Some good info here

[Ears]

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

[VacuumVoodoo]

Yes it is. especially section on screen grids should shed some light on the unenlightened

[Ears]

Yes it is. especially section on screen grids should shed some light on the unenlightened :wink:

Yes, and a very timely find for me, I was about to ask a few questions on exactly that topic, so was pleased that my apparent lack of knowledge was at least excusable, since...(quote):

""As long as the Plate voltage is higher than the Screen voltage, Plate Current in a Screen-Grid tube depends to a great degree on the Screen voltage and very little on the Plate voltage" (RCA Manual RC14)

Beyond the above basic design criteria, little discussion is offered in manufacturers' tube handbooks regarding the effects of Screen-Grid voltage on Plate current."

The approach generally taken is to promote the application and use of vacuum tubes by publishing "typical" operating conditions for vacuum tubes, including recommended Grid #2 operating voltages.

In the case of audio tubes there are copious examples provided that cover likely popular uses - often taken by designers verbatim or as recommended by tube manufacturers, without researching alternatives - resulting in copycat, "more-of-the-same" designs. After all, why go the the trouble and expense of researching something that someone else had already pre-determined and/or recommended - particularly if that "someone" has the expertise of a tube manufacturer?

Curves are nearly always provided for Control Grid (Grid #1) modulation characteristics, but no so for Grid #2.

Hence little published data is available to demonstrate the effects upon Plate Current from varying Grid #2 voltage."..........

[David Root]

I have been reading and re-reading this Aussie's writings for about 6 weeks now. Very interesting. I agree the article on screens is excellent and much needed.

He gives a fancy name to basically using a diode in the screen supply line as a means of preventing the AC current stream from cathode to anode from diverting some of itself thru the screen circuit, thereby boosting the plate current and output power.

Now I know this is a simple trick that can be used in the power supply line too, to prevent big power section transient current draw from sucking supply from the preamp section, see the Komet 60 schematic for example.

One thing bothers me: if this diode in the screen line is so powerful, how come no manufacturers use it, even in hi fi amps, which is what this guy is really discussing. Is it possible this was just overlooked at the end of the tube era? At that time silicon diodes had been in common use for several years, so I find it difficult to believe no-one latched on to it.

He does have what sounds like a good point about keeping screen voltage just high enough to pass the electron stream along to the plate without collecting any, in order to increase plate efficiency and output power.

I've been considering building a pair of 60W monoblocs for a CD power amp, using 6550s and using his techniques including the separate screen supply.

I guess my question is does anyone see this as an improvement over a simple UL OT, which is a known improvement over straight pentode mode for Hi Fi, (but not for MI amps). Or maybe even steel guitar amps for that matter.

[jelle]

I have read this diode application in the screen supply too where it will prevent the unhappy few electrons that are caught by the screen grid to be shunted to ground and go to the plate instead. I have not dared to test it.

I'd love to see results of an actual bench test to learn what the effect is and if a bigger OT is needed etc...


Basically I'm interested in this info because I feel it is a great idea.

[David Root]

He tested it on a 1959 Marshall. Took out the screen resistors and replaced them with diodes. With 520 V plate at idle, the output power went from 90W to 160W, the plate current went to 180 mA per tube and the dissipation to 55W! Would have to adjust the bias to suit, that would cut the power down to size.

Said it sounded great but neither the tubes nor the iron would last too long. No s**t Sherlock!

I would have left the resistors in I think. He recommends that for hi fi.

I'm thinking of trying it in the amp I'm just finishing up, which is a modern 4-stage crunch machine with a 50W full bandwidth toroidal OT at 3k4 Zaa and a pair of 6550s. He has an interesting way of sizing screen grid resistors: (Vs/Vp)*(Zaa/4). I'd never seen any calcs. for this, just "1k 5W" fits all. In my amp above that would be (355/490)*(3400/4) = 616R, not 1k0. Wouldn't need 5W rating either.

