I re-did the 36MC6 with the bunched curves left in and the screen V increased.
100K plate Fdbk resistor and 22K grid feed resistor for all cases. 20 uF cap in P Fdbk too.
1) 6L6WXT+ in pseudo triode 20 mA/div Vert., 50V/div Horiz, 3.75 V/step
2) 6L6WXT+ in Schade 20 mA/div Vert., 50 V/div Horiz, 8 V/step, with 230 V on g2
3) 36MC6 in pseudo triode 50mA/div Vert, 50V/div Horiz., 7.5 V/step
4) 36MC6 in Schade with 108V on g2, 50 mA/div Vert., 50V/div Horiz., 7.5 V/step
The Vg1 = 0 is indeterminate on the Schade pics. My adjusting Vbias on the tracer only ends up as the approx. center trace since the AC coupled 1/2 Sine Wave plate voltage gets attenuated, and cap coupled with it before going to the grid.
Again, it looks like the Schade models have more even spacing and less roll over at HV.
-----------------------------
On another note, I recently got a 36MC6 off Ebay that has the screen grid maximally mis-aligned with the g1 grid wires (ie, 50% in between) It overheats quickly on the tracer with 2.5X normal g2 current initially, causing the plate curves to collapse as the g2 current climbs into meltdown territory. It is recoverable after cooling though, and repeatable. Made perfectly wrong.
I think I will plot the cathode current versus g2 voltage to see if it has a 1.4 or 1.5 power law (I = kV^x) . I have observed 1.5 power law for plates and for non aligned g2 (with different winding pitch), and a 1.4 power law for aligned g2 grids. So I would expect a 1.5 power law curve here.
They didn't ask for it back. So maybe I should give it the TubeLab treatment.
100K plate Fdbk resistor and 22K grid feed resistor for all cases. 20 uF cap in P Fdbk too.
1) 6L6WXT+ in pseudo triode 20 mA/div Vert., 50V/div Horiz, 3.75 V/step
2) 6L6WXT+ in Schade 20 mA/div Vert., 50 V/div Horiz, 8 V/step, with 230 V on g2
3) 36MC6 in pseudo triode 50mA/div Vert, 50V/div Horiz., 7.5 V/step
4) 36MC6 in Schade with 108V on g2, 50 mA/div Vert., 50V/div Horiz., 7.5 V/step
The Vg1 = 0 is indeterminate on the Schade pics. My adjusting Vbias on the tracer only ends up as the approx. center trace since the AC coupled 1/2 Sine Wave plate voltage gets attenuated, and cap coupled with it before going to the grid.
Again, it looks like the Schade models have more even spacing and less roll over at HV.
-----------------------------
On another note, I recently got a 36MC6 off Ebay that has the screen grid maximally mis-aligned with the g1 grid wires (ie, 50% in between) It overheats quickly on the tracer with 2.5X normal g2 current initially, causing the plate curves to collapse as the g2 current climbs into meltdown territory. It is recoverable after cooling though, and repeatable. Made perfectly wrong.
I think I will plot the cathode current versus g2 voltage to see if it has a 1.4 or 1.5 power law (I = kV^x) . I have observed 1.5 power law for plates and for non aligned g2 (with different winding pitch), and a 1.4 power law for aligned g2 grids. So I would expect a 1.5 power law curve here.
They didn't ask for it back. So maybe I should give it the TubeLab treatment.
Attachments
Last edited:
Thanks for doing these plots. It is nice to see things work out in real life like they do on paper.
Can your tracer trace up to higher voltages? I'm just curious how much the curves start to lay down at low currents and higher voltages with the feedback.
Another note about the curve bunching near saturation: It might be hard on g1, but if you were to drive the triode connected examples with positive g1 voltages, you should see the same curve bunching that you see with the pentode and feedback.
You can see it a bit in these data sheets:
KT88
10
812A
It's just what happens when you approach saturation.
Can your tracer trace up to higher voltages? I'm just curious how much the curves start to lay down at low currents and higher voltages with the feedback.
Another note about the curve bunching near saturation: It might be hard on g1, but if you were to drive the triode connected examples with positive g1 voltages, you should see the same curve bunching that you see with the pentode and feedback.
You can see it a bit in these data sheets:
KT88
10
812A
It's just what happens when you approach saturation.
I can get up to 1500 V on the plate, but the grid drive is what is limiting me, only about 80V available there.
Best I could get:
1) 6L6WXT+ pseudo triode with positive going g1, 10 mA/div Vert, 50 V/div Horiz., 8 V/step
2) 6L6WXT+ pseudo triode 10mA/div Vert, 50 V/div Horiz, 8 V/step
3) 6L6WXT+ Schade 10 mA/div Vert, 50 V/div Horiz, 8 V/step
Hmmm, maybe was less than 8V /step on the triode mode ones, one of them was giving me a problem with the trace just going horizontal suddenly at HV. The Schade ones look correctly scaled (Mu).
Ohh, the higher Mu in psuedo triode limits the V range, but there is an extra trace there too. I guess I could have gotten 1 more trace in the Schade mode.
Well, I think Schade wins over triode.
Best I could get:
1) 6L6WXT+ pseudo triode with positive going g1, 10 mA/div Vert, 50 V/div Horiz., 8 V/step
2) 6L6WXT+ pseudo triode 10mA/div Vert, 50 V/div Horiz, 8 V/step
3) 6L6WXT+ Schade 10 mA/div Vert, 50 V/div Horiz, 8 V/step
Hmmm, maybe was less than 8V /step on the triode mode ones, one of them was giving me a problem with the trace just going horizontal suddenly at HV. The Schade ones look correctly scaled (Mu).
Ohh, the higher Mu in psuedo triode limits the V range, but there is an extra trace there too. I guess I could have gotten 1 more trace in the Schade mode.
Well, I think Schade wins over triode.
Attachments
Last edited:
Here is a low gm DHP (dual Pentode, paralleled here, total gm = 3000), the 1E7G,
which has remarkably nice pseudo triode curves.
Lets see how it compares with Schade mode, which may be at some disadvantage here with the lowish gm.
1) Pseudo triode 10 mA/div Vert, 50 V/div Horiz, 8 V/step
2) Shade mode 10 mA/div Vert, 50 V/div Horiz, 8 V/step, 223 V on g2
Both look quite good. (the one mis-positioned trace in the Schade picture is from a retrace interaction with the camera, the correct trace is dimly visible in the proper position)
This is a tube I have considered for use in a tube emulator circuit (as the triode model). A differential tube pair driving a single Mosfet output, with N feedback of the Drain back thru the plate of the 1E7G (on one side of the differential pair). The Mosfet has to emulate the tube model accurately.
------------
which has remarkably nice pseudo triode curves.
Lets see how it compares with Schade mode, which may be at some disadvantage here with the lowish gm.
1) Pseudo triode 10 mA/div Vert, 50 V/div Horiz, 8 V/step
2) Shade mode 10 mA/div Vert, 50 V/div Horiz, 8 V/step, 223 V on g2
Both look quite good. (the one mis-positioned trace in the Schade picture is from a retrace interaction with the camera, the correct trace is dimly visible in the proper position)
This is a tube I have considered for use in a tube emulator circuit (as the triode model). A differential tube pair driving a single Mosfet output, with N feedback of the Drain back thru the plate of the 1E7G (on one side of the differential pair). The Mosfet has to emulate the tube model accurately.
------------
Attachments
Last edited:
I keep comparing to the 300B, but this plot has less curve lean-over at high voltages, more gain, and lower plate resistance. How much does a 36MC6 cost compared to a 300B?
That 36MC6 is the best one of 30 that I have, so don't expect them all to be that good. I got mine off Ebay over the last year for around $8 or $9 each + shipping. Some were duds though. But the ones still listed there are mostly up around $20 except for some Raytheon (Japan) or used looking ones.
