Joined the club, Rod pentode superhet radio.

[Murphyv310]

Hi Adrian.
Not sure if the 1ZH37b will work to 108mhz, they do work very well just below 60mhz though. I have a gut feeling they will work even higher than 108mhz, inter electrode capacitance is very low.

[BusyBee]

Can I ask if the 1J37b valves would work around the 88 to 108 MHz band or higher, I have ordered a few and would like to use them as rf input and mixer if possible down to 10.7 MHz IF.  I am guessing they would be better than just a normal 1J18b or 1J24b for this?

It could be a while before they arrive so at this stage just planning.

Adrian

I have built a mixer but have yet to test it properly. I have only put 10.7 MHz through it to verify that it was basically working but maybe I can feed it with the generator and a signal and see. I just copied my 455kHz module but with appropriate changes. I'll check it today if I get chance.

Tracy

[AdrianPH]

Thank  you very much Tracy, as I said it will be a while before they arrive so no rush, but it is appreciated.

Adrian

[BusyBee]

Hi Adrian,

I haven't advanced far with the mixer, mainly due to my lack of test equipment for VHF frequencies. My signal generator has a dead band between about 38 and 90 MHz and my scope is almost out of steam at 100 MHz. Your talk of VHF did prompt me to think again. I made the mixer originally as I had made a 455kHz mixer, which does work well, but got sidetracked with FM and 10.7 MHz due to coming across a 6BN6 which seemed to be an ideal for simple FM demodulation (another story). I progressed a little with testing a dual band IF stage but really need to build a local oscillator for FM.I was thinking the other day and did come up with what I thought a good physical construction for VHF. If you are interested - A sort of bobbin layout with the valve screening tube in the centre soldered to manhatten layout surface mount washer looking discs at either end for the components. The valve connections coming out of the holes next to the board for short and easy connection to the rest of the circuitry. I have yet to determine the oscillator details though. Until I get this made though I think I will be fighting against the elements to prove anything but will keep going and keep you updated with progress.

Tracy

[AdrianPH]

Hello Tracy, I am thinking of following the FM stage of this http://www.hi-ho.ne.jp/ux-45/schmatic/Fm.GIF, but where I can use rod valves as I have some 37's on the way using one of them for the first stage/mixer. The layout of the circuit I have not got down to yet but I think will need a different idea to what I did or the am superhet.

Adrian

[Mike Watterson]

You know the Vidor 1957 FM Radio uses a DF97 triodised as an oscillator-self mixer on VHF as do all battery valve sets except early ones using a DC90.
The interesting thing on the Vidor is that on AM, that's a LO and the first FM IF (a DF97) is then the AM mixer, signal to g1 and LO to g2.

I'd not attempt a separate LO & Mixer at VHF. Almost ALL mains valve sets use the same idea of osc-mixer using a triode, or a triodised Pentode. Either EF80 or 1/2 ECC85 or part of ECF82. On a mains valve set the 1st EF80 or other RF stage is more to buffer the aerial so the roof top H or Yagi isn't radiating the LO. The Battery valve sets just have a whip, loop or two whips and a carefully tweaked balun between aerial and the DF97. Most transistor sets except for posh ones with HF, use s single transistor osc-mixer on VHF and a separate one for AM.

I'd only use the 1j37b for AGC at IF, or mixer at IF for BFO or for 10.7 MHz to 455 MHz on a dual conversion AM IF. I'd use the SAME triodised self oscillating mixer as every VHF radio. The only difference using the 1j24b or the 1j29b (not sure which would be better, but the ONLY two to consider) compared with an EF80 or the DF97 is the connection of g3, it goes to f- or f+ or 0V. The g2 connects to anode. I've tested swapping a DC90 and DF97 as VHF mixer/osc and only one connection needed cut on the Vidor valve base, which had no effect to the DF97 fitted.

I've tested the Vidor 1957 idea of a Pentode mixer at AM. Which was used on early 1930s superhets. It works driving the g2 of mixer from the anode of triodised osc direct. So 2x 1j24b can plug-in replace any battery valve heptode or Octode mixer osc (on those the g2 is really a sort of triode osc anode, g1 is feedback and some higher g number is RF in). The Pentode mixer on AM (like a Hexode with separate triode osc) has less LO leakage on the aerial than most heptodes or octodes.

I'm not saying you can't use a 1j37b as either FM mixer or 1st AM mixer, but it's going to be poorer (if it works at all on VHF) than the VHF mixer-osc triode or triodised pentode. Similarly on AM, the Pentode with RF to g1 and LO to G2 is how it's done. The 1j37b will be poorer and takes more filament.

The wunderlicht (sp) is an old 1930s dual grid triode that was used mainly as a detector, and sometimes as mixer. The 1j37b isn't quite the same, but it's better as IF DC controlled gain, a detector, frequency doubler or an IF frequency changer because the g1a and g1b need driven by a centre tapped transformer. Signal to centre tap and the drive is LO. For DC gain you feed RF or IF equally to g1a and g1b and differentially change the DC bias (you can fix either grid to 0V or f- and then you can use positive OR negative AGC volts. Max gain is DC of g1a and g1b are equal.

The 1j37b (and its cousins) don't at all work like a dual gain FET or the RF and LO grids on Hexode, Heptode, Octode and Nonode. Common mode signal on g1a and g1b is amplified. ANY differential signal (AC or DC), in any phase or polarity reduces the gain. So a rotten starting point to design a VHF front end. Use self osc mixer in triodised mode. It ought to be a little better than triodised EF 80 which barely reached 101 MHz. That's why the 108 MHz sets usually used ECC85 or occasionally ECF82.

