Triode vs Pentode vs Ultralinear

[TerrySmith]

I've noticed that almost all EL34 based amps have UL outputs, and 7591 amps are pentode.

My question is what is the advantage & disadvantage between UL and pentode connected outputs?

Some circuits use a regulated screen circuit, like the 50w amp in the RCA book, could that be better?

The reason I'm asking the questions is I have quite a few sets of iron from 7591 based amps without UL taps that I would like to use.

[WA4SWJ]

Hi Terry,

See:

http://www.normankoren.com/Audio/Feedba ... l#Clipping

near the very bottom of the page for one explanation of the connection differences. I'm sure there are other pages that talk about this.

All of the various connections work fine I'm sure. I think it's mostly a matter of what sounds the best to you within the parameters spelled out in the table shown on that Web page.

Regards,

Ed Long

[EWBrown]

From experience, I have found that some tube types work very well in UL mode, and then there are a few a that do not, perhaps the 7591 falls into the latter category?

/ed B in NH

[dhuebert]

http://www.webace.com.au/~electron/tubes/screens.htm.

also:

http://www.aikenamps.com/UL.pdf

In fact if you haven't already, read all the technical articles @ Aikenamps.

My understanding is: Triode gives more even order harmonic distortion for a sweeter sound (low power tho), pentode gives more power but with a small sacrifice in sound quality (I don't notice) and ultralinear lowers output impedance and reduces distortion due to local feedback.

Some tubes are too stressed by UL to last long in this mode as screen grid voltage is equal to plate voltage. This puts a large current through the screen and can shorten its lifespan. Again read the above articles.

"Some circuits use a regulated screen circuit, like the 50w amp in the RCA book, could that be better?"

The more tightly regulated the screen grid voltage is, the lower the distortion is. G2 appears in the electron stream just like G1, which means it can control signal as well, ala UL. In pentode mode, however, G2 gets its signal from power supply artifacts which degrade distortion performance. A quick perusal of guitar amp schematics reveals a choke between B+ and G2 and large value capacitors to ground, a simple L-C regulator.


If you look at:
http://www.nj7p.org/Tube1.php?tube=6CA7 (American EL34)
You will see G1 and G2 @ 450V for AB1 operation.

However, if you look at:
http://www.nj7p.org/Tube1.php?tube=7591A
You will see G1 @ 450V and G2 @ 400V for AB1 operation.

Perhaps, due to G2 having a lower voltage rating on the 7591 the manufacturers felt that UL was a poor choice for this tube.

Another question to consider: what kind of amps do the 7591 transformers come from? In my experience (all too brief!) pentode was used for maximum effiency, ie guitar and PA amps and UL was used for maximum fidelity, ie home entertainment. I hope this was helpful



Don

This editor needs a swift kick in the a$$ as me old pappy used to say

[EWBrown]

Somewhere on the net was an article that listed the best tubes for UL usage, can't find it now... Yellow_Light_Colorz_PDT_09

I did a quick lok through the RCA tube manual, and the VTL book, and it appears that the 7951, the 8417 and the 7355 are not suited for UL usage (G2 voltage is substantially lower than Plate voltage ratings). Same applies to 6CW5, 6P15P (SV83), 6AR6, 6098, 6384, 6146 and 807 (and most other RF and "sweep" tubes).

OTOH, many beam power tubes and pentodes are good for UL usage, among them, the 6L6, 6L6G, 6L6GB and 6L6GC, 5881, 7581A, KT66, KT88, 6CA7/EL34, 6550, 7027A and the 6BQ5 family: 6BQ5, 8BQ5, 10BQ5, EL84, 7189, 6P14P, 6GK6, 10GK6, 16GK6, 29GK6 (?) 6973, 6CZ5, and 6V6, 6V6GT, 7408, 7C7, 6AQ5, 12AB5 and even 6BM8, 16A8, 8B8 and 6F3P.

Most of these tubes, the SG tap should be about 40-43% of the plate voltage, with the exception of the 7868, which should be 50%.

This list is by no means complete, but it's a start...

/ed B in NH

[Shannon Parks]

Hi Gents,

I'm still a little ways from getting my Eico HF86 amp tweaked, but I'm close.

