If you are talking of screen drive for output valves then the main advantage is better linearity. This is because the screen and anode are likely to be of similar shape so provide the right conditions for constant mu operation i.e. ideal triode. The big disadvantage is that you need a beefy driver.
That would be a rather high Rp triode with g2 as the input. The output Z is still the same as in pentode mode. Some local neg. feedback can fix that however.
Some 6LG6 typical TV sweep tube curves for evaluation:
http://frank.pocnet.net/sheets/123/6/6LG6.pdf
On page 3, the top curves are for g1 drive, and the bottom curves are for g2 drive. Notice how much more "equal" the spacings are for the bottom g2 curves. But they still have the same Rp slope to them as the usual pentode curves. And notice how much more g2 voltage variation there is across their curve set than for the g1 curve set. g2 operation closely conforms to 3/2 power law of Ip versus Vg2 (which conforms to the plate back effects), whereas g1 operation is more like square law Ip versus Vg1 (over a limited range) due to grid wire proximity (to the cathode) effects.
I usually hear that the g2 drive is more suited to class aB (mostly B) operation, where it gives good efficiency, which in turn can provide more power output or cooler operation at normal output.
The danger with g2 drive, especially for high power output, is of heating up the screen grid when it has high positive drive voltage applied, and the plate voltage is dipping well below it (heavy screen grid current). And as George (Tubelab) has determined by extensive testing, this can cause grid 2 electron emission when the plate swings back up above B+ (and possibly g2 swings negative also), and catastrophic meltdown ensues. A blocking diode in the g2 circuit might help.
Some ideas for combined/scaled g2 and g1 grid drives have been hashed around some to avoid this safety issue (could halve the g2 voltage drive requirement), still evolving, nothing much published in the way of a tested amplifier yet.
Some 6LG6 typical TV sweep tube curves for evaluation:
http://frank.pocnet.net/sheets/123/6/6LG6.pdf
On page 3, the top curves are for g1 drive, and the bottom curves are for g2 drive. Notice how much more "equal" the spacings are for the bottom g2 curves. But they still have the same Rp slope to them as the usual pentode curves. And notice how much more g2 voltage variation there is across their curve set than for the g1 curve set. g2 operation closely conforms to 3/2 power law of Ip versus Vg2 (which conforms to the plate back effects), whereas g1 operation is more like square law Ip versus Vg1 (over a limited range) due to grid wire proximity (to the cathode) effects.
I usually hear that the g2 drive is more suited to class aB (mostly B) operation, where it gives good efficiency, which in turn can provide more power output or cooler operation at normal output.
The danger with g2 drive, especially for high power output, is of heating up the screen grid when it has high positive drive voltage applied, and the plate voltage is dipping well below it (heavy screen grid current). And as George (Tubelab) has determined by extensive testing, this can cause grid 2 electron emission when the plate swings back up above B+ (and possibly g2 swings negative also), and catastrophic meltdown ensues. A blocking diode in the g2 circuit might help.
Some ideas for combined/scaled g2 and g1 grid drives have been hashed around some to avoid this safety issue (could halve the g2 voltage drive requirement), still evolving, nothing much published in the way of a tested amplifier yet.
Last edited:
The other advantage is efficiency. Screen drive amplifiers can be run near class B with excellent results.
Disadvantage is the need for a driver stage that can swing lots of volts and is unaffected by the non-linear current draw of the screen grid. If you're curious about how to engineer a screen drive stage properly, go to the Support section of David Berning's website and look at the schematics for the old BA-150 (Audionics) and the Berning EA-230.
Disadvantage is the need for a driver stage that can swing lots of volts and is unaffected by the non-linear current draw of the screen grid. If you're curious about how to engineer a screen drive stage properly, go to the Support section of David Berning's website and look at the schematics for the old BA-150 (Audionics) and the Berning EA-230.
Thanks all,
I was just curious. I have read somewhere that screen connected was close to a true triode... so I tried it for a preamp...it gives lower gain... However it seems that this connection is better for output applications...and not Single Ended output.
Maybe its time to invest in 2 opt's. I don't need many watts so SE will be fine...and I can get replacement tubes for a fraction of the EL84's my ST-35 takes.
I was just curious. I have read somewhere that screen connected was close to a true triode... so I tried it for a preamp...it gives lower gain... However it seems that this connection is better for output applications...and not Single Ended output.
Maybe its time to invest in 2 opt's. I don't need many watts so SE will be fine...and I can get replacement tubes for a fraction of the EL84's my ST-35 takes.
I am usually very forthcoming with my discoveries but I have seen some of the circuits that I designed be copied and sold by others before I can prove to myself that they are worthy of the Tubelab name. Life has a way of getting in the way of fun sometimes, and my recent "retirement" has me packing up Tubelab and moving everything out of Florida. It's a few years sooner than anticipated, but I will make it work. It could be a year or more before I have a lab set up, so I must remain silent for now.
I have seen over 100 watts of clean power flow from a pair of $1 tubes for over an hour. The same amp sounds quite good playing music, very loud music, and positively screams when driven to 140 watts of severely overdriven guitar. The driver was the exact same board I laid out for the 6L6GC in AB2 thread about 5 years ago (2 6SN7 LTP's).
I designed 3 different versions of G1-G2 drive and simulated them in LTspice. I built 2 and both work fairly close to the simulations. All I am willing to divulge at this point is that it works, and really isn't that hard to make work. All of the individual elements have been discussed on this forum.
It sounds and simulates more like a pentode to me....high output impedance and all.
I have made lots of Zero Bias 811/572/805 amps and they certainly don't sound like any Pentode amps; I have heard,.
Best sounding amps I have ever made; they can drive any speakers even speakers big transistor amps struggle with. The Russian 811A's last not like KT88's. The 811 amps run at over 100 watts with only 600 volts and 70 mills current..
Phil
Best sounding amps I have ever made; they can drive any speakers even speakers big transistor amps struggle with. The Russian 811A's last not like KT88's. The 811 amps run at over 100 watts with only 600 volts and 70 mills current..
Phil
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.