Feedback windings in output transformers?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

rascalseven

Well-known member
Joined
Jun 3, 2004
Messages
906
Location
"Tejas"
Is there a good text to read that explains the idea of using feedback windings in output transformers? I know that the nicer Focusrite stuff uses feedback windings, as has some older Calrec stuff, and I've looked at schems (app notes) at Cinemag, Jensen, and Lundahl, but those don't really provide enough information to understand what all is happening in the circuit itself (either that, or I'm just slow).

Where can I go to learn more about this? I'm trying to put together an output stage using a drawing from the Lundahl tech sheet (http://www.lundahl.se/papers.html -- the "Mixed feedback.... " doc at the top of the list) using an Edcor XS1100 (their equivalent to Lundahl's 5402). The drawing give some formulas for determining component values for the circuit, but I think they expect a bit more of a working knowledge of transformers than I have.

Help?

Thanks so much,

JC
 
Walt Jung - OpAmp Applications has some stuff in the Signal Amplifiers chapter...line drivers.

You can download the whole pdf free from analog.com here or just that chapter.

Not much though but its a start. Isn't this some kind of german patent?

-T
 
Werner Baudisch seems to be your man...page 73 in the chapter I linked.

In fact, there is more there than I remembered, formulas and a nice explanation from Mr. Jung.

Should sort you out on your way.
 
the Neve 3415 module has feedback and a tertiary output in it. I had 4 and still have the original transformers.

might be an interesting schematic to reference.
 
in another application, the neumann u67 microphone uses the bv12 transformer which has a tertiary winding sandwiched between the primary and the secondary.
u67 schematics
some modifications on this mic alter the circuit and take out the feedback though.
-max
 
I like the sound of all the various preamps I have that use a tertiary feedback winding. I like the idea that the transformer is somewhat self-compensating when the feedback is sourced from it.

There's at least one compressor mic amp that uses a tertiary winding on an input transformer as a spot to insert an active variable load, which shunts out the input signal to achieve compression. The input impedance can vary from a few ohms to over 600, with 150 ohms quoted as the average. Not exactly the hi-fi way to do it, but kinda interesting to hear as an effect. Bonus points to the person who knows what obscure unit I am referring to.
 
Okay... here's a first attempt based on the info given above:

Mixed_Feedback_Output_XS1100.png


Basically the article mentioned above used the same circuit as the lundahl app note with the addition of a 500-ohm trimpot to adjust for minimum THD at 20Hz. Based on the description given the ratio between the input to feedback resistors should be the same ratio as the lower right resistor to the DCR of the primaries.

According to Edcor, the primaries of the XS1100 have DCR's of 104 and 120 ohms. Connected in parallel (with the secondaries in series for 1:1 operation) the primaries' resistance should be 55.71 ohms, right? Therefore, the associated resistor (which should be effectively 1/2 the DCR of the connected primaries) is 28 ohms.

I don't have the XS1100's to test this yet, but does this look like an appropriate starting place for my testing?

Is there anything I've overlooked?

Thanks so much!

JC
 
the only snag i can see is that the xs1100 is not really wound in a fashion that is optimal for this kind of feedback through the transformer circuit. it may still work, but you might find it hard to achieve much improvement over the usual configuration.

the lundahl transformers that are specified in walt jung's book are wound rather differently than the xs1100. the jensen and cinemag transformers referenced in their whitepapers (which use a negative feedback arrangement) are wound in yet a different way.

consider getting a lundahl or an oep if you want to use the lundahl circuit. the cinemag and jensen variants have their benefits as well, and you can use most any quadfilar wound transformer for them.

ed
 
I am not at all informed on this but
In the diagram above, how is there any feedback? Theres only 2 nodes in and 2 out--I'm guessing the primary offers its inductance as a part of the active circuit but through the same potential variations (nodes) that are driving it.?

conventionally either the pri and sec windings share a common common and a portion of the signal on the out is (-) fed back to an active stage to smooth out any variations introduced between by the xfmr or electronics--

or a tertiary winding is used in the same way, except without the requisite of common neutral, allowing electrical isolation of the trans
 
shabtek-

the lundahl circuit is an example of *positive* feedback being used to lower the output impedance (below zero) and effectively cancel out the DCR of the primary winding. part of the distortion mechanism of a transformer is that the primary is not "perfect" and has its own DCR, which, when the magnetizing current flows through it, causes distortion.

i would recommend reading the chapter 6 of walt jung's Opamp Applications book, posted by tom above, for a more thorough explanation of the circuit.

what you are talking about are methods of *negative* feedback, to actually bring the distorted voltage drop across the primary back into the amplifier to cancel it out.

ed
 
Alright, here's my redraw of a Jensen circuit now using a 5534 instead of the AD845 in the Jensen drawing (I have 5534's, and also a couple of the Profile 4804's, so I wanted to use what I have first, particularly since the iron I have is essentially similar to the JT-123 in Jensen's drawing).

Picture2.png


Two questions: 1) is there any reason why this might not work well with a 5534, and 2) in order to have a unity gain circuit (as drawn the gain is 6dB) can I just simply connect the secondaries in series instead of parallel as it is currrently drawn?

Thanks so much for your input, guys. This is interesting stuff. I've never built an output stage with feedback and really want to understand it more.

JC
 
Hi Joel!
Take a look at the outputstage in the ISA430, an IC and a transistor pair. It's the same type of outputstage as the one that was used in the AIR Montserrat consol (31106 channel amps)
 
Sorry for the OT, but which drawing program are you using JC ?
Looks nice & clear. Can you run a sim from it or is it just for drawing ?

Regards,

Peter
 
Peter, I'm just using Illustrator. No, I can't do anything with it other than look at it. I don't yet have a PCB program, but would love to get one (I'm on a Mac) and learn my way around it.

TK, where can I find the ISA430 drawing? Doesn't that use the 'pumpkin' output trafo made by Carnhill?
 
[quote author="rascalseven"]Peter, I'm just using Illustrator. No, I can't do anything with it other than look at it. I don't yet have a PCB program, but would love to get one (I'm on a Mac) and learn my way around it.[/quote]
OK, thanks JC, was just curious.
Imho the most 'pleasant' drawing style is the one as used in older EW+WW schematics, with those rounded corners where wires make a +/-90 degree turn.
For some reason your drawings made me think a bit of that look (except for the corners).
 

Latest posts

Back
Top