G1=G2/mu Scaled Drive Strawman Design - Page 2 - diyAudio
Go Back   Home > Forums > Amplifiers > Tubes / Valves

Tubes / Valves All about our sweet vacuum tubes :) Threads about Musical Instrument Amps of all kinds should be in the Instruments & Amps forum

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 28th March 2010, 05:56 PM   #11
diyAudio Member
 
Michael Koster's Avatar
 
Join Date: Jan 2008
Location: Eureka, CA
Quote:
Originally Posted by smoking-amp View Post
Here are some pics. These were taken on a timebase scope rather than the curve tracer, so are versus time rather than plate voltage as on the tracer.
(...)

1) cathode current 50mA/div
2) plate current 50 mA/div
3) g1 current 10 mA/div
4) g2 current 10 mA/div

I'm puzzled by the cathode current and plate current traces, should be more similar I would think. Maybe I got the scale wrong on one. Well, at least one can see that the g2 current is the most spikey when plate voltage is low. Looks like g2 baseline current ramping is about half the g1 current ramp. Close to linear current ramps overall for g1 and g2, but some curvature upwards, probably a diode V/I curve. g2 looks more curved. g1 being closer to a straight ramp (resistive Z) is desirable as far as using just one Mosfet driver with a resistive divider for g1.
Thanks all for the comments!

It looks to me like the cathode current is close to the expected sum of all other electrode currents. Peak g1 and peak g2 current are on the same order of magnitude as each other, but G1 seems more linear resistive and g2 acts more like a diode as plate voltage goes lower.

I'm not too worried about the G1 current, but the allowed G1 dissipation is not specified for sweep tubes. I am assuming it's good for at least one watt. The current would need to be >200mA at +20V peak to dissipate anywhere near 1 watt.

I think some trial and error is going to be needed; I will do a lot of in-circuit test and measurement. I will stick with the g1 follower for now so I can easily experiment with varying the DC bias offset on g1.

@kenpeter: I think 40K/10K will drive the Crss of the scaler follower OK. Aren't the Ciss and Coss bootstrapped in this circuit? However there's not a problem lowering the impedance of the divider down below 1K if needed e.g. to drive G1 directly. The mu-gyrators may need a bigger heatsink though...

@wrenchone: If this drive method needs more than 100V peak, the gm will be too low for the local feedback topology in my application. I don't think tubes with low pentode gm (6L6 types) or higher mu (EL34, 6550) will work as well as sweep tubes in this circuit, but 6l6s and 6550s work well enough as pentodes with G2 at B+ anyway.

@George: This does seem to be very little different from the "high impedance triode" mode of operation you describe. This same circuit without the extra g1 scaler/follower will drive high impedance triodes and give low effective output resistance.

@Anatoliy: I have never understood the popularity of ultralinear. By the time you add enough NFB to get a decent output resistance you can consider a number of other options like local Schade feedback *instead*. I suppose I need to read H&K again ... or not ;-)

Cheers,

Michael

Last edited by Michael Koster; 28th March 2010 at 06:15 PM.
  Reply With Quote
Old 28th March 2010, 09:04 PM   #12
diyAudio Member
 
smoking-amp's Avatar
 
Join Date: Dec 2001
Location: Hickory, NC
Here is a plot of actual g2 and g1 voltages using just the resistive divider to drive g1. I adjusted the divider to get twice the current level of plate curves with g1 and g2 versus g2 only. Then I checked the divider resistance, and its running at 268 Ohm from g2 to g1 and 100 Ohm from g1 to cathode. Tube Mu is nominally 4.2. So would expect 300 Ohm and 100 Ohm if g1 weren't drawing current. Can see from the plot that g1 voltage is drooping at higher stepping levels. I adjusted the attenuator on the scope to get matched 1st step levels between them. Or put a ruler to the g1 steps and see the droop.
Scale on the g2 curve is 5 V per div. according to the scope (but 5.3 V steps according to the curve tracer, someones calibration?), g1 curve should be about 1/4.2 of that, the variable attenuator was used on g1 to get a match.

I just checked the tracer step output using a DVM and get 3 V steps. Looks like the scope is right. Which means I've only been getting 30 V max for 10 steps.

Ugh, just found the problem, the variable step gain control on the tracer wasn't at its max setting.

I'm going to set up a Mosfet follower for g1 next.
Attached Images
File Type: jpg 6HJ5_g2_g1_current.jpg (38.0 KB, 287 views)
__________________
I want a Huvr-Board!!

Last edited by smoking-amp; 28th March 2010 at 09:29 PM.
  Reply With Quote
Old 28th March 2010, 09:50 PM   #13
diyAudio Member
 
smoking-amp's Avatar
 
Join Date: Dec 2001
Location: Hickory, NC
Re-did the g2 and g1 voltage plot for 6HJ5 with the max step setting of 5.3 V steps per the tracer (6V steps per the DVM, and about 5.9 V steps according to the scope). Vertical scale here is 10 V per div. (scope) for g2. This is using the same resistive divider for g1 again.

