On that calculation of input impedance for the 13FM7 triode #1 (above) with positive grids, I probably have to -add- (rather then subtract) the two tube absolute grid current deltas, since the drive polarities are opposite. So that would give
Rin = 4X (1.1mA + 0.5 mA )/ (1.9V) the 4X for G-G instead of CT to G.
So Rin = 3368 Ohms. But we can likely reduce the current swing to 1/4 of that (by using higher plate load resistors or even CCS's), so 4X 3368 = 13473 Ohms. A reasonable input Z for some things at least.
Hmmm, I guess CCS loads would even raise the input Z drastically, since no current change is even needed then. Just Mu operating at constant current, so linear even.
That's an interesting observation. Positive grid V gain (only) stages can still have high input Z, just power stages are stuck with low input Z. Right? That's a new one for my book.
So one could just stack these 13FM7 tubes up in P-P parallel for whatever power level is needed from the big output #2 triodes. No follower needed for the input then. Can use an Edcor XSM15K/15K to do the splitting at the input. And maybe a GXPP OT. Or a CXPP OT for total Hi-FI.
This is getting ridiculously easy now to build a high quality (linear) reasonable power LV B+ amplifier.
The more 13FM7s in parallel, the lower the output Z too. This needs to be proven!!
Most any of the dissimilar dual triode Vertical TV amp tubes should work I would think. 13GF7 say.
This would set a new "paradigm" for entry level, "safe to build", tube amplifiers. And dirt cheap too.
They said it couldn't be done...
Done....
Rin = 4X (1.1mA + 0.5 mA )/ (1.9V) the 4X for G-G instead of CT to G.
So Rin = 3368 Ohms. But we can likely reduce the current swing to 1/4 of that (by using higher plate load resistors or even CCS's), so 4X 3368 = 13473 Ohms. A reasonable input Z for some things at least.
Hmmm, I guess CCS loads would even raise the input Z drastically, since no current change is even needed then. Just Mu operating at constant current, so linear even.
That's an interesting observation. Positive grid V gain (only) stages can still have high input Z, just power stages are stuck with low input Z. Right? That's a new one for my book.
So one could just stack these 13FM7 tubes up in P-P parallel for whatever power level is needed from the big output #2 triodes. No follower needed for the input then. Can use an Edcor XSM15K/15K to do the splitting at the input. And maybe a GXPP OT. Or a CXPP OT for total Hi-FI.
This is getting ridiculously easy now to build a high quality (linear) reasonable power LV B+ amplifier.
The more 13FM7s in parallel, the lower the output Z too. This needs to be proven!!
Most any of the dissimilar dual triode Vertical TV amp tubes should work I would think. 13GF7 say.
This would set a new "paradigm" for entry level, "safe to build", tube amplifiers. And dirt cheap too.
They said it couldn't be done...
Done....
Last edited:
RE: jhstewart9
Interesting read about the space charge tube by Norman Pickering. Thanks.
The TV Sweep tubes obviously have grid 1 too close to the cathode for space charge use. (too high grid 1 current) It is surprising that the 26A7 tube seems to do so much better than the commercial space charge tubes mentioned so far. (rather low grid 1 current as space charge electrode for 26A7) Something unusual about 26A7 construction obviously.
I hate to bust a new one apart, but I would sure like to see how they are made. One thing I have noticed however is that the grid3 (beam forming plate actually) has a much wider opening than usual. I think this explains the bad kink in the 26A7 pentode curves! Grid 3 is not doing its job of repelling secondaries sufficiently, leaving tetrode like kinks. The kinks are not cleared until 20V is reached, ruining the 26A7 LV advantages in pentode mode.
If the grid 3 opening had been smaller, or the grid 3 brought out on a pin, this could have been fixed. But that might have impaired the low Rp some. The very close spacing of the plate box (only 48V B+ rated) may explain the low grid 1 current in space charge mode. Due to Rp only 2400 Ohms. Sure wish grid 3 were available on a separate pin. (put some -B on it)
The article mentions higher power space charge tubes that were to be introduced soon (late 1947). Did these ever make it on the market? tube #s?
The article mentions "hangover" at low frequencies is improved with the lower Rp of the space charge tube. I guess this is something related to damping?
The 26A7 is providing around Mu 17.5 in space charge mode (versus Mu 2.9 in triode), which is close to the Mu 20 mentioned in the article for their space charge tube. But 26A7 produces less useful current in the LV constrained space charge mode 4.5 mA peak versus 15 mA peak in normal triode mode (LV constrained). Each variant has its advantages, hi Mu gain stage, or low Mu power stage.
--------------------------
So 26A7 was looking reasonably OK, until the lowly $1 13FM7 came along. It does need CCS plate loads for the (Mu 66) gain triode #1 in the 13FM7 however, for running LV B+ with positive grid 1.
Well, if anyone wants to experiment with space charge tubes, the 26A7 is the best so far. (even though not designed specifically for that) Most linear, best power, best -plate- efficiency. (not best heater though)
.
Interesting read about the space charge tube by Norman Pickering. Thanks.
