http://www.t-linespeakers.org/tubes/triode-trick.html
I understand the article, but...
What about this?
http://www.webace.com.au/~electron/tubes/oes.html
http://www.webace.com.au/~electron/tubes/triode.htm
I don't quite understand what this guy says. If I'm not wrong, there are a few errors in there.
Could someone explain me what's behind putting a single forward biased diode in series with the screen? I understand the first article (where a Zener string, AC bypassed, is used), but can't comment on the others.
Thank you in advance, and excuse me if it has already been topic of discussion.
I understand the article, but...
What about this?
http://www.webace.com.au/~electron/tubes/oes.html
http://www.webace.com.au/~electron/tubes/triode.htm
I don't quite understand what this guy says. If I'm not wrong, there are a few errors in there.
Could someone explain me what's behind putting a single forward biased diode in series with the screen? I understand the first article (where a Zener string, AC bypassed, is used), but can't comment on the others.
Thank you in advance, and excuse me if it has already been topic of discussion.
Thank you all!
But this idea
http://www.t-linespeakers.org/tubes/triode-trick.html
doesn't seem so bad to me. I will try it.
I'm also thinking that if we can establish a fixed potential difference between anode and G2, we can up plate voltage (gaining power) mantaining G2 in its ratings.
Or not?
But this idea
http://www.t-linespeakers.org/tubes/triode-trick.html
doesn't seem so bad to me. I will try it.
I'm also thinking that if we can establish a fixed potential difference between anode and G2, we can up plate voltage (gaining power) mantaining G2 in its ratings.
Or not?
SY said:
I'm not familiar with that, but a few years ago there was an article published in an electronics magazine (not audio based) where the author suggested using a floating supply to raise the screen voltage above that of the plate. The claimed result was that the lower plate voltage meant that more current could be run without exceeding the plate dissipation limit. I don't think the author considered screen dissipation, nor did he consider that a simple zener could provide the same level shifting without the capacitive load of a floating supply.
EDN design idea
Yea,
was a boosted triode presented as one EDN Design Idea
by Dave Cuthbert
Here's the link:
http://www.edn.com/contents/images/61203di.pdf
Good to squeeze out all the milliWatts from a chep tube, at a price of power supply complication. The increase expense maybe used to buy a better tube with better and safer results.
Yea,
was a boosted triode presented as one EDN Design Idea
by Dave Cuthbert
Here's the link:
http://www.edn.com/contents/images/61203di.pdf
Good to squeeze out all the milliWatts from a chep tube, at a price of power supply complication. The increase expense maybe used to buy a better tube with better and safer results.
kathodyne said:
Maybe, but not complete BS. If you follow his advice, then you will get the screen voltages of small signal pentodes down, and reducing the screen voltage does improve the performance of pentode based small signal voltage amps.
As for that "Optimized Electron Stream Technology", well, perhaps it works and perhaps it doesn't. Trioding a pentode with zeners and diodes at least won't hurt. Some folks swear by it (even right here) but like anything else, is that for real, or do they only think it improves the sonics because they expect it to improve the sonics?
rdf said:
In fact, I still can't understand WHY should G2 emit electrons. Isn't it designed to NOT allow this? Or maybe only pricey ones do? Maybe cheap ones, cheap tubes, gas in the tube etc can cause the G2 to emit electrons... since it's always negative w.r.t. anode, there should be a steady current from G2 to anode, even with no signal.
Mah...
And thank you Piergiorgio for the link. I see that we're going very near the limits of the tube if we raise G2 voltage w.r.t. anode voltage, maybe going out of G2 dissipation range. It would be interesting to see Ig2 plotted against Ip for Vp << Vg2 in those conditions.
Hi, logging in on a borrowed computer 1200 miles from home:
I read the web page that talked about the "optimized electron stream" stuff. It talks a lot but presents no data to back up the story. I haven't tried it, so I really can't say, but I don't see it working well myself. As for the diode in series with the screen, it sounded terrible on a 6AV5 pentode amp and worse with UL connection.
It is possible for the screen to emit electrons if it is hot (from near limit dissipation) and the plate voltage dips below the screen voltage. This condition will occur on signal peaks in true pentode configuration. Keeping the screen voltage constant (regulating it) usually offers the best tube linearity under these conditions. Nevertheless there are people who claim that the magic diode does wonders for the sound.
I read the article in EDN when it first appeared (we get that magazine at work). Of course I went home and tortured a 6L6, since that particular tube was mentioned in the article. The story implied that if the screen voltage was raised above the plate (triode connection) voltage the tube would conduct heavier at a lower plate voltage, but remain linear. These statements are TRUE. I could indeed operate a 6L6 in triode mode with only 150 volts on the plate. This could be useful in cases where limited voltage was available. As I cranked up the plate supply, which brought the screen voltage up higher, the amp worked, and sounded good, and made good power. At 300 volts on the plate (400 on the screen) the screen grid was glowing red and there was a blue glow inside the tube. Reducing the offset voltage stopped the glow. At plate voltages above 400 I could not apply any extra screen voltage without provoking a display of colors, and at higher plate voltages I found that reversing the offset power supply (operating the screen at a fixed voltage below the plate) allowed the plate voltage to operate at 550 with the screen at 400 volts. This combination allowed me to crank some serious power out of a triode connected 6L6 without frying the screen grid.
