1j42b Valves

[Murphyv310]

Hi.
I received my consignment of the Dual control 1j42b Russian rod pentodes today.
They are thinner and shorter than the other ones I have. I had a play about with one of them and feeding the filament with 1v I get a forward voltage (Diode test) between filament and either of the control grids of 1.9v, using both I get 1.1v in fact strapping all electrodes together I get 0.9v between them and the + of the filaments.
Next test was to try this set up as a detector in the frame aerial crystal set. Unbelievably this actually works between the joined electrodes and filament - (filament - to phones) performance on strong signals is slightly lower than a geranium diode but weak stations come through much easier, no doubt the valve is conducting where the germanium doesn't....... quite amazing.

Next project is now on the cards

[Mike Watterson]

Performance actually gradually deteriorates above 10V HT!

See
http://www.radiomuseum.org/tubes/tube_1j42a.html
This is seriously the oddest of the Rod Pentodes, which are pretty odd.

Datasheet English PDF here
http://www.radiomuseum.org/forum/142_dat...glish.html

If you ground g2, g3 and anode, then the two g1 grids will work in gammatron mode, with filament as cathode and one grid as grid and the other as anode (either way round).

It may be possible with a 19kHz common mode choke for filament to feed stereo multiplex to filamentary cathode and have it work both as a frequency doubler 38kHz and mix in the 38KHz L-R to g1a and g1b, and then have a single valve stereo multiplex decoder. Grundig had a one valve decoder for early stereo on exports as Majestic to USA, though the mixer part was done with semiconductor diodes as switches.

I'm transferring workshop to a different shed right now.

[Pat Pending]

Trevor, have you tried adjusting the filament voltage while using them as a detector? You can often find a sweet spot where they really perform well, when running at reduced (critical) filament voltage. I've found examples of the 6418 hearing aid valve that are good detectors when used as you describe.

Cheers, Andy.

[Murphyv310]

Thanks Mike for the links
Thanks too to Andy.

A few years ago I made a one valve radio with Ferrite rod aerial and a Rod Pentode, it featured reaction but I found that if you tailored the HT volts to around 30v you could dispense with the reaction control (either cap or variable resistor) as the valve would self limit, like an AGC feature, it would always sit at the point of oscillation, it would throttle back on strong signals but the gain would rise on weak stations. Higher HT and the circuit would take off, lower HT and the reaction would stop.
Regarding using a valve detector in a crystal set circuit, I've never had a result like today, I did get some success but only on a really strong signal, I find that I can hear stations not audible with a germanium diode, OK it's not a proper crystal set as I'm using power for the filament but I'm amazed.
I did also try varying the filament volts, I got results at 0.3V full volume at 0.7v with no change to the max of 1.2v

[Mike Watterson]

Most Rod Pentodes were designed to "still work" at 0.7V.
In contrast a DK91 / DK92 / DK96 / 1L6 is unreliable below 0.9V. The DK96 HF performance is poorer below 1.1V

At 0.7V it must be only about 70mW of power consumption? I suppose over a 1000 hours from a D Cell. Why bother with an on/off switch?

If you examine the g2 current vs Anode current vs g1 voltage for Rod Pentodes, you can see why the usual technique of a simple resistor for screen grid fails badly, as Ia rises, the Ig2 falls to almost zero, so the Vg2 would rise to full HT. Hence the technique is either voltage tap, full HT (if it's a low HT) or a pair of resistors as voltage divider. g2 and Anode is almost a current mirror.

[Murphyv310]