The 1j37b was an interesting novelty looking for an application. It's only used as IF gain control or 2nd stage mixers or mixers for BFO on the Russian sets.

[AdrianPH]

Hello Mike, it will take some time for me to take in what you have printed here.  I will have to bring up the circuits of the sets you mention (If I have them) to try and follow what you are saying.  I had been reading through some of the posts by Joe Sousa on the Radiomuseum site and it was the statement of: - 'Model r326, by Aleksandrov Radio Works in Russia, uses the 1ZH37B dual grid pentode in mixer service.'  That made me want to try it and I have assumed that the rf input onto one grid and the local osc on the other grid would give isolation between the antenna and local osc.  If I am reading the thread correctly from Joe although the valves seem to be rated to 60 MHz, his tests suggest they will work considerably higher frequency.  I was going for a separate local osc valve as I thought it easier to construct and get right, it even allows me to use one of my signal generators as a local osc source initially while getting the mixer stage working.

It is all  learning exercise for me, just to see if I can using these valves, because I will have them?  I do appreciate the thread I mentioned gives a good account of them as an agc stage.

This is what I had in mind, sorry for the quickly hand drawn sketch.
   

OK I can start to appreciate it may not work as expected, but could you please clarify something for me
'The 1j37b isn't quite the same, but it's better as IF DC controlled gain, a detector, frequency doubler or an IF frequency changer because the g1a and g1b need driven by a centre tapped transformer. Signal to centre tap and the drive is LO.'

I am struggling a bit to understand this and wonder if this is the type of thing you mean: -
     

If not could you do a simple hand sketch some time.

I am considering a second mixer stage down to 455 at some point going from 10.7 to 455 for NBFM, in effect having two demodulators.  I think I can see that it is easier at 10.7 to 455 because the centre tapped transformer could be tuned for a single frequency and not across 88 to 108 or what ever band I go for.

Head hurts.
Cheers

Adrian

[BusyBee]

Hi Adrian,

I did a bit of background work, looking for links to the R326 but it seems like you have found them anyway. Here is another which does give a bit more on the actual radio, and some possible limitations of rod tubes,  and is an interesting site anyway: ANTENTOP.

Mike has hit upon a few excellent points here which will be relevant to you. It's not something that I am very familiar with as I tend to concentrate more on design / construction and research more in the areas of ham or, in this case, military radio rather than consumer equipment. As such things like cost are a lesser factor so combined mixer oscillators are thinner on the ground. I do wonder? My view is that the Russians didn't seem to make any major errors so their approach should work. Being military it is designed for different criteria. The way I see it reliability and ease / standardisation of maintenance may be far more important than absolute performance of cost. (I do find it interesting how boxes of rod and other tubes came with spec / connection sheets). A point I would make though is that rod pentodes were a clever design to circumvent some of the limitations of traditional valves. As such they will almost certainly have limitations in comparison, as well as their intended advantages.

If / when I find more I will update you.

Tracy

[Mike Watterson]

I had been reading through some of the posts by Joe Sousa on the Radiomuseum site and it was the statement of: - 'Model r326, by Aleksandrov Radio Works in Russia, uses the 1ZH37B dual grid pentode in mixer service.' '[/i]

I am struggling a bit to understand this and wonder if this is the type of thing you mean: -
 

If not could you do a simple hand sketch some time.

I am considering a second mixer stage down to 455 at some point going from 10.7 to 455 for NBFM, in effect having two demodulators.  I think I can see that it is easier at 10.7 to 455 because the centre tapped transformer could be tuned for a single frequency and not across 88 to 108 or what ever band I go for.

The R326 is using it at HF, not VHF *I think*, and I'm not sure that's it's the front end mixer. I'd have to check, in any case that's a pretty crazy set. Certainly most Russian VHF applications were below 70 MHz.

Certainly you can use a 1j37b as 10.7 MHz to 455 kHz for NBFM or HF dual conversion. However I'm not sure that the 1j24b doesn't work better (feed LO to g2 and RF or IF to g1) and certainly dramatically lower filament current.
The 1j37b doesn't need as much LO drive, but needs more complicated centre tapped transformer. It's certainly likely to work well as a detector or frequency doubler if driven with a differential signal on g1a and g1b, with DC bias to centre tap (often 0V via decoupled leak resistor of 2.2 M Ohms.)

Joe is the expert.

[Mike Watterson]

A point I would make though is that rod pentodes were a clever design to circumvent some of the limitations of traditional valves. As such they will almost certainly have limitations in comparison, as well as their intended advantages.

I think it's because they didn't have Silicon till later and the RF performance of their Germanium transistors was rubbish at the time of the early 1j17b.
Another factor is that for MiGs and Satellite the Rod Pentodes are more robust to EMP from weapons or from Cosmic Rays. The Rod Pentodes were not better than mains valves the Russians had, but better than B7G miniature 1.4V valves (they did 1.2V variants of those for NiCd) and much better than the USA & UK wire-ended sub-miniature types. The later Rod Pentodes are also staggeringly economic giving better performance than 50mA 1.4V B7G types but at as low as 11mA and 1.2V NiCd supply.
They did use AF power Germanium for simple HT inverters for most of the Rod Pentode sets. Compare Ge transistors as inverters and audio out on 1950s valve car radios in the west.