The Zenith ones listed in quantity on Ebay are actually GE ones (they are missing the little winglets on the sides of the plate), and the two GE ones I have are not that great. Sylvania is the super good one I have, but most of the Sylvania and RCA ones I have are fairly similar and not so super. I've been hoping to find another super one, but at $8 or $9. I'll take one of them over a 300B any day. But if I could get consistently super ones at the going price I would be happy too.
36MC6/6MC6 seems to be an RCA designed clone of the 6LX6 tube. I only have one of the 26LX6, and they are more expensive on Ebay, but similar price on the lists below.
36MC6 are $18 here:
30 to 698
and $17 here:
Tube List range 13CW40 - 90C1 @ ESRCVacuumTubes.com
On the other hand, I haven't tried the less than super ones in Schade mode yet, and high gm is the key there. So maybe they will look good that way too. Have to try some yet.
For high gm though, 35LR6 is one of the best, and I have yet to try one in Schade mode yet. Will let you know how that looks. They are a bit more plentiful I think.
For the most gm and Watts for the buck, 6HJ5 is the real winner (24 Watt and no plate cap and $5). I have to try them out too. I have a nice box of them here. They do have nice "triode" curves.
Another one I have to check is 21HB5, these are high gm and only 18 Watt, but the plate is the same size as other 24 Watt tubes. (no plate cap) These were on the $1 list for years. I have a BIG box of them.
Another dark horse is 6CB5, cheap 24 Watt, good triode curves. Got some here to try.
So I'm going to be busy here in Schade mode for the next week I think.
-----------------
The Zenith ones listed in quantity on Ebay are actually GE ones (they are missing the little winglets on the sides of the plate), and the two GE ones I have are not that great. Sylvania is the super good one I have, but most of the Sylvania and RCA ones I have are fairly similar and not so super. I've been hoping to find another super one, but at $8 or $9. I'll take one of them over a 300B any day. But if I could get consistently super ones at the going price I would be happy too.
36MC6/6MC6 seems to be an RCA designed clone of the 6LX6 tube. I only have one of the 26LX6, and they are more expensive on Ebay, but similar price on the lists below.
36MC6 are $18 here:
30 to 698
and $17 here:
Tube List range 13CW40 - 90C1 @ ESRCVacuumTubes.com
On the other hand, I haven't tried the less than super ones in Schade mode yet, and high gm is the key there. So maybe they will look good that way too. Have to try some yet.
For high gm though, 35LR6 is one of the best, and I have yet to try one in Schade mode yet. Will let you know how that looks. They are a bit more plentiful I think.
For the most gm and Watts for the buck, 6HJ5 is the real winner (24 Watt and no plate cap and $5). I have to try them out too. I have a nice box of them here. They do have nice "triode" curves.
Another one I have to check is 21HB5, these are high gm and only 18 Watt, but the plate is the same size as other 24 Watt tubes. (no plate cap) These were on the $1 list for years. I have a BIG box of them.
Another dark horse is 6CB5, cheap 24 Watt, good triode curves. Got some here to try.
So I'm going to be busy here in Schade mode for the next week I think.
-----------------
Last edited:
Here is the 12GE5 (same as 6JN6, 6GV5, 6FW5, 6DQ6B )
The 6JN6 was used in Pete Millett's DCPP "red board", "Engineer's Amplifier" which was quite popular and it used Schade mode for the output tubes too.
1) pseudo triode 50 ma/div Vert, 50V/div Horiz, approx. 8 V/step
2) Schade mode 50 mA/div Vert, 50V/div Horiz, approx 8V/step, 150 V on g2
Obvious that the curve spacing is more uniform in Schade mode.
The 6JN6 was used in Pete Millett's DCPP "red board", "Engineer's Amplifier" which was quite popular and it used Schade mode for the output tubes too.
