Rod Pentode TRF

[Murphyv310]

Hi.
Perhaps some will remember I made a two valve TRF using 1J42A rod pentodes. The radio worked amazingly but then I popped one of the filaments and when I replaced  it the performance could never be replicated. Its bugged me for a year or more.
So I decided to have another go. These low voltage 1J42A's are like hens teeth and I've never seen any more than the 10 I bought a good few years ago.
I have now three valves in circuit and although it works very well it is not as good as the original that was literally ear splitting.
I did substitute the output with a 1J18B and it's ear splitting, I'd think if all the valves were the higher voltage versions the radio would be really excellent. Reaction is very smooth with one setting covering half of the medium wave band. Sensitivity tails off though at lower frequencies.
I include the diagram. The audio stages share a common screen feed and decoupling, I've purposely not used potential dividers so I don't need to add an HT switch, valve voltages are low and initially I had 2v on the RF amp anode with good results, it's now at 4v with a slightly higher gain, lowering the anode load resistor although giving more HT its negated by damping the RF, a choke works but the reaction is harder to get working smoothly I may try a choke in series with a lower value resitor, a bit of trial and error.
Overall pleased with the result.

[Amie]

Well done Trevor I like the speed of construction with that method you've used, youve done a neat job too. Plus the old school Phenolic front panel love it!... I've done a few circuits with the Ruskies now, amazing for their size, I'll have to get some circuits drawn up of the little audio amps I've been playing with, particularly pleased with the push pull class A self split 3 valve using a pair of 1j29b as output.

Keep up the good work.

[Murphyv310]

Those 1J24P power power rod pentodes are even more amazing Amie.
I'm going to construct a low (ish) power AM transmitter for 40 meters, there are still some AM nets on Amateur radio. Perhaps in the next few weