Ultra Linear Hi-Fi ?

[Amp Kat]

Anyone hook up a Hi-Fi in ultra-Linear ? I have a Fisher SA300 that the owner replaced the OT's but they are UL and I've allready hooked them up but was wondering about how it would effect the AD & DC bias and how much dissipation would the screens use. I guess I could throw a current meter in between the screen and tranny and monitor it. Just wondering if you guys had done it ?

[bob p]

as you may already know, the whole idea behind UL is to offer the best features of both triode mode and pentode mode operation. (or the worst features of both, depending upon your philosophical point of view). the whole idea is that you get the lower distortion like you would get with triode mode operation, and the higher power output like you would get from pentode mode operation.

depending upon what tubes you're using, you might just want to take a peek at spec sheets. some spec sheets have separate tables for UL operation that should answer your questions.

the original paper on UL operation by Hafler & Keroes is worth taking a look at. I don't know a link, though its pretty easy to find with google.

[Rob Mercure]

Hey KB,

I've owned several power amps using UL connection and I like 'em for HiFi use - my currrent system consists of a Scott 340B using 7591s in UL and I'm completely satisfied with 'em.

Bob mentioned the "lower distortion" of UL use but lowering the impedance of the output stage and increasing speaker damping are the benefits usually touted (which my fall within "lower distortion" I'll admit) and where specificially designed in UL operation can allow good performance without any NFB.

As the screen taps on the OT are "closer" to the CT than the plate "ends" the screens usually run slightly "hotter" than the plates and there is some increased plate current. While I've not done any A/B using HiFi amps I did monkey around with the 130W Fender Twin to determine for a customer whether he should remove it - and thankfully he left this to my discretion cuz I'd never have taken on the job if I had to A/B UL/not UL for him with different speakers and sources, etc. The "experiment" was conducted using brand new, slightly cooked in, SED 6L6GCs that I had carefully matched myself (at BRM I used to just buy the unmatched boxes of 100 and then match up pairs/quads/hexes ("the goblins will get you if you don't watch out")) and I really didn't notice any great change in cathode current between the types of operation - but there did seem to be a "second" break in period for the output tubes once I wired out the UL. The SEDs seemed to want another 4-5 hours under load to stabilize but once stabilized there really didn't seem to be much change.

As far as the Fischer I assume that the original circuit included some sort of screen grid current limiting resistor? If so I'd just leave the resistor in place and try 'em UL - if Fischer didn't include a limiting resistor I'd add 'em until you're satisfied that the UL connection operate within parameters that you are satisfied with. And I know you've already checked the bias supply to assure that there is sufficient DC swing if case you need to limit the idle current a tad.

Not sure that this helps much but I hope so. While I know it's probably silly the inefficiency of triode output stages "annoys" me and I've never been convinced that DHTs were crafted within the palms of god so UL does seem to nicely "split the difference" (but if I were really nosey about this I guess I could unhook the UL taps on one channel of my Scott and after biasing each side identically have a friend sneak in and "randomly" click the "channel reverse" switch when I'm not looking to see if I noticed a difference <grin> - or I could take up earthworm ranching <'nother grin>).

Rob