1) pseudo triode 50 ma/div Vert, 50V/div Horiz, approx. 8 V/step
2) Schade mode 50 mA/div Vert, 50V/div Horiz, approx 8V/step, 150 V on g2
Obvious that the curve spacing is more uniform in Schade mode.
Attachments
Last edited:
One thing to keep in mind about making the tubes super linear, is if you are going to run them in class aB, where they will overlap partially. Then it could be useful to have some drop off in gain in the overlap area, otherwise one will have gm doubling in the overlap, causing crossover distortion. Certainly some attention required for optimum biasing and a comparison needed of the Schade versus psuedo triode in P-P class aB.
Last edited:
Some more "triode" and "Schade" curves. These are all 50 mA/div Vert. and 50 V/div horiz. and 7 V/step
and for Schade mode: 100K Ohm plate Neg. Fdbk R, and 22K Ohm grid 1 feed R and 20 uF Cap in the plate Fdbk.
36MC6 Schade plots used 171 V on g2
6HJ5 Schade plot used 200 V on g2
1) 36MC6 average type tube in triode
2) 36MC6 average type tube in Schade
3) 36MC6 selected Super tube in triode
4) 36MC6 selected Super tube in Schade
5) 6HJ5 average type tube in triode
6) 6HJ5 average type tube in Schade
I suspect one would find a "Super" version of most any tube type if one selected through 30 or 40 tubes, as I certainly see some variation between tubes.
On the 36MC6 tubes, it now appears that the "Sylvania" marked tubes are actually just remarked RCA tubes, exactly the same construction and using RCA's signature dark heater wires on both. The GE tubes are different though.
and for Schade mode: 100K Ohm plate Neg. Fdbk R, and 22K Ohm grid 1 feed R and 20 uF Cap in the plate Fdbk.
36MC6 Schade plots used 171 V on g2
6HJ5 Schade plot used 200 V on g2
1) 36MC6 average type tube in triode
2) 36MC6 average type tube in Schade
3) 36MC6 selected Super tube in triode
4) 36MC6 selected Super tube in Schade
5) 6HJ5 average type tube in triode
6) 6HJ5 average type tube in Schade
I suspect one would find a "Super" version of most any tube type if one selected through 30 or 40 tubes, as I certainly see some variation between tubes.
On the 36MC6 tubes, it now appears that the "Sylvania" marked tubes are actually just remarked RCA tubes, exactly the same construction and using RCA's signature dark heater wires on both. The GE tubes are different though.
Attachments
Last edited:
Some more "triode" and "Schade" curves. These are all 50 mA/div Vert. and 50 V/div horiz. and 7 V/step
and for Schade mode: 100K Ohm plate Neg. Fdbk R, and 22K Ohm grid 1 feed R and 20 uF Cap in the plate Fdbk.
1) 35LR6 average triode
2) 35LR6 average Schade
3) 21HB5A average triode
4) 21HB5A average Schade
5) 21LG6 average triode
6) 21LG6 average Schade
and for Schade mode: 100K Ohm plate Neg. Fdbk R, and 22K Ohm grid 1 feed R and 20 uF Cap in the plate Fdbk.
1) 35LR6 average triode
2) 35LR6 average Schade
3) 21HB5A average triode
4) 21HB5A average Schade
5) 21LG6 average triode
6) 21LG6 average Schade
Attachments
More "triode" and "Schade" curves. These are all 50 mA/div Vert. and 50 V/div horiz. and 7 V/step
and for Schade mode: 100K Ohm plate Neg. Fdbk R, and 22K Ohm grid 1 feed R and 20 uF Cap in the plate Fdbk.
This is a comparison between 26LX6 and 36MC6, supposedly equivalent.
