OT Primary Impedance for 7591 P-P

[EETStudent]

Hey guys,

I would like some advice regarding the output impedance of a push-pull pair of cathode biased 7591's at 375vdc. I've looked at the data sheets, and most examples indicate 6.6k for all fixed bias settings. For cathode biased at 450v plate voltage, the impedance jumps to 10k. For my current project, I have a 20W 8k tweed deluxe style OT, and a 6.6k 30W beefy 'Trainwreck' style OT. My guess is that these tubes would prefer to see upwards of 8k, but what technical qualities should rule out the 6.6k? Thank you

[loudthud]

The 7868 is almost identical to the 7591 except for the Novar envelope. Most examples (RCA RC-26) for fixed bias give 6.6K for a load. I would go with that

[Chuck H]

OTOH Fisher ran 7591's into a 7.6k load.

I fail with tube charts. But there is a way to plot the other parameters (THD, Vp, THD, etc.) to determine a desirable load. Maybe someone will chime in.

[soundmasterg]

My Bogen CHB100 uses four fixed bias 7868's, and the OT has a 2.5k load for four tubes...so thats 5k for two tubes. I think the 5k would be max power, but that amp is also running at 480v with today's wall voltages or 450v according to the schematic, so that changes things of course. The OT is small for the tube complement so the amp is 55 watts RMS when it could be 88 watts according to the RCA tube manual specs. Cathode bias tends to run a higher impedance, and I think lower voltages also tend to run a higher impedance too. I think either transformer would work...and heck, why not try both and see which you like better?

Greg

[EFK]

I've built a few amps with those as I think they're great tubes. All were cathode bias in the 325-350V range, and I used a 6.6K trainwreck type OT with all of them. Worked beautifully and a stout chunk of iron at that.

[Gingertube]

6K8 in this old Ampeg ET1B with 7591

Schematic Trace from my restoration job.
http://music-electronics-forum.com/t33513/

Cheers,
Ian

[potatofarmer]

Like loudthud said, electrically they're pretty much identical to 7868s. Check out the last page of the RCA 7868 datasheet for some suggested load impedances: http://www.guitaramplifierblueprinti...files/7868.pdf

I just built an amp with a pair of cathode-biased 7591s, plates around 450V, 8k4 p-p. You just have to keep an eye on your screen dissipation; in my amp ~600R screen stoppers were sufficient but I went with 1k anyway.

Check out Merlin's page about picking a load impedance and the consequences of low vs high: The Valve Wizard -Push-Pull

[wizard333]

I think lower voltages also tend to run a higher impedance too

I thought the opposite was true for all tubes, generally speaking?

[loudthud]

I thought the opposite was true for all tubes, generally speaking?

In general if you wanted to keep the power the same you would lower the impedance. What you run into is the tubes can only handle so much current, so at some point lowering the impedance won't makeup for a lower B+. The 7591 can handle lots of voltage, which is the way to go for more power. For tone, a moderate B+ is probably where the sweet spot is.

[imaradiostar]

I have EH 7591's back biased with a triode electronics 6k6 output tf (wired ultralinear with 470 ohm stoppers) and it's my favorite amp. Your mileage may vary.

Jamie

[jazbo8]

6.6k maximizes the output power, but with higher distortion.

[J M Fahey]

In general if you wanted to keep the power the same you would lower the impedance. What you run into is the tubes can only handle so much current, so at some point lowering the impedance won't makeup for a lower B+.

That's the point.

A lot of people speak "tubes" but think "transistors" .... or so it seems.
Tubes are current starving or current limited, period.

Using lower than correct impedances is rampant everywhere.
Of course, amps may still sound good, that's subjective or personal, but *real-measured* power same as stated in the brochure is nowhere to be found.

The curves shown (thanks jazb8) show this painfully:
he graphs a 6600 ohms load, choosing peak current as slightly over 200mA , say some 203/205 mA or so, and then he has a saturation voltage of less than 40V . *Very* good .
Remember saturation voltage must be substracted from available +B (under load) to get *useful* peak voltage, transmissible to the load.

Now if you want to pull meager 250mA out of same tubes, very little more, saturation voltage increases to jaw dropping 125/130V !!!!!!

So an average designer, trying to pull extra power from his tubes, lowers load impedance, raises +B to hairy levels (say, 500V or thereabouts), which he wastes uselessly as saturation loss in the tubes, and all he gets is abusing tubes to death.

What's the point of slamming 500V on tubes and then losing 125V out of them?
I just don't get it.

km6z has already mentioned meeting old Russian Musicians, (no, not Rock guys) using their old workhorse tube amps every weekend, still with the original power tubes .... many 20 years old or even more.

The secret?: they are used "by the book": 6L6 fed 360V, screens are not abused, and loaded as the datasheet says.

Ok, they get 30/32W RMS where Leo would have gotten 40/45W RMS ..... so what?

It's not that significative *loudness* difference anyway, and reliability is impressive.