This offset screen grid trick shows promise for triode commected sweep tubes which have a low screen grid voltage rating. It can be used in triode or UL connections. The drawback is the requirement for a floating power supply. There have been circuits that attempt to eliminate this requirement by using zener diodes or VR tubes. These all have a problem when the plate voltage drops under a strong signal condition which tries to force the screen grid below ground. This causes the zener or VR tube to stop conducting leading to abrupt distortion. I mentioned this (particularly nasty sounding) distortion in the 6CD6 thread. I was breifly experimenting with a mosfet feeding the screen grid (similar to PowerDrive). This offers the ability to adjust the DC voltage on the screen independently from the AC. It also allows a continuously adjustment from true pentode to triode connection including all ranges of UL tap positions, and does not need a UL tap on the transformer. The drawback is a requirement for a negative supply on some tubes (sweep tubes with a low screen voltage).
This was in the early experiment stages when I left on this trip, and I don't have any way to post a schematic untill I return, but it will show up soon.
I read the web page that talked about the "optimized electron stream" stuff. It talks a lot but presents no data to back up the story. I haven't tried it, so I really can't say, but I don't see it working well myself. As for the diode in series with the screen, it sounded terrible on a 6AV5 pentode amp and worse with UL connection.
It is possible for the screen to emit electrons if it is hot (from near limit dissipation) and the plate voltage dips below the screen voltage. This condition will occur on signal peaks in true pentode configuration. Keeping the screen voltage constant (regulating it) usually offers the best tube linearity under these conditions. Nevertheless there are people who claim that the magic diode does wonders for the sound.
I read the article in EDN when it first appeared (we get that magazine at work). Of course I went home and tortured a 6L6, since that particular tube was mentioned in the article. The story implied that if the screen voltage was raised above the plate (triode connection) voltage the tube would conduct heavier at a lower plate voltage, but remain linear. These statements are TRUE. I could indeed operate a 6L6 in triode mode with only 150 volts on the plate. This could be useful in cases where limited voltage was available. As I cranked up the plate supply, which brought the screen voltage up higher, the amp worked, and sounded good, and made good power. At 300 volts on the plate (400 on the screen) the screen grid was glowing red and there was a blue glow inside the tube. Reducing the offset voltage stopped the glow. At plate voltages above 400 I could not apply any extra screen voltage without provoking a display of colors, and at higher plate voltages I found that reversing the offset power supply (operating the screen at a fixed voltage below the plate) allowed the plate voltage to operate at 550 with the screen at 400 volts. This combination allowed me to crank some serious power out of a triode connected 6L6 without frying the screen grid.
This offset screen grid trick shows promise for triode commected sweep tubes which have a low screen grid voltage rating. It can be used in triode or UL connections. The drawback is the requirement for a floating power supply. There have been circuits that attempt to eliminate this requirement by using zener diodes or VR tubes. These all have a problem when the plate voltage drops under a strong signal condition which tries to force the screen grid below ground. This causes the zener or VR tube to stop conducting leading to abrupt distortion. I mentioned this (particularly nasty sounding) distortion in the 6CD6 thread. I was breifly experimenting with a mosfet feeding the screen grid (similar to PowerDrive). This offers the ability to adjust the DC voltage on the screen independently from the AC. It also allows a continuously adjustment from true pentode to triode connection including all ranges of UL tap positions, and does not need a UL tap on the transformer. The drawback is a requirement for a negative supply on some tubes (sweep tubes with a low screen voltage).
This was in the early experiment stages when I left on this trip, and I don't have any way to post a schematic untill I return, but it will show up soon.
There was another thread a month or 2 ago regarding what to do with G3 in pentodes where this was brought out separately, with quite a bit of discussion on this. Some tried the screen with serie diode and ruined the tube.
Not having tried this myself, I would reserve judgment like a true scientist. But I would agree with SY (post #2) that some very bland assumptions are made along the line, and the serie diode leaving the screen "unattended" under certain conditions, means that its behaviour would be controlled mainly by circuit capacitances - very suspect. I did not read all of the quoted sources (thanks to honourable members for supplying those), but I did not find any figures or graphs quantifying said advantages - perhaps in the parts I skipped. I would be intrigued if this arrangement can provide significant improvement on a well designed UL output stage.
As Dave suspects (post #8), using a higher G2 voltage than plate definitely puts the screen in danger dissipation-wise. (Some wants the g2 voltage to be lower, others say it should be higher - where are we going?)
There is at present another thread "Measuring Triode connected Pentode Curves" broadly dealing with the same topic.
Regards.
Not having tried this myself, I would reserve judgment like a true scientist. But I would agree with SY (post #2) that some very bland assumptions are made along the line, and the serie diode leaving the screen "unattended" under certain conditions, means that its behaviour would be controlled mainly by circuit capacitances - very suspect. I did not read all of the quoted sources (thanks to honourable members for supplying those), but I did not find any figures or graphs quantifying said advantages - perhaps in the parts I skipped. I would be intrigued if this arrangement can provide significant improvement on a well designed UL output stage.
As Dave suspects (post #8), using a higher G2 voltage than plate definitely puts the screen in danger dissipation-wise. (Some wants the g2 voltage to be lower, others say it should be higher - where are we going?)
There is at present another thread "Measuring Triode connected Pentode Curves" broadly dealing with the same topic.
Regards.
i messed about with this, and it definately changes the sound, but not for the better. Harsher and brighter. Elsewhere in the article it suggests adding in more screen resistance. This takes away some of the harshness - nut seems a sticking plaster for a problem created by the diode. I seem to remember seeing imbalance in the waveform with the diode (rectification of the signal maybe).
Bad, bad, bad. Opinion with no science.
Shoog
Bad, bad, bad. Opinion with no science.
Shoog
I don't mean to revive an old thread but what about taking a second silicone diode, reversing the polarity and wiring it in parralel with the first one. that should allow both halves of the wave to pass with the same level of voltage drop. or maybe some sort of ac filter to keep it from ever reaching the diodes.
A small length of wire or a series screen resistor will do a better job of passing both halves of an AC signal. This whole scheme is an unnecessary complication that makes the sonic performance worse, not better.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.