I followed a link while reading another thread and came across this article or rather copy of a email:
http://ken-gilbert.com/techstuff/triode_connected_pentodesbts.html
This is of great interest to me. Can anyone explain the advantage of this method over that of using a simple resistor. Also, what is the advantage of biasing the "screen grid to -5v? Why not connect it to the cathode in the traditional method? I would like to try this on my new EL34 amp but I would like some opinions first. Will this method make a strapped EL34 sound more like a "real" triode?
http://ken-gilbert.com/techstuff/triode_connected_pentodesbts.html
This is of great interest to me. Can anyone explain the advantage of this method over that of using a simple resistor. Also, what is the advantage of biasing the "screen grid to -5v? Why not connect it to the cathode in the traditional method? I would like to try this on my new EL34 amp but I would like some opinions first. Will this method make a strapped EL34 sound more like a "real" triode?
Stopped reading about 2 pages down. AHHHH THE WHITE ON BLACK IS THE HURTING OF EYES NOW! ...Fading... still got lines across my eyes... ugh... anyways.
The very logic this guy has is bad, a screen's potential for acceleration has nothing to do with how far it is inside the electron beam. Shadowing it by the grid is fine AFAIK. Although... I suppose the density of electrons at that point would tend to produce a generally negative field there (space charge) meaning that putting a screen wire in the middle of this charge would tend to assist things. But then, the repulsion of this charge on the electrons from any given point to another would be negative closer to the cathode (tending to repel backwards) and positive closer to the plate (pushing them into it). So I say it averages out and you're better off with screens in the grid's shadow to reduce dissipation.
Don't know what else he's going on about. Reducing screen voltage will ALWAYS reduce plate current, so you'll certainly get less power with the zener modification. Thus, being closer to max. output, it could be argued that distortion will be higher for a given signal level.
Typically, using a screen voltage lower than plate reduces distortion, but as I recall it's mostly odd order. SE will always have 2nd H distortion.
Tim
The very logic this guy has is bad, a screen's potential for acceleration has nothing to do with how far it is inside the electron beam. Shadowing it by the grid is fine AFAIK. Although... I suppose the density of electrons at that point would tend to produce a generally negative field there (space charge) meaning that putting a screen wire in the middle of this charge would tend to assist things. But then, the repulsion of this charge on the electrons from any given point to another would be negative closer to the cathode (tending to repel backwards) and positive closer to the plate (pushing them into it). So I say it averages out and you're better off with screens in the grid's shadow to reduce dissipation.
Don't know what else he's going on about. Reducing screen voltage will ALWAYS reduce plate current, so you'll certainly get less power with the zener modification. Thus, being closer to max. output, it could be argued that distortion will be higher for a given signal level.
Typically, using a screen voltage lower than plate reduces distortion, but as I recall it's mostly odd order. SE will always have 2nd H distortion.
Tim
I'd be interested to see if there are any actual (as opposed to claimed) performance differences from using a zener to stand off the screen from the plate by a few volts. The poster in this reproduced exchange doesn't cite any. I've got a guess, but would rather see data.
It's interesting to consider the magnitude and relative phase of the voltage variation with signal caused by the strapping resistor.
It's interesting to consider the magnitude and relative phase of the voltage variation with signal caused by the strapping resistor.
Oh yeah, the strapping resistor. That does basically nothing in normal operation with normal values. Even if screen peaks to 50mA, it'll drop only 100 * .05 = 5V. A 1k resistor will soften things up a bit more, but it's important to note the voltage across such a resistor varies wildly with operating point, making the tube wheze and cough as a signal swings near Vg=0V.
Tim
Tim
You're being too severe, even with those conservative numbers. Let's consider some tubes that might actually be likely to be strapped and see the difference between zero signal and max screen current per tube (tube manual numbers):
6L6GC: 5-8 ma
7591A: 5-10 ma
7027A: 5-10 ma
8417: 12-15 ma
6550: 15-18 ma
So, in your worst case, the screen fluctuation due to a strapping resistor is something like 1 or 2 volts out of a 300 volt swing.
Sch3mat1c said:
In my browser it is Black on Black, makes it even harder to read, but at least i'm not cross-eyed. Here is the original email that i clipped:
I never dismiss anything Bill says lightly (at least about electronics)... i haven't tried this yet, but i have seen Bill install it into an amp for a client.
dave
SY said:
That would be with a suitable capacitor bypassing the resistor. I mean instantaneous figures you find on the left side of the loadline. 🙂
Tim
huh?
From reading Planet10's copy of the email, it seems to me that even just 5 mA of screen current will charge the zener limited capacitor string to the 31V (5*6.2V). The two 10K equalizing resistors in series with 5mA, would give .005 * 20,000 = 100V, so no problem charging up to the 31V limit. The 10 Ohm resistor and ferrite bead is probably just to prevent parasitic oscillation like a grid stopper?
Having 31 V less on the screen grid than the plate will drop the max. intercepted plate current (= screen current) down on the left, or high current, side of the plate characteristics nicely, causing the screen to run cool. Seems like a nice idea for the tube's longevity. But it will limit max plate current and, hence power output a bit, since the screen voltage determines the max plate current. If one could boost the B+ voltage by 31V as well, then it would be a "wash". If you don't care about the tube life, then Tim is probably right about more headroom giving less distortion, so skip the mod. The -5V on the Suppressor grid may be a little overkill, since it usually is a pretty open spaced grid. My 2 cts. I could be wrong.
