thoughts and ?s on UTC transformers for PP preamp circuits

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

emrr

Well-known member
Joined
Apr 12, 2006
Messages
8,525
Location
NC, USA
Every now and then I consider the UTC LS-51, HA-114, and A-26 for push-pull tube output stages. The LS and HA both mention 1 mA of unbalanced DC in the primaries, while the A does not mention DC at all. Has anyone built a PP circuit with these units, and did you use DC in the transformers, or did you use parallel feed? I cannot think of a single circuit I've ever seen using these pieces. The quoted 1 MA of imbalance would be fine with low current tubes, but seems like it could be problematic with something like 12AU7's or 6SN7's. Sometimes I think parallel feed may really be the only safe way to go with these, and I'm not inclined to see if I blow one up experimenting.

Can anyone address, from experience, the suitability of the various UTC LS, HA, and A series pieces for push pull use? Obviously UTC has some that are designed specifically for PP use (LS-12, LS-51, HA-101, HA-114, A-12, A-26), and others that are only mentioned in regards to single ended use that have split secondaries (LS-10, LS-50, HA-100, HA-113, A-10, A-26). Is this later group wound to less balanced tolerances, or are they perfectly suitable for PP use as well? I have certainly seen plenty of LS-10's on PP input stages. Has anyone tried paralleling windings for lower ratios, and looked for detrimental frequency effects due to the paralleling? Can anyone address the use of the PP types in single ended applications (input and parallel feed output), either at series or parallel connection? Say, like an LS-12 paralleled for single ended 30K secondary and an LS-51 paralleled for P.F. 7.5K primary?

In related thoughts, I see a lot of discussion here concerning the difficulty of winding good iron for push pull preamps like the Langevin AM-5116-B, AM-16, Altec 458A/459A, output of the Altec 428B, output of the RCA BA-21A, etc etc. If you look at the early transformer literature, you find lots of units that are specified for either SE or PP use, especially inputs. Are we to assume that they were not overly concerned with high frequency imbalance in those days, or that we are talking about a difference between 'usable colored sound' and 'modern lab grade' sound when discussing these winding difficulties?

I realize I'm asking for a small dictionary here. I'm mainly curious about folks real-world experiences with these pieces.

Best,
 
> push-pull tube output stages. The LS and HA both mention 1 mA of unbalanced DC

If your push and pull are not mis-matched, you can run 5mA or 10mA each side and hardly get 1mA unbalance. That will cover many 6SN7 rigs.

> see if I blow one up experimenting.

It won't blow up. The bass won't meet spec.
 
There are tons schematics around with push pull UTC iron.
Look in the UTC catalogs for circuits.
Or, visit a good engineering library.
I am fortunate enough to have acess to 3 floors of nothing but engineering books.
First floor is Master Theses' from Stanford PHD's.
Second floor is journals and periodicals.
Included on the 2 nd floor is a complete set of AES journals, varoius audio engineering magazines.
Third floor is nuthin but transformner books. :razz:
I just saw a UTC push pull circuit in one of the Audio Electronics back issues.
I will see if I can take a jpg for you.

cj
 
Thanks PRR and CJ. I just haven't seen those particular pieces in use, which strikes me as odd. I've got a room full of schematics, yet somehow those have never shown up. The PP UTC sample circuits I have seen are the power amp types, rather than the line level stuff.

Unfortunately, there are apparently no EE libraries in my near vicinity of NC. They just grow tobacco here, and do bio-med research. You can get engine blocks machined here, but machine shops won't touch non-automotive work. Everyone wants to work for NASCAR.....
 
Resurrecting this after shooting some raw response curves with various loading and ground references; tried to cover all bases.

I'm finding in SE connection that most split winding UTC's have an upper response dip that recovers right before upper roll-off. Like a 1/3 octave shallow notch knocked out of a proper smooth curved roll-off. True for SE specified types and PP types. All of them exhibit textbook upper roll-off curve when connected in PP. All of the normally single ended types (A-10, A-24, LS-10, LS-50) I've looked at look much better in PP. The workings of distributed capacitance in a high impedance split winding as related to ground reference?

Worst I've seen was an ADC PP type connected SE; narrow -12db notch at 12.5kHz. Same unit in stock amp connection used only one of the split secondary windings SE, with other winding floating. That looked almost normal, with a high boost probably chosen on purpose. Same in PP connection looked great in a modern sense.
 
I've built 3 SR-71's using customer-provided UTC A-26's as grid transformers (PP) and I'm running 100v on the ct at times of great GR. (cath balance pot helps!)

No issues.

Sounds wonderful!

I was surprised myself.

The very top octave does what you say, dips and returns.
But low end (core material) is just lovely!

Output, needing more current then a typical PP preamp, of course is a different fish.
They spec'd HA-114's there.=) A 30's classic, eh?
 
Are you seeing the dip in PP? I'm only seeing it in SE connection, and going away in PP on my samples.

I do really like the UTC bottom end you get with LS series iron.
 
interesting. I wonder if this is a 'luck of the draw' thing as to when it's seen (PP vs. SE). Am I right in thinking this to be related to distributed capacitance differences/interactions in the two high impedance windings?
 
Watch your setup, scope leads and stuff, the can shift your curves quite a bit.
As an illustration, sweep a UTC with 1:1 and 10:1 leads and plot them both on the same page.

Magnetic aging, lams change, grains move around, become less aligned?
Ask Bozworth.
 
[quote author="CJ"]Watch your setup, scope leads and stuff, the can shift your curves quite a bit.
As an illustration, sweep a UTC with 1:1 and 10:1 leads and plot them both on the same page.

Magnetic aging, lams change, grains move around, become less aligned?
Ask Bozworth.[/quote]


Spectrafoo Complete RTA software with pink noise, resistive loading on one side, series resistance on input side if needed to get reasonable match, loading tweaked (using decade resistors) and tried in reverse connection to see if any differences appear. Tried in an amp circuit if also possible.

Seperately, am seeing that mic pre inputs in circuit don't give correct high frequency response if you use a 40-50 db line pad (U type). Am assuming the excessive series resistance is to blame, even though the load resistor is 'setting' the input Z of the transformer. Switch to lower level signal and a standard mic pad (or none if possible) extends the metered top end response. And I speak of extended flatness, not any sort of ringing boost.
 
[quote author="rafafredd"]I´ve built a Langevin 5116 type circuit with the A26. Works really nice![/quote]


5116, or 5116-B? One is SE, one is PP.
 
Glad to hear it's working well in that app. I also saw another 5116-B clone made with an A-24 in PP.
 
I am going to try some "in circuit " sweep tests, leave the xfmr bolted in, maybe then fire up the first stage 12AX and measure the output, more real world.
 
Back
Top