It is very similar to the ST35, but is a pentode amp:
- 12DW7's voltage gain stage has ~300k plate resistor
- 12DW7's other half set up as a concertina phase splitter (but with higher value resistors)
- drives four PP EL84's, with about 340V regulated on the screens and 350V unregulated on the plates

At this point, it looks like the UL ST35 really has the Eico beat by the numbers. The Z565 sure seems like superior iron, with more primary inductance, so from the get go this drag race isn't fair. But I do plan plenty of side-by-side listening 'tests'.

What about patent laws regarding Ultra-Linear transformers at the time? Were people paying royalties to Hafler and Keroes?

Shannon

[EWBrown]

I also have Fisher SA-100 that was "rescued" from the local town dump several years ago. Probably quite similar to the HF-86, it's GZ34 rectified,
2 12DW7/7247s, 4 7189As, "pentode" output circuit. For the price, ($0) it sounds great, but the ST35, both original Dynaco and Baby Blue far surpass it. I replaced the original (still OK) 7189s with a MQ of Russki 6P14P-EVs, and re-biased it, with not much audible difference. It's had some minor mods done by the previous owner(s), but only in the AC primary wiring. And the top cage and bottom plate are MIA, buf for freebies, who cares... Yellow_Light_Colorz_PDT_06
As you can see in the URL below, the OPTs are MASSIVE for the power level.

Sams Photofact PDF here:

http://www.quadesl.com/pdf/sa100_photof ... r%20SA100'

Even though the URL looks kinda funky, it works...


/ed B in NH Yellow_Light_Colorz_PDT_11

[TerrySmith]

The 7591 transformers I was referring to came from a Heathkit AA-100. Outputs are larger than the Dyna A-470, and no screen taps. I have 3 sets to use.

I might try a pentode circuit with the regulated screen like in the RCA book.

[erichayes]

Hi All,

I've been following this thread with some interest due to what's been going on in my lab/shop the last couple of months.

RCA didn't recommend the 7591 for UL configuration because it wasn't RCA's tube (it was GE's). However, RCA did recommend the 7868, which was electrically identical to the 7591 but used RCA' proprietary "novar" base, for UL use. Similarly, RCA didn't recommend the generic 6BQ5 for UL, but did recommend their 7189--a beefed up 6BQ5.

The secret for good UL operation is proper tracking of the screen grid. I have the dubious luxury of prototyping my OPTs with screen taps at different points on the winding, rather than having to settle for the cookie cutter 43%. Some tubes want to see more, some less. It doesn't really have much to do with plate vs screen ratings...look at the EL34. 800 volt plate, but only 400 volt for the screen. Yet it performs admirably in UL mode...as long as you don't exceed about 500 volts on the plate, at which point you start getting really spectacular interelectrode pyrotechnics.

The main reason most straight pentode amps don't sound all that great is that the screen supply voltage isn't regulated. If your screens aren't following (accurately) the plates a la UL, they have to be clamped to within no more than 5 volts deviation. If the screen voltage is allowed to wander about, the linearity of the tubes goes straight into the toilet and you get--guess what?--distortion.

I'm not a big fan of gas-discharge VR tubes, as they introduce hash and noise into a circuit (the same reason I don't like Hg vapor rectifiers). But they do look cool and are a helluva lot better than no regulation at all.

[evsentry3]

Somewhere on the net was an article that listed the best tubes for UL usage, can't find it now... Yellow_Light_Colorz_PDT_09

Hey Ed,
Is this the article that you're thinking of... (about half way down)

http://www.webace.com.au/~electron/tubes/ulo.html

Good read, just wish I could understand it all!!

[dhuebert]

really spectacular interelectrode pyrotechnics

Are these the screen grids melting and shorting to the suppressor due to excessive current or inter-electrode arcing due to high voltages?

Don

[EWBrown]

Speaking of Hafler and Keroes:

First a 56 year old article:


http://www.dc-daylight.ltd.uk/Valve-Aud ... v1951.html

and then there is the expected l dissident opinion from the UK over 50 years ago)

http://www.dc-daylight.ltd.uk/Valve-Aud ... -1952.html

Makes for some interesting and occasionally technically heavy reading.

/ed B in NH

[lynxx]

I have several Kenwood and Fisher and one HH Scott and all of them are set to pentode mode. all of them but 1 run the 7591 tubes in them.