, thanks Lar :thumb:

 

[Larrchild]

Not that your circuit needs it, but the secondary could also be used for 'transforming' additional L and C back to the anodes for tone or emphasis or ? etc. An OPT stepdown might be a better choice in that case though. I recall someone here posting an old Altec comp schematic that utilized this technique.

Thats the beauty of not being tied to an interstage xfmr only for this "use the primary as a choke" notion. Any xfmr w/ suitable primary will do.

Since the only interstages I like are made by Tango and cost 300 dollars, Parafeed using less erudite iron is a sultry-siren calling.

I love the- squirt some signal from the secondary back to the anode thru a filter network for free-eq idea. Thats strong, cause it technically only places the xfmr in the signal path for the amount of eq added (or subtracted), and cause of the choke situ, does not subtract from the headroom of the circuit, because the eq voltage is merely "added" to the main signal voltage. Sweet.

I have to get this thing back and try all this. I should have never loaned it out.

:grin:

Peace. :guinness:

 

[Marik]

[quote author="Larrchild"]

Thats the beauty of not being tied to an interstage xfmr only for this "use the primary as a choke" notion. Any xfmr w/ suitable primary will do.

Since the only interstages I like are made by Tango and cost 300 dollars, Parafeed using less erudite iron is a sultry-siren calling.

[/quote]

Hey Larr,

If it is of any help, I tried Lundahl interstage and it sounded better as a parafeed. In the end I actually ended up with a cheap Hammond :grin: .

I remember Magnequest had a center tapped plate choke. I have a pair. IIRC, payed something like $70 each a few years ago.

Check them out if it is within your price range.

Otherwise, all the questions to daddy CJ. :wink:

 

[Larrchild]

Well, regarding a cheap transformer instead of chokes:
My theory..is that if you use the core and winding for a choke, you are bound to it's frequency response via its characteristic impedance throughout the passband, but you are eliminating it's transfer function issues of sending the actual signal thru the core for a given quality transformer. hi-z to hi-z is a ballbuster.

I hope thats the way it pans out.

Here's another deep thought: if you did want to use the secondary to add , say a midrange control or something like that, the response of the transformer in that capacity need only be as good as the mids anyway.

 

[CJ]

I found a cool surplus store that has stuff pertaining to machine shops and their tooling.

Well, they had some mag wire, for cheap!

So I grabbed as much as I could carry.
15 bucks a piece for these spools.
All different gauges.
Should last a lifetime, as long as I don't drop them and break another flange.\

And, I have 80 different bobbins coming in the mail from Ram Sales.
I told the guy to grab one or two of every size he had from 75EI to 175EI, and thats what he did.
Twin and triple chamber bobbins included.

So we are still going at it over here, yesssirrrrr..

Just for the heck of it, I sampled some suppermalloy in 75EI.
160 henries with a 1 inch stack and 1000 turns, but thats not gapped.
Still, a big jump over the M6, but running Ni with DC, well, I don't know.
I might send you a Ni version just for laughs.
Just don't short out one of the tubes with your screwdriver, oththerwise you will have a really mellow sounding preamp.

 

[Larrchild]

Nice Score, CJ! yer supplied for a long time now!

just don't short out one of the tubes with your screwdriver, oththerwise you will have a really mellow sounding preamp.

If that ever happens, is there a way to demag the xfmr to restore its non-magnetized core?

 

[CJ]

Yeah, Metalica at 180 db.

 

[CJ]

sent Ram Sales 110 bucks cash in the mail, letter gets lost of course, comes back two weeks later, with the cash!

OK, so a win/lose there, the only time I have a letter get lost is when there is cash inside! :twisted:

Anyway, the Post Office computer put the wrong bar code on the thing.

But, the bobbinns finally got here.

so onward and upward.....

 

[Larrchild]

Oh Henry!

Nice!

Combined with your core-na-copia, you can do it all.

 

[Guest]

[quote author="Larrchild"]Schematic



Whaddya say? Chokes in the 1st stage too? :razz:[/quote]

Constant current source using couple of transistors would be fine.

 

[Larrchild]

Chokes do have an advantage, in that the audio signal waveform can exceed the rail voltage. That intrigues me.

 

[Guest]

[quote author="Larrchild"]Chokes do have an advantage, in that the audio signal waveform can exceed the rail voltage. That intrigues me.[/quote]

Do we speak of a mic preamp, or plate modulation exciter?

Now, about disadvantages:
chokes add frequency dependent phase products, as the result your feedback presence; 2'nd order harmonic (that is frequency dependent!) coming through the loop are distorted and intermodulates input, and so on, that adds higher order distortions that also are frequency dependent, how does it sound? Also, any choke core is non-linear...

It is up to you, however, what kind of ditsotrions you want to generate. My suggestions were based on my taste.

 

[Larrchild]

My suggestions were based on my taste.

Easy enough to find out.

I guess I'll try both soon, and post some sound files.
It's a question that has been on my mind for awhile.
The non-linearities mentioned were considered, I just want to hear them, and decide.

 

[Marik]

I tried a monolitic SS CCS in an output stage, and while it sounded nice I had to abondon the idea because of its high noise. The noise of that stage was dominating :shock: . Something to consider.

It is my understanding the 2 transistor CCS has the same noise problems as monolitic one. IIRC, Gary Pimm has developed a CCS, working around that problem, but it considerably complicates the circuit.

The choke does sound different and for my applications nicely "rounds off" the edges.

 

[Guest]

My pet has the tail as the neighbour's dog, but it is not a dog at all...

Any transistor working in OB or OE connection is a CCS. Any pentode tube is a CCS. Any active element is the source of noises. Any passive element under current and/or voltage is the source of noises. The question is, what is better noise source, what is worse.

However, the most famous noise sources are p-n junctions when reverse broken down, for example Zener diodes and base-emitter junctions in reverse connections. However, they may be used as voltage references in some ICs called "CCS diodes", but it does not mean that all CCSes are good noise generators.
If you use directly biased Vbe as a reference voltage source you will get much-much-much less noises than if you take a reverse biased one.
Any modern OpAmp you may use in your microphone preamp has already current sources both in series with input and in series with the output of the every first differential stage. Are they noisy? Are they too complex like your friend developed?