1) 26LX6 triode
2) 26LX6 Schade
3) 36MC6 average triode
4) 36MC6 average Schade
5) 36MC6 selected super triode
6) 36MC6 selected super Schade
Looks like the 26LX6 has a somewhat lower internal Mu than the 36MC6. The Schade curves come out the same. I only have one 26LX6, so I can't really say whether it is representative of the average tube, but it is similar to the super 36MC6 as far as curve roll-over in triode mode.
I do have a pair of 26HU5, which is supposed to be equivalent to 26LX6, so I'll take a look at that tomorrow when I get to the octals: 6CB5A, 36LW6, 6DQ5
and for Schade mode: 100K Ohm plate Neg. Fdbk R, and 22K Ohm grid 1 feed R and 20 uF Cap in the plate Fdbk.
This is a comparison between 26LX6 and 36MC6, supposedly equivalent.
1) 26LX6 triode
2) 26LX6 Schade
3) 36MC6 average triode
4) 36MC6 average Schade
5) 36MC6 selected super triode
6) 36MC6 selected super Schade
Looks like the 26LX6 has a somewhat lower internal Mu than the 36MC6. The Schade curves come out the same. I only have one 26LX6, so I can't really say whether it is representative of the average tube, but it is similar to the super 36MC6 as far as curve roll-over in triode mode.
I do have a pair of 26HU5, which is supposed to be equivalent to 26LX6, so I'll take a look at that tomorrow when I get to the octals: 6CB5A, 36LW6, 6DQ5
Attachments
Last edited:
More "triode" and "Schade" curves. These are all 50 mA/div Vert. and 50 V/div horiz. and 7 V/step
and for Schade mode: 100K Ohm plate Neg. Fdbk R, and 22K Ohm grid 1 feed R and 20 uF Cap in the plate Fdbk.
1) 6LB6 in triode
2) 6LB6 in Schade
3) 24JE6A/24LQ6 in triode
4) 24JE6A/24LQ6 in Schade
next is a $2.95 30MB6 off Ebay, right away I'm suspicious of the quality
5) 30MB6 in triode
6) 30MB6 in Schade
7) 30MB6 in Beam pentode
There are very noticeable kinks in the 30MB6 Schade curves, the triode curves don't look so hot either. So then I did the pentode curves to see what was up, and this tube has BIG kinks from screen grid current. No wonder it was $2.95, someone's reject. I inspected the tube through the inspection windows in the plate, and the grids are well aligned, and well centered on the cathode. I don't know what the problem is.... Locating the Hammer.
and for Schade mode: 100K Ohm plate Neg. Fdbk R, and 22K Ohm grid 1 feed R and 20 uF Cap in the plate Fdbk.
1) 6LB6 in triode
2) 6LB6 in Schade
3) 24JE6A/24LQ6 in triode
4) 24JE6A/24LQ6 in Schade
next is a $2.95 30MB6 off Ebay, right away I'm suspicious of the quality
5) 30MB6 in triode
6) 30MB6 in Schade
7) 30MB6 in Beam pentode
There are very noticeable kinks in the 30MB6 Schade curves, the triode curves don't look so hot either. So then I did the pentode curves to see what was up, and this tube has BIG kinks from screen grid current. No wonder it was $2.95, someone's reject. I inspected the tube through the inspection windows in the plate, and the grids are well aligned, and well centered on the cathode. I don't know what the problem is.... Locating the Hammer.
Attachments
Last edited:
Just for reference, here are some beam pentode curves for the tubes so far. Notice how much better the kinks are in these (than that flaky 30MB6 up above). These are all on a 50 mA/div Horiz., 50 V/div Vert scale, except for the 6L6WXT which has 20 mA/div Vert.
1) 6L6WXT+
2) 12GE5 (6JN6...)
3) 36MC6
4) 26LX6
5) 6HJ5
6) 35LR6
7) 21HB5A
8) 21LG6
9) 6LB6
10) 24JE6A/24LQ6
1) 6L6WXT+
2) 12GE5 (6JN6...)