The jaggies in the g1 curves on each step are from when plate voltage is near zero (occurs on leading or falling edge depending on the phase of the 60 Hz plate voltage).

Clearly, g1 voltages are drooping off with high g1 current and/or low plate voltage, when just using a resistive divider.

OK, on to getting a Mosfet driver set up for g1 finally.
Attached Images
File Type: jpg 6HJ5_g2_g1_current.jpg (36.6 KB, 281 views)
__________________
I want a Huvr-Board!!

Last edited by smoking-amp; 28th March 2010 at 09:58 PM.
  Reply With Quote
Old 28th March 2010, 11:48 PM   #14
diyAudio Member
 
smoking-amp's Avatar
 
Join Date: Dec 2001
Location: Hickory, NC
Here is Mu scaled g2,g1 6HJ5 with a Mosfet driver for g1 now. 50 mA/div vert. and 50 V/div horiz. 5.3 V steps on g2. Approx 4/1 scaling. The divider is still between g2 and cathode here, but g1 is buffered by a Mosfet follower (which has a few volt gate threshold unfortunately)

2nd plot is the earlier resistive divider curves. Same setup except no Mosfet driver.

Looks like the Mosfet driven g1 curve set is more squared up (no more g1 sag at low plate V). The lower bunched together plate curves are due to the Mosfet turning off below its gate threshold, so only g2 drive there. I'll have to rig up some DC biasing for the g1/Mosfet next, to fix the gate threshold problem.
Attached Images
File Type: jpg 6HJ5_Mosg1_g2.jpg (26.3 KB, 273 views)
File Type: jpg 6HJ5_g2g1drive.jpg (27.3 KB, 273 views)
__________________
I want a Huvr-Board!!

Last edited by smoking-amp; 28th March 2010 at 11:59 PM.
  Reply With Quote
Old 29th March 2010, 01:02 AM   #15
diyAudio Member
 
tubelab.com's Avatar
 
Join Date: Jul 2005
Location: West Virginia panhandle
Quote:
I have never understood the popularity of ultralinear. By the time you add enough NFB to get a decent output resistance you can consider a number of other options
I try to keep an open mind toward most things involving audio and electronics. 5 or 6 years ago when I was advocating the use of a mosfet follower for grid drive the response from the tube audiophile community was overwhelmingly negative. No look how far we have come.

I don't have a strong opinion on UL but I have found applications where it works well. Consider my Simple SE design. It is a simple circuit using a CCS loaded triode (half a 12AT7) driving a pentode (EL34, 6L6GC or KT88). The output stage can be connected in triode, UL, or pentode via jumpers or a switch. Local cathode feedback can be enabled via a jumper or a switch. 90% of the time I have the amp in triode mode without feedback using KT88's. There are times however where UL with cathode feedback just sounds better. These times usually involve music like Pink Floyd or Depeche Mode. Other Simple SE users have come to similar conclusions.

Quote:
This does seem to be very little different from the "high impedance triode" mode of operation you describe.
I like the curves you post better than the typical high Mu beam triodes like the 6HV5. Unfortunately I didn't get to experiment this weekend. Instead I spent the weekend outside making my yard look better than the empty forclosures on the block. The weather radar has been green, yellow, and red all around us, but I only got wet once today.

Even though I was outside working for most of the weekend, by brain kept dreaming up circuits. For now my plan will be to use a pentode from the dollar menu as a driver / splitter. Multiple mosfet buffers will feed each grid. Schade style feedback will be used. Dollar menu sweep tubes and high Mu beam triodes are the initial "test subjects".
__________________
Too much power is almost enough! Turn it up till it explodes - then back up just a little.
  Reply With Quote
Old 29th March 2010, 02:31 AM   #16
diyAudio Member
 
smoking-amp's Avatar
 
Join Date: Dec 2001
Location: Hickory, NC
Here is the scaled g2,g1 6HJ5 using Mosfet drive on g1, with DC biasing to fix up the Mosfet gate threshold.

I ran into a problem here though using a cap to couple divided down g2 signal to the Mosfet gate. A protection zener diode across the Mosfet gate to source conducts when the drive signal is large enough, changing the charge on the cap (a blocking problem). I have a 2.7K Ohm resistor loading the Mosfet source to 0 V at the cathode, which provides the conduction path when AC drives it below 0 V. I was able to adjust the bias upwards (half of AC signal approx.) to get things to work, but if I change the # of steps on the curve tracer, I have to re-adjust the DC bias every time. Not good for real signals.

SOooo, either I have to return the Mosfet loading resistor to a negative (more than -1/2 max peak drive AC on g1) voltage to avoid blocking, or go with some type of DC coupling offseted to fix the Mosfet gate threshold.

