UTC 20 Watt Amp - Any Experience?

[SigloOne]

I have a friend who is going to be helping me build a tube power amp. One design he has recommended is a 20 Watt UTC design.

I found a schematic:

http://www.ampslab.com/SCHEMATICS/Utc20W.gif

Does anyone have any experience with this design?

Also, does anyone know of a suitable replacement for the UTC HA-136, H73, and H84 transformers. Looking for quality iron. Tried searching for these online for used or NOS UTC parts, but no luck.

Any help would be appreciated,

--Steve

[sorenj07]

the schematic doesn't appear... try http://imageshack.us

[dsavitsk]

Probably this one http://www.triodeel.com/images/utc20w.gif

[sorenj07]

Seems like a textbook Williamson topology but a couple things seemed weird to me. The 300pF/3.3K from the plate of the first stage, for example.. Presumably to provide stability of some sort, but isn't it transformer-dependent? The 500pF from the first stage cathode to ground is also weird, I have no clue what that does. If you're not using the same OPT's, I'm not sure that the capacitance on the NFB loop will be equivalent either.

My recommendation for what you do instead of spending years hunting down the original UTC transformers is a direct function of the budget you're willing to allow this amp.

[SigloOne]

Yes, its the circuit that dsavitsk posted.

If this is a design that is transformer dependent and therefore hard to accomplish due to the scarcity of UTC transformers, what do you recommend?

I was considering building a preamp and two monoblocks and am still leaning this way, however I would consider a stereo design so that I don't have to buy as many transformers and can instead put the money into higher quality iron.

My goal is 20 Watts per channel. I want something reliable that I wont have to monkey with all the time.

Thanks for all your help.

-Steve

[richwalters]

Originally posted by sorenj07
The 300pF/3.3K from the plate of the first stage, for example.. Presumably to provide stability of some sort, but isn't it transformer-dependent?

-->Yes it is. If you don't fit them, the amp is a near certainty will oscillate at the upper frequency end. These values determine the HF pole roll off. If you're not using the same OPT's, I'm not sure that the capacitance on the NFB loop will be equivalent either.
--> Yes, also the cap in the global nfb loop will interact. As I am also using a 4 stage amp, with triodes, and sim split load design interesting my values aren't that dissimiliar.

When you get to the testing stage then a bit of skill with square waves can trim these values. However;
If you want to save all this hassle, then a kit form is best.
As for hunting down orig transformers. This is looks a remote prospect.
What's your budget ?

riichj

[Miles Prower]

Originally posted by sorenj07
Seems like a textbook Williamson topology but a couple things seemed weird to me. The 300pF/3.3K from the plate of the first stage, for example.. Presumably to provide stability of some sort, but isn't it transformer-dependent?

Yes, it's a pure Willy. The 300pF/3K3 network is required for high frequency stability. You probably won't need it if you cut back the gNFB to sane levels. The A Number One problem with the original Williamson was excessive gNFB. Of course, back in those days, speeks weren't very good (since there was no analytical methodology for design, and it was very much a "black art") and tended to have sloppy bass. Giving the Williamson some 26+db of gNFB forced the Zo low enough to tame those wild woofers. Reduce the gNFB and high frequency stability isn't a big enough problem to require such stabilization networks.

The 500pF from the first stage cathode to ground is also weird, I have no clue what that does.

Bypass to keep RF riding in on the speaker cable out of the front end. Look at its fc= 677KHz, almost ideal for rejecting the AM BCB frequencies and anything higher.

My recommendation for what you do instead of spending years hunting down the original UTC transformers is a direct function of the budget you're willing to allow this amp.

Since this is a UL design, I wouldn't skimp on the OPT. Bad UL OPTs sound, well, really bad. The originals aren't really necessary if you cut back on the gNFB. You won't need much more than about 12dbv anyway. In my designs, 12dbv of gNFB definitely makes for a solid statey, "Big Box" type of sound. More would make that even worse, and less will improve the sonics for the most part. With less gNFB, there won't be the stability problems either.

[SigloOne]

I haven't really given a thought to the budget yet, however I agree that attempting to find the UTC transformers is an exercise in futility.

I had considered going the kit path, but as with anything where someone else is choosing the components, I figured I'd be happier piecing it all together myself.

Since going with this exact design is a moot point, what would any of you recommend as a good solid design, possibly stereo, based perhaps on KT66 and 5U4 tubes?

--Steve

ps. Although my friend is has tube experience from many years past, I don't. Any good books for someone just getting into tubes?

[chrish]

Doing a search for a possible williams topology amp came up with this.

I have a set of Tamura F-684 6.6K 30watt output transformers, freq resp 30-50K apparently. I know that the williamson requires a reasonable output transformer to avoid instability. Is this topology acceptable/recommended with these transformers?

What output valves work well with this topology? Have some Russian 6L6GT (low voltage types) that would be nice to use, but not essential.

Also read that this topology runs a little more stable with much lower feedback, and that this was a good combination with transmission line speakers due to the lower damping. As I have transmission lines, this sounded interesting.

Look forward to your comments

Cheers,

Chris

[kevinkr]

Originally posted by SigloOne
I haven't really given a thought to the budget yet, however I agree that attempting to find the UTC transformers is an exercise in futility.

I had considered going the kit path, but as with anything where someone else is choosing the components, I figured I'd be happier piecing it all together myself.

Since going with this exact design is a moot point, what would any of you recommend as a good solid design, possibly stereo, based perhaps on KT66 and 5U4 tubes?

--Steve

ps. Although my friend is has tube experience from many years past, I don't. Any good books for someone just getting into tubes?

Any of the books by Morgan Jones - Valve Amplifiers 3rd Edition, and Building Valve Amplifiers. Both are available from Amazon.com and Olde Colony Books. (AudioXpress)