Interstage curious

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

CurtZHP

Well-known member
Joined
Mar 21, 2005
Messages
634
Location
Allentown, PA
My last tube build, "The Heavy Artillery," built around a UTC LS-10X and an LS-50 has been working great.  Still very grateful for all the help I got here in bringing it to life.

I recently acquired an LS-21 interstage transformer.  I'm curious as to how I might incorporate this into the existing design, or something similar.  As I understand it so far, this is fed from the plate of a tube preamp stage, and the secondary is then usually connected to the grids of a Push-Pull stage.

What are the advantages/disadvantages to such a design?  Does it have to feed a PP stage, or can it feed a single tube output stage prior to the output transformer?

The few Google searches I did brought up a load of power amp schematics and a few old radios.  Not much in the way of preamps.

Thoughts?

 
Why ? Probably from A time when valve gain was expensive.
Dont know specs but maybe useful in a varimu dynamics compressor type circuit
 
Would it eliminate the need for a coupling capacitor between stages?  Perform better than a cap?  It seems it would, but a few people I read seemed to have mixed feelings about running the B+ supply through the primary.  What's interesting is that the pins on this transformer are not labeled with numbers like most LS iron.  The primary has pins labeled "P" (for plate?) and "B" (supply?); and the secondary has two pins labeled "G" (Grid?).  The other two pins on the secondary are labeled "F."  Filament?  Those two already have a jumper between them.

I recall someone once suggesting a similar substitution.  They suggested swapping out the cap for a large inductor.  I'll have to see if I can find that thread....
 
Labeled that way, it sounds intended for B+ in the primary.  Don’t recall what the manual says. 

Catalog shows no DC in the primary, so parafeed usage.  The P and the B indicate recommended connection orientation, highs may suck the other way. 


With a pair of G, it suggests the windings are capacitively balanced for PP use and may not have good response SE.  Try it and see.    Every UTC humbucking coil I've ever tested looks better in PP than SE. 

Look at the RCA 76 console program amp, that’s the classic usage. 

That ratio buys you about 10dB gain. 

Look at RCA 86 and UA 175(?) limiters, a transformer like this is run in reverse so the gain is a step-down, purpose being to make the load on the previous stage gain reduction tubes as light as possible. 
 
The 76 PGM amp has one more gain stage than all the other derivative RCA PGM amps.  This was their basic recipe from 1934 on through 1945.    The simplest version was a single tube before the  IT with gain roughly 50dB. 

The PP stage gets a higher max output level when that's needed, so if you want max clean headroom, it's one path.  That RCA approach aims for 0.5W / +27dBm output with less than 1% THD.    With the right OT, you can also use something like a 6SN7 for the PP stage, like the Collins 12X and 12Y.    Collins 12Y and 7JA is in the Tech Docs. 
 
As a small detour, please allow me to bring up this schematic-snippet here as well.

Obviously solid-state, 1977-ish Peavey PA-amp.

As I've always thought to have understood, the use of an interstage-TX in a power-amp was
partly due to the familiarity of the designers with tube-designs
(which actually makes little sense right away since many tubed power-stages used phase-splitter tubes...)

... or could also have been due to minimizing the number of costly transistors.
(but a medium-power BJT still more expensive than an interstage-TX in the mid-seventies?  ::) )

Curious...
 

Attachments

  • 200-TX.JPG
    200-TX.JPG
    113.2 KB · Views: 39
> use of an interstage-TX in a power-amp was partly due to the familiarity of the designers with tube-designs

Not really. Transistors can have LOW impedances. The driver at 30V 50mA "likes" a 600r load. The power transistor bases are more like 10 Ohms. And one of them must be inverted. The transformer does all this in one lump. The low impedances mean that freq response is not hard. While this scheem was around all over, a notorious promoter was Delco, selling their early big transistors. Some of the Fender Rhodes used Delco's plan verbatim.
 
The assumed cost of that interstage-TX remains puzzling, but who knows it wasn't that much higher than a few more transistors back then...?

But wondering why that required signal polarity inversion wasn't simply solved by means of say a quasi complementary output stage? (Sziklai + Darlington combination)
 

Attachments

  • Sziklai-Darlington.jpg
    Sziklai-Darlington.jpg
    48.3 KB · Views: 13
> why that required signal polarity inversion wasn't simply solved by means of say a

That PV was old-skool in 1977; but it worked good.