3) 36MC6
4) 26LX6
5) 6HJ5
6) 35LR6
7) 21HB5A
8) 21LG6
9) 6LB6
10) 24JE6A/24LQ6
Attachments
Comparing the 26HU5 with the 26LX6 now. Supposedly the same tube.
1) 26LX6 in triode
2) 26LX6 in Schade
3) 26LX6 in B. pentode
4) 26HU5 in triode
5) 26HU5 in Schade
6) 26HU5 in B. pentode
a 2nd 26HU5, not as good in triode or Schade:
7) 26HU5 #2 in triode
8) 26HU5 #2 in Schade
9) 26HU5 #2 in B. pentode
10) 36MC6 (average tube) in triode
Well, the better 26HU5 triode is close (a bit better actually) to the average 36MC6 triode.
1) 26LX6 in triode
2) 26LX6 in Schade
3) 26LX6 in B. pentode
4) 26HU5 in triode
5) 26HU5 in Schade
6) 26HU5 in B. pentode
a 2nd 26HU5, not as good in triode or Schade:
7) 26HU5 #2 in triode
8) 26HU5 #2 in Schade
9) 26HU5 #2 in B. pentode
10) 36MC6 (average tube) in triode
Well, the better 26HU5 triode is close (a bit better actually) to the average 36MC6 triode.
Attachments
Last edited:
Lets see how the better 26HU5 (from above) compares with the super 36MC6 triode then:
1) 36MC6 super triode
ehh, very close to the better 26HU5 triode!!
and some more octal TV Sweep tubes:
2) 6CB5A triode
3) 6CB5A Schade
4) 6CB5A B. pentode
5) 26DQ5 triode (nice!)
6) 26DQ5 Schade
7) 26DQ5 B. pentode
8) the better 26HU5 triode again for easier comparing with the super 36MC6 triode
1) 36MC6 super triode
ehh, very close to the better 26HU5 triode!!
and some more octal TV Sweep tubes:
2) 6CB5A triode
3) 6CB5A Schade
4) 6CB5A B. pentode
5) 26DQ5 triode (nice!)
6) 26DQ5 Schade
7) 26DQ5 B. pentode
8) the better 26HU5 triode again for easier comparing with the super 36MC6 triode
Attachments
Last edited:
Some more. The 12BQ6GA is at 50 mA/div Vert., 50 V/div Horiz.
The 6HB6 and 29GK6 are at 10 mA/div Vert., 50 V/div Horiz.
Schade is using 100K plate Fdbk R with 20 uF in series, and 22K grid 1 drive R
1) 12BQ6GA in triode
2) 12BQ6GA in Schade
3) 12BQ6GA in pentode
4) 6HB6 in triode
5) 6HB6 in Schade
6) 6HB6 in B. pentode (note, the top curve should be spaced a little higher, I forgot to remove the 22K R, grid current near Vg1 = 0 ....)
7) 29GK6 in triode
8) 29GK6 in Schade
9) 29GK6 in pentode (actual. ie NON beam)
The 29GK6 is equivalent to 6BQ5 and EL84 (this tube was an actual pentode, non beam)
.....
The 6HB6 and 29GK6 are at 10 mA/div Vert., 50 V/div Horiz.
Schade is using 100K plate Fdbk R with 20 uF in series, and 22K grid 1 drive R
1) 12BQ6GA in triode
2) 12BQ6GA in Schade
3) 12BQ6GA in pentode
4) 6HB6 in triode
5) 6HB6 in Schade
6) 6HB6 in B. pentode (note, the top curve should be spaced a little higher, I forgot to remove the 22K R, grid current near Vg1 = 0 ....)
7) 29GK6 in triode
8) 29GK6 in Schade
9) 29GK6 in pentode (actual. ie NON beam)
The 29GK6 is equivalent to 6BQ5 and EL84 (this tube was an actual pentode, non beam)
.....
Attachments
Last edited:
Really nice curves...
No wonder RH84 gets praises... These curves explain a lot...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.