[LOUDthud]

I read the article referenced above. (Boring) Although there are some valid points, it reads more like the dissertation of someone that never built an amplifier. One example sited is a 100W amplifier with 600V B+ and a 5K plate to plate load. He then calculates the RMS voltage for full output and subtracts half of that from the B+ to get a safe screen voltage value. Hello, the voltage across each tube at the respective peak of the waveform is about 100V. Should the screen voltage be less than 100V? I doubt the tube will be able to conduct the 200ma peak necessary to produce full output. In another place he suggests using a 10,000uF cap to bypass the screens. Such caps are called "widow-makers" for a reason. Towards the end of the article he finally relents and suggests that some resistance in the screen circuit is proper design practice.

When the plate voltage of a pentode falls near the knee of the plate curves, the screen current shoots up to a pretty high value. This is sometimes shown on the plate curves as the dashed lines Ic2 on the right graph. It is this area where a 1K resistor in series with each screen performs it's magic. The left graph shows the plate current vs voltage for different values of screen voltage with grid voltage at zero. It gives a pretty good idea of what the screen voltage needs to be for a desired maximum plate current. Remember, that's the screen voltage after the dropping resistor.

No mentioned in the article is what happends when the load is mismatched. It appears that when a 16ohm load is connected to an 8ohm tap that the screens are abused and when a 4ohm load is connected to the 8 ohm tap that the plates are abused.

Just my .022

[krash]

Given that I recently experienced a baffling issue with one of my amps that was solved by screen R values, I read this article with interest.

I think for the most part, some of his recommendations and conclusion are overly conservative, maybe over the top, but he does make excellent points. Mostly the good point is that you should design in your screen R value and screen supply circuit on purpose, knowing the expected behavior.

Now, since we tend to discuss TW amps here, I think the standard implementation in TW amps is probably ok. However a 100 ohm screen R in a Marshall 18W for each tube is probably totally inadequate IMHO, at least if my recent experience is any indicator.

Please note I am not masquerading as an expert here. Just trying to move a notch or two up the learning curve.

[jelle]

I agree. I learned a lot from this website. I would recommend people to read it but it can be somewhat dry stuff maybe...but overall I liked it!

[Ears]

I've corresponded with the site's owner who was friendly and provided me with some more info which I attach.
By the way I discovered his claim re max value of useful screen voltage is also supported by the data sheets of the transmitting valves TT21 TT22(basically KT88 on steriods) - also attached.
I'm fairly sure Grimwood is open to discussion.

[David Root]

So putting in the diode immediately behind the screen grid resistor does not affect possible overheating from excessive screen supply DC current. Therefore a 5W screen grid resistor value is still valid, assuming the plate voltage is not dragged down into the area where the grid current skyrockets and the screen goes into meltdown.

The diode does prevent the AC and possible collected DC in the cathode to plate current stream from getting into the screen supply behind the diode and reducing output power.

Ears, did you get into the diode thing with him? It certainly appears he's right about limiting screen voltage, which of course most MI amps don't do. But then his perspective is primarily hi fi. Which reminds me, I didn't agree with most of his stuff about MI amp designs.

[Ears]

Ears, did you get into the diode thing with him? It certainly appears he's right about limiting screen voltage, which of course most MI amps don't do.

No I didn't. But his email address is on his site somewhere . I'm still trying to clarify a number of things about screens in my own mind, before thinking about performing tweaks. Whether he is always right about screen voltage is not quite what I meant by his claim being supported by the data for TT21. But the claim is there in the Marconi data: "there is no advantage in setting the fixed screen voltage supply in excess of about 300V". But the first graph provided by LOUDThud reminds us that there is an advantage in increased power output. I think I need to re-read a whole heap of stuff.

[krash]

you know Randall Aiken's article evaluating EL84s from JJ and EH remarked about the screen resistors and noted no noticeable change in tone going from 100 ohms to 2.5K but a 1-2 watt power reduction.

I think the tradeoff is more than likely worth it. These days I am finding very few guitar players who are _really_ all bent up about not having quite enough amplifier power, to the tune of one or two watts making or breaking the deal. Especially if it improves tube life or adds some great margin of safety, the added reliability of the amplifier is likely a good compromise against that one or two watts IMHO.

[David Root]

Perhaps the diode trades off the lower screen voltage as far as output power is concerned, with the possibility that bias adjustment to prevent overdissipation (perhaps in both plate and screen) is necessary.

He mentions lower distortion levels and too as a result of adding the diode on each screen, and no AC crosstalk between tubes. How would this work?