Here's another thought: DRIVE. It takes real power to drive a transistor. The stage before it, "driver", has to make this audio power. The resistance-coupled plan you show is at best 8% efficient, which may (does) mean high dissipation in the driver. The transformer-coupled plan approaches 50% efficient. This is 1/6th the dissipation in the driver. Or in this case instead of double drivers your Q2 Q4, your Q1 through a transformer can do the job.

Also I think many large shops had tooling for small transformers, which was going idle as transistors took over. So $4 for another transistor, or $2 for an in-house ($3 bought-in) transformer... it can add-up. None of the buyers know what is inside the box. Put the money in knobs and trim! Also the direct-coupled amplifier is prone to blow ALL its devices domino-style in any hard fault, the transformer plan limits the spread of damage.
 
a few circuits i  have heard that use innerstage transformers seem to have less hum,  don't know why,

how are innserstage  transformers judged?  frequency response? pretty easy to get good response from a low ratio transformer,  more important would be distortion, as now you have a non linear device (i guess caps are non linear also)  in the signal chain, thus the praise of such beauties as the Triad HS-29 for having gobs of primary inductance for low bass distortion.  how does phase shift differ from xfmr to cap coupled? good question, would be a good experiment to perform,

stumbled across this>

 

Attachments

  • crowhurst_basic_audio_vol2-130.png
    crowhurst_basic_audio_vol2-130.png
    36.7 KB · Views: 33
It's on my bucket list to stitch up a build  using    'transistor gain' functional blocks with a 'low ratio interstage'  transformer  :)

I have a couple of old UTC in the low Z winding config ...  I did do some preliminary tests driving it with some JLH class A headphone amps modules  ..  a pair of single ended JLH headphone modules, driven out of phase (wrt each other)  and ... 

...  driving the utc  'push-pull'  [250ohm+250ohm seperate winds primary,  and one 500ohm winding on secondary], about 5Wrms power handling, with a reasonably  wide bandwidth [approaches 30Hz -3dB with good fidelity.

My initial measurements were good - much better than I expected, even in this most simple of application examples ...  enough to make me want to use the scheme as the 'output amps' on a nice 2RU  'hybrid'  build  (tubes and traffos on top, point-point wiring inside with some  'pcb assemblies'    here and there    (JLH '5 transistor amp' and 'liniac' modules)

So hey!  no shame in using these kinds of 'hybrid'  transistor, transformer circuits in the old school fashion.

That's one thing I like about modern '1ru headphone amplifier'  units  ..  a 'quad headphone amp'  on the 'low-cost side'  can give you a nice '4 channel *mono*    'drive amplifier'    ...    [balanced]    ..  which is fantastic for investigating    'driving transformers'  in hifi,  push-pull, single ended ...  you can do it all!
 
This is shaping up to be an interesting project.

Figuring out the push-pull stage between the interstage transformer and the output transformer is what I'll likely get stuck on, since I've never done one before.

Is it just a simple matter ("simple!"  Already I'm laughing...) of two triode stages, one for each side of the secondary winding? 
Seems the literature for the LS-21 implies as much.  "Plate to PP grids."  So, from the plate of the previous tube stage to the grids of two separate stages, but obviously with their respective outputs at opposite polarity to each other.  How am I doing so far?
 
Anyone happen to know what the story is with running plate supply voltage through the primary of a UTC LS-50?

According to their 1963 catalog, the LS-50 is a "single plate to multiple line" transformer.  I'm already using one as an output transformer coming from a single tube stage (not push-pull), but it's capacitor coupled.

I'm guessing an LS-21 can handle it, since it's labeling seems to imply as much.
 
CurtZHP said:
Anyone happen to know what the story is with running plate supply voltage through the primary of a UTC LS-50?

According to their 1963 catalog, the LS-50 is a "single plate to multiple line" transformer.  I'm already using one as an output transformer coming from a single tube stage (not push-pull), but it's capacitor coupled.

I'm guessing an LS-21 can handle it, since it's labeling seems to imply as much.

Uh, no, unless I'm reading you incorrectly. 

No DC in an LS-21. 

LS-50....no DC. 

Right there in the catalog, both of them. 
 
EmRR said:
Uh, no, unless I'm reading you incorrectly. 

No DC in an LS-21. 

LS-50....no DC. 

Right there in the catalog, both of them.

Nuts!  Somehow I missed that.  Capacitors it is.  But now I'm really confused, because I was under the impression that the LS21 was specifically for going between a plate and the grids of a PP pair.  Just about every other design I've looked at shows the plate supply for the preceding stage running through the IS primary.
 
Current in a PP configuration is balanced between the two halves and in opposite directions. So the core will see net DC of zero.
 

Latest posts

Back
Top