The TV Sweep tubes obviously have grid 1 too close to the cathode for space charge use. (too high grid 1 current) It is surprising that the 26A7 tube seems to do so much better than the commercial space charge tubes mentioned so far. (rather low grid 1 current as space charge electrode for 26A7) Something unusual about 26A7 construction obviously.
I hate to bust a new one apart, but I would sure like to see how they are made. One thing I have noticed however is that the grid3 (beam forming plate actually) has a much wider opening than usual. I think this explains the bad kink in the 26A7 pentode curves! Grid 3 is not doing its job of repelling secondaries sufficiently, leaving tetrode like kinks. The kinks are not cleared until 20V is reached, ruining the 26A7 LV advantages in pentode mode.
If the grid 3 opening had been smaller, or the grid 3 brought out on a pin, this could have been fixed. But that might have impaired the low Rp some. The very close spacing of the plate box (only 48V B+ rated) may explain the low grid 1 current in space charge mode. Due to Rp only 2400 Ohms. Sure wish grid 3 were available on a separate pin. (put some -B on it)
The article mentions higher power space charge tubes that were to be introduced soon (late 1947). Did these ever make it on the market? tube #s?
The article mentions "hangover" at low frequencies is improved with the lower Rp of the space charge tube. I guess this is something related to damping?
The 26A7 is providing around Mu 17.5 in space charge mode (versus Mu 2.9 in triode), which is close to the Mu 20 mentioned in the article for their space charge tube. But 26A7 produces less useful current in the LV constrained space charge mode 4.5 mA peak versus 15 mA peak in normal triode mode (LV constrained). Each variant has its advantages, hi Mu gain stage, or low Mu power stage.
--------------------------
So 26A7 was looking reasonably OK, until the lowly $1 13FM7 came along. It does need CCS plate loads for the (Mu 66) gain triode #1 in the 13FM7 however, for running LV B+ with positive grid 1.
Well, if anyone wants to experiment with space charge tubes, the 26A7 is the best so far. (even though not designed specifically for that) Most linear, best power, best -plate- efficiency. (not best heater though)
.
Last edited:
So summarizing the 26A7 tube
26A7 linear operating triode modes
(pentode mode sucks...)
10V/div Horiz.
1) 26A7 in normal -grid 1 triode mode, 10 mA/div Vert., -2.5V g1 steps, Mu 2.9
2) 26A7 in space charge mode, 2 mA/div Vert., Vg1 = +0.33V, -1V g2 steps, Mu 17.5
----------------------------
Uh-Oh, I just realized that 7.25V/(.25V) is not Mu 2.9, it would be Mu 29
I'll have to re-check this on the curve tracer or datasheet.
http://tubedata.milbert.com/sheets/049/2/26A7GT.pdf
Oh, OK. I must have written the g1 step size down wrong when I named the normal triode file.
The datasheet confirms normal Mu -is- near 3. Never-mind. We return your TV set to normal programming...
26A7 linear operating triode modes
(pentode mode sucks...)
10V/div Horiz.
1) 26A7 in normal -grid 1 triode mode, 10 mA/div Vert., -2.5V g1 steps, Mu 2.9
2) 26A7 in space charge mode, 2 mA/div Vert., Vg1 = +0.33V, -1V g2 steps, Mu 17.5
----------------------------
Uh-Oh, I just realized that 7.25V/(.25V) is not Mu 2.9, it would be Mu 29
I'll have to re-check this on the curve tracer or datasheet.
http://tubedata.milbert.com/sheets/049/2/26A7GT.pdf
Oh, OK. I must have written the g1 step size down wrong when I named the normal triode file.
The datasheet confirms normal Mu -is- near 3. Never-mind. We return your TV set to normal programming...
Attachments
Last edited:
26A7 as well as 28D7 are kinked badly on purpose, 28D7 mostly for stupidly low-V knee, whereas 26A7 actually benefits from secondary emission by having very low THD in SE and P-P, especially for a low-V tube. Here's an article on 26A7 and how it benefits from secondary emission, made it a while back.
Vacuum tube / Valve appreciation Society
However, it is worth noting, with a Cavitrap / Slatted anode like 48 and more modern tubes, it would be possible to achieve far better results with this kind of tubes, although 26A7 is pretty much perfect for what it is, except the massive heater consumption.
Vacuum tube / Valve appreciation Society
However, it is worth noting, with a Cavitrap / Slatted anode like 48 and more modern tubes, it would be possible to achieve far better results with this kind of tubes, although 26A7 is pretty much perfect for what it is, except the massive heater consumption.
They are meant for aviation use where the only supply is 25-28V (12 cell storage, charges at 28V or so). When you say microphonic, how was this measured, because considering their intended use, I'm sure that would be the first issue they would address, either way may very well be just a faulty valve, I dropped an N339 Beam Tetrode yesterday and can now hear what sounds like one of the grids being loose when I carefully lightly shake the tube, so that one or both tubes may have been damage. The movement may be of the grid supports, which is what suspect has happened to the N339.
Measured using my ears. A slight tap on the tube was really loud. Fidelity was probably at the bottom of the requirements list for these tubes.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.