1J26A Rod Pentode

[Mike Watterson]

Maybe one wire is a screen if it has no filament centre tap.
Blue paint on outside can come off and then you'd see connections a little better. Still, the pair of g1 are innermost plates and pair of anodes the outermost plates.

I think I can almost see an anode plate to a wire!

[Murphyv310]

Hi.
Another couple of photos.

[BusyBee]

I have had a look through my collection of books but, apart from the info I gave above, I have not found any more. I did give it a quick search on a couple of Russian forums but had no joy. I do need to get organised for this though as I am finding that, as time moves forward, sites like Google in particular and getting very fixed in the direction you need to get their results from.

Tracy

[Murphyv310]

Very true Tracy
I've messaged my friend Adam in Poland to see if there might be any info available in Eastern Europe. If so I'll update in due course.

[Kalee20]

What about the supplier? They got the valves from somewhere, by following back they may get data! Sod's law says it will be in Russian, but it would be a start!

[Murphyv310]

Good point Peter.
The seller had posted the pin out for the 1j36b in the listing, it's still worth a try though.

[Murphyv310]

Hi.
The plot deepens with these valves. I used 9v to use as HT. The control grid was easy to find on the ohms range but two wires gave identical readings which initially I thought was odd. I then identified the G2 with 9v. Biasing on what I thought was the G1 gave me current flow, strapping the other wire to the G1 nearly doubled the current so I assume this is a dual control grid valve, I then found the Anode which didn't conduct too well at 9v, removing the G2 stopped anode current, the remaining wire I believe is a screen not a G3 as although a tiny change was noted on anode current its insignificant.

The filament is still visibly growing at 0.92v and 94ma current, raising the filament above 1v the efficiency lowers at the low HT volts of 9v.

I wish someone else had one these to confirm my tests.

[Murphyv310]

Update.
The seller on ebay replied to say he was unaware the connection diagram was the wrong one. He also said he spent a long time trying to get data for this valve. They are ultra rare apparently, probably an appreciating asset.

[Mike Watterson]

Yes. Seem to be very rare and quickly replaced by other models.
The filament current is interesting compared to these models:
   
(My own measurements)

It's about 1/2 of a 1p24b and about twice a 1j29b or a 1j37b. But the 1p24b and 1j29b are really dual rod pentodes, that's why they both have two filaments with a common at the top to give the illusion of a centre tap. The 1j24b is extremely efficient, but the the 1j29b is about twice the power rating of the 100mA DL94, but at about 1/2 the filament current. Thus the 1j29b has about 4 times the filament efficiency per mW as the DL94. The DL92 is miserable at about half the output power of the DL94 and thus the 1j29b is 8 times it's efficiency, if you used enough DL92's in parallel for the same power.

So I suspect the 1j26a is a bit like half of a 1j29b or 1p24b in construction using a similar filament to the 1p24b, which has two and thus three g1 plates:

Top View

| . | . | The control grid plates and filaments on all dual filament, thus "centre tapped" tubes, 1j29b, 1p22b, 1p24b

And single filament tubes:
| . |

Not shown: g2 and g3 being pairs of rods.

At least one rod tube has a single anode, thus is 1/2 a regular rod pentode.

I suspect the few applications where it was used the 1j26a was replaced by the also rare 1j36 and more commonly by the high power 1j29b or massive power 1p24b.

The 1j36b, quoted as the replacement is about 74mA filament current vs 130 mA.

[Murphyv310]

Hi.
I decided to go through this again but using 27v. They follow typical rod pentode characteristics with more output on the G2 with the anode grounded. Between 18v and 27v the output from the anode rose considerably. No doubt with higher HT the anode will perform much better. Triode connecting the valve even up to 27v reduces gain. Using my Fluke multimeter on Diode check give 1.2v between each control grid and f- strapping them gives 0.8v very low indeed for a valve! This also in my mind confirms the valve is dual control grid construction.
I've been using one of those plastic breadboarding experiment panels for tests, easier to work with rather than a birds nest.
I did this in the living room this evening so only a meter to hand. I used two piezo high impedance phones to make a simple AF amp to get some idea of whether it would work. There was reasonable gain between 18 & 27v and even results at 9v on the G2. The valve does feel slightly warm at 0.9v filament volts and the filament is just visibly heating, looks the same as the 1j42a running 1.1v but at a much higher current on the 1ZH26A.
To be fair I really don't see any great advantages using this valve and it may have a shorter lifespan than others, the greedy filament is also another downside.