[sdiy] Urei 1176 old version single ended output - how does it work?

[René Schmitz]

Hi Jürgen,

I think the winding does provide some local negative feedback, and also 
uses the collector AC current a second time. I would guess that the 
transformer is wound such that you transform the emitter output 
impedance and the collector output impedance to the impedance on the 
secondary.
There are some tube amps which use such a split not quite plate, not 
quite cathode circuit. Namely the Quad II. 
http://www.drtube.com/schematics/quad/quad-ii.gif

Here we have a push pull variant, where certainly the magnetic flux 
would cancel, but still they use this split circuit.

Cheers,
  René

jhaible at debitel.net wrote:
> Hi.
> 
> I have a question about something I don't understand.
> 
> I have built a clone of the Urei 1176 Limiter some years ago.
> But there I have used the driver station of later versions,
> which is an ordinary class AB power amp and output transformer.
> (http://www.oldcrows.net/~jhaible/compressor/jh_compress.html)
> 
> But ever since I built this, I was wondering about the older
> versions' output stage, which is single ended class A with a
> very special transformer coupling.
> 
> Here are two links to show what I mean:
> http://www.waltzingbear.com/Schematics/Urei/Urei_1176_page_2.htm
> and
> http://www.waltzingbear.com/Schematics/Urei/Urei_1176LN_series_D.htm
> 
> These circuits leave me puzzled.
> I know that it's possible to build transformers that work directly
> in single ended circuits - they carry the whole plate or collector 
> DC current, they have an air gap to avoid satuaration, and they 
> need to be bigger than a push-pull type or ac coupled transformer
> of similar power rating and frequency response.
> So far everything is clear. Most single ended tube amps work like
> this, and there was a Neve (?) transistor amp like this as well.
> 
> But the Urei circuit has one primary transformer winding connected
> to the collector of the power transistor, and another primary
> winding connected to the emitter. So my first thought was they made
> some clever circuit where two identical DC currents thru two indentical
> transformer windings will make their magnetic field cancel (like
> in a push-pull stage), such that no special air gap and oversized
> transformer is needed. And if you look at the polarity of the
> primary windings, apparently that's exactly what they are doing.
> (The "black" wire apparently is tied to a positive supply voltage;
> there must be a connection dot missing where it corosses a horizontal
> line.)
> 
> So the DC current's influence will cancel. Good.
> Only that the AC will cancel as well. 8-(
> And here is where I don't understand it. It looks like a phase
> splitter configuration, but it isn't. When the power transistor
> is conducting more than its DC bias current, both collector and emitter
> current will increase - same for decreasing - so if there was
> a cancellation effect for DC, there will be the same cancellation
> for AC. I can turn it around as much as I try, it's always that way.
> Unless, of course, one of the primary windings would be _bypassed_
> for AC, by connecting a capacitor in parallel to the winding.
> Now there is such a capacitor, C14 in 
> http://www.waltzingbear.com/Schematics/Urei/Urei_1176_page_2.htm ,
> but it's way too small to bypass a considerable amount of AC
> current.
> So I'm really puzzled.
> 
> Of course not knowing the transformer data makes interpretation
> of the circuit difficult, and the asumption of a DC currentn
> compensation may be wrong from the start. I can't even read all
> of the resistor's component values exactly.
> The purple audio version of the schemos show readable component values,
> and the connection dot mentioned above, at least:
> http://www.purpleaudio.com/pdflib/PurpleMC76.pdf
> 
> I thought I'd share my thoughts; maybe somebody else on the
> list can shed some light.
> 
> JH.
> 
> 
> -------------------------------------------------
> debitel.net Webmail
> 

-- 
uzs159 at uni-bonn.de
http://www.uni-bonn.de/~uzs159