Another problem also came up too. I had to operate with reduced g1 AC (1/6 of g2 AC here for the plot shown) in order to avoid some kind of saturation problem of the tube plate current. (I had plenty, 36V, of g1 Mosfet drain voltage to avoid saturation there) The top curves start bunching together (at the low plate voltage end of the curves) if I increase g1 drive beyond 1/6 of g2. Not sure what the problem is yet. Maybe g2 starts eating the plate current too much? More testing in order.
Attached Images
File Type: jpg 6HJ5_Mosg1bias_g2.jpg (27.1 KB, 264 views)
__________________
I want a Huvr-Board!!
  Reply With Quote
Old 29th March 2010, 02:59 AM   #17
diyAudio Member
 
smoking-amp's Avatar
 
Join Date: Dec 2001
Location: Hickory, NC
edit (late):
The bunching effect at high plate currents may just be a visual artifact due to the curve tracer load limiting resistor not allowing/showing more than the low plate voltage part of the curves there. I also couldn't see any particular pick up in screen currents there either with the divider cranked up and a current probe. When I switched to a 140 Ohm load resistor (220 Watt limit) on the tracer from the usual 650 Ohm (50 Watt limit), the bunching effect disappeared. (The cardboard the tube was sitting on started smoking though!) More testing in order.

HeHe:
6HJ5 at 220 Watt peak limiting plot attached, scaled g2g1 drive. 100 mA/div vertical. Note that the increased g1 drive here is causing the knees of the plate curves move to higher plate voltages than before. The g2,g1 divider was at 82 Ohms from g2 to g1 pickoff, and 100 Ohms from there to ground.
Attached Images
File Type: jpg 6HJ5_g2g1_220Watt.jpg (26.6 KB, 16 views)
__________________
I want a Huvr-Board!!

Last edited by smoking-amp; 29th March 2010 at 03:15 AM.
  Reply With Quote
Old 29th March 2010, 04:03 AM   #18
diyAudio Member
 
smoking-amp's Avatar
 
Join Date: Dec 2001
Location: Hickory, NC
Since Michael is using 6DQ6 outputs, I dug out some 12GE5s which are essentially the same tube as 6DQ6B or 6JN6A. This tube has a Mu of 4.4, 17.5 Watt diss. rating.

1st plot is operating in scaled g2,g1 mode with the Mu divider at 4.4 and using a Mosfet driver with gate biasing. Scale is 50 V/div horiz. and 50 mA/div vert. 5.3 V/div stepping. (usual 50 Watt peak load limiting on the tracer)

2nd plot is with 220 Watt peak load limiting on the tracer (140 Ohm load resistor), same vert and horiz scales, same 5.3 V steps and Mosfet g1 driver. Mu divider was 79 Ohms from g2 to g1 pickoff and 100 Ohms from there to cathode.

Hey, you ought to be able to get 400 Watts out of these in P-P for a few seconds at least.
Attached Images
File Type: jpg 12GE5_g2g1_Mosfetdriver.jpg (24.7 KB, 13 views)
File Type: jpg 12GE5_g2g1_220W.jpg (27.3 KB, 17 views)
__________________
I want a Huvr-Board!!

Last edited by smoking-amp; 29th March 2010 at 04:07 AM.
  Reply With Quote
Old 29th March 2010, 04:15 AM   #19
diyAudio Member
 
tubelab.com's Avatar
 
Join Date: Jul 2005
Location: West Virginia panhandle
Quote:
Hey, you ought to be able to get 400 Watts out of these in P-P for a few seconds at least.
That sounds like a challenge to me. Well I got two $1 tubes wired together as a phase splitter. With 2 volts into the stage, I get two out of phase signals at about 350 V P-P. Buffer that with mosfets and I should be able to melt any grid ever made! I may have some ugly looking old 6DQ6's somewhere.
__________________
Too much power is almost enough! Turn it up till it explodes - then back up just a little.
  Reply With Quote
Old 29th March 2010, 04:28 AM   #20
diyAudio Member
 
smoking-amp's Avatar
 
Join Date: Dec 2001
Location: Hickory, NC
I'll be listening for the News: "Fireball envelops South Florida"
If you want some cheap 6BQ6GAs, I've got a box of them (40 or 50) from the last tube sale, I'll probably never use them now. Too big for fishing bobbers.
__________________
I want a Huvr-Board!!
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
screen-drive (g2 driven) plate curves spacing linear! oldeurope Tubes / Valves 27 26th June 2008 07:43 AM
HK 'Deuce front end on G2 drive output stage? GordonW Tubes / Valves 0 9th June 2008 03:22 PM
6JN6 screen-drive (g2 driven) plate curves? GordonW Tubes / Valves 26 23rd April 2008 02:56 PM
Anyone know the g1-g2 capacitance of 6AU6? ray_moth Tubes / Valves 10 29th August 2004 08:19 AM


New To Site? Need Help?

All times are GMT. The time now is 12:23 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2