Don
From reading Planet10's copy of the email, it seems to me that even just 5 mA of screen current will charge the zener limited capacitor string to the 31V (5*6.2V). The two 10K equalizing resistors in series with 5mA, would give .005 * 20,000 = 100V, so no problem charging up to the 31V limit. The 10 Ohm resistor and ferrite bead is probably just to prevent parasitic oscillation like a grid stopper?
Having 31 V less on the screen grid than the plate will drop the max. intercepted plate current (= screen current) down on the left, or high current, side of the plate characteristics nicely, causing the screen to run cool. Seems like a nice idea for the tube's longevity. But it will limit max plate current and, hence power output a bit, since the screen voltage determines the max plate current. If one could boost the B+ voltage by 31V as well, then it would be a "wash". If you don't care about the tube life, then Tim is probably right about more headroom giving less distortion, so skip the mod. The -5V on the Suppressor grid may be a little overkill, since it usually is a pretty open spaced grid. My 2 cts. I could be wrong.
Don
Hi,
I don't think the voltage as such matters but the polarity does. It does seem rather far fetched, something I'd like to see on a plotter first.
BTW, this will only be of use on the Euro/Russian EL34s.
Not the current production Ei EL34s which are tetrodes like the US 6CA7s with the suppressor tied to the cathode.
I'm putting my
on you...
Cheers,😉
I don't think the voltage as such matters but the polarity does. It does seem rather far fetched, something I'd like to see on a plotter first.
BTW, this will only be of use on the Euro/Russian EL34s.
Not the current production Ei EL34s which are tetrodes like the US 6CA7s with the suppressor tied to the cathode.
I'm putting my
on you...Cheers,😉
I'm so glad "G" asked this question, since I was puzzling over it myself and was going to offer it up to the group for discussion.
I'm still not sure I understand it its circuit effects yet, but I'm extrapolating from everyone's comments that:
1. No-one here has yet tried it, and so can't yet comment on it sonically or electrically (ie its actual measured performance in a circuit)
2. Despite the pedigree of the original designer; the theoretical foundations that appear to support his mod are either not clear or unsound, thus negating the desire for someone to actually want to try it.
<goes back to trying to think it all through...>
I'm still not sure I understand it its circuit effects yet, but I'm extrapolating from everyone's comments that:
1. No-one here has yet tried it, and so can't yet comment on it sonically or electrically (ie its actual measured performance in a circuit)
2. Despite the pedigree of the original designer; the theoretical foundations that appear to support his mod are either not clear or unsound, thus negating the desire for someone to actually want to try it.
<goes back to trying to think it all through...>
Majestic said:
LOL. I think you have a firm grip on the situation Majestic. Does anyone have a breadboard they could try this on? I haven't been able to find a good source for the OsCons.
LOL - well I guess there's first time for everything 😉 😀G said:
I'm sure (ok, well I'm not!, but I strongly suspect...) that any other high quality electro. should suffice.
However, Googling does turn up more than a few entries.
eg:
- http://www.capacitorsplus.com/banner_main.htm
- http://diyparadise.com/movingsale.html (undated page - always a worry, but still worth trying since they have the exact value you want too.)
- http://www.bdent.com/
or you could always ask the natiional distributor for a source: - http://www.sanyo.com/industrial/electronic_components/capacitors/os_con/
Hi,
The key to good performance here is the low ESR of the cap so other may well do as well as may film caps.
Cheers,😉
The key to good performance here is the low ESR of the cap so other may well do as well as may film caps.
Cheers,😉
Majestic, that's a very perceptive analysis!
If I had all the time in the world, I'd try a lot more things than I do. But the rationale and analysis make little sense to me, and given my null results with similar stuff (specifically, going between 100 ohm strapping resistors and direct connections in an EL34 amp), I'm not optimistic.
If I had all the time in the world, I'd try a lot more things than I do. But the rationale and analysis make little sense to me, and given my null results with similar stuff (specifically, going between 100 ohm strapping resistors and direct connections in an EL34 amp), I'm not optimistic.
I agree with that too, Fred. But I did want to leave the door open on that one since the designer did appear to be quite specific about his preferences.fdegrove said:
And, given if we were to read in between the lines, we might assume that in order for him to be so specific, that he's played around with other manufacturer's offerings and found them 'wanting' in this application.
(Still, I always wonder if I'd be able to hear the difference...
)Cheers SY. I'm knowledge-and-budgetry constrained myself at the moment! So I'm with you in spiritSY said:
But back to the circuit, I think I'm slowly gleaning some of the designer's original intent. Obviously the whole point of this is, ultimately the(/any) audiable sonic difference that this is going to make. But I'm trying to work out whther there is something in there that a simulaton (eg a SPICE model) might be able to show too.
To do that I guess I'm going to have to have a better understanding of why the current "tried and true" resistor method is supposedly not as good. Or, restated, what Bill (the designer) is trying to do differently, since it seems there is a bit of subtlety in the solution of his.
(--Either that, or as noted; vodoo curcuit improvement obscurity! 🙂
- Status
- Not open for further replies.