Hi.
Indeed Michael you are right about these being the oddest of the Rod Pentode valves, In fact they are the oddest valve I have ever come across.
I have thrown the valve theory books in the bin and looked at these valves from a different perspective.
I spent about an hour and a half with one I'll use for experimentation and without regard for valve theory.
On the bench I had an AA Ni-Mh cell, PP3, a 22ohm resistor a few small value caps and assorted resistors 100k and above, high impedance phones and other bits and bobs.
I only used the 1j42b as I'm now quite familiar with the other types and their idiosyncrasies.
Initially I wanted to try it as a simple audio amp, after trying some combinations I found that using the G2 as a control grid with about 1.2v positive bias and the anode as the output I got around 4x gain, very odd indeed. I then tried using the valve as a leaky grid detector, results were pretty poor no matter what combination was tried. Next was a simple crystal osc, simple feedback from anode to G1 this works well but the bias is touchy and how it's connected, it must be between Filament - and needs to be 1-1.2 v to work reliably, a 44mhz crystal started and ran as good as a 200khz one I have, anode load was a choke for the 44mhz and a 100k for the 200khz crystal, connections are conventional but the G2 produces nearly the same op as the anode, a very high value of resistor (1meg up to 8.2meg) increases the output, G3 damps the op if connected to either polarity of the filaments on the 200khz crystal and stops oscillation on the 44mhz crystal. As an RF amp I was getting useful gain at 100mhz but using the G2 as a control grid got more gain than using the dual G1's how very odd.
I'll be doing more tests over the next few days but have to say if anyone is experimenting expect the unexpected!!!!!

[Mike Watterson]

It does really comply with valve theory, but of the advanced kind!

All the rod tubes are really electron guns with two beams in opposite directions from filament and g1, g2, g3 are really electrostatic lenses. The 1j29b and 1p24b are actually two sets in parallel, hence two filaments and a an extra shared plate/rod for g1, g2 and g3 between the two parallel beams!

Americans tried making rod electrode tubes in 1930s but it failed because the cathode was indirect. The Russian guru that invented these realised that unlike a CRT the idea would ONLY work with a very very thin perfect direct filament cathode with the g1 plates very very close on either side to form the pair of beams.

All the rod tubes have dual g1 plates (internally connected) and four or six g2 and g3 rods or plates internally connected. With the 1j37b they removed some shielding rods so that access to g1a and g1b are separate. The 1j42a is last model.

Most gain:
6V HT! (Equivalent to tetrode kink area of beam pentode)
g2 to direct to HT, g3 to f- or f+ (forget which)
tie g1a and g1b together. It's the opposite effect to an op-amp.
If you have differential RF in with a centre tap on transformer to g1a and g1b, then volts on centre tap adjusts threshold / gain and the RF is rectified (or frequency doubled) depending on if AF load (resistor or audio transformer and capacitor to OV) (or if RF tuned / RFC).

Have fun.
I'd only use one of these as mixer/osc ,RF AGC, frequency doubler, unusual detector etc, maybe IF. The 1j24b will work with 2 x 12V alkaline A23 or similar and takes similar heater current.
It will be interesting to try it at 88 to 108MHz and 110MHz to 137MHz as a mixer/osc on a superhet or super-regenerative. They could be used also as IF dual 470KHz & 10.7MHz with AGC but not a regular IF amp. I have Russian schematic with older 1j37b as AGC stage. I'll look for it.

All the rod tubes will work as logic gates (Resistor Nor Gates) or amps using positive g2 with no need for capacitive coupling or level shifting, with suitable HT!
Wrecked moving junk out of new workshop.

[pwdrive]

What would have those types of valves been used in originally?

Lawrence.

[Mike Watterson]

Note!
There is almost no electrostatic shielding (none on 1j37a), just rods.
Because it's a pair of planar electron beams they are also magnetically sensitive!

All photos of Russian parts have copper or tinned sleeves on the tubes (for electrostatic shields) and they always lie flat on the PCB (older models) or alumina substrate (surrounded by SMD parts!). The cabinet kept them away from magnets.
   
   

  czech_rod_article.doc (Size: 298.5 KB / Downloads: 6)

[Murphyv310]

Hi Lawrence.
I don't purport to be an expert on these valves like Michael but I believe they were used in Sputnik. MIG fighters and a civilian Walkie Talkie, also as proximity fuses too. There may be other uses too.
The Russians had and do have some very clever engineers, even in the consumer world produced fantastic cameras with amazing lenses, their watches too were precision, the old Poljot watches were a work of art in engineering terms.
I believe the Russians most likely could have done even more re-search and made further strides with valves whereas the west concentrated on silicon, all very interesting.

Michael thanks for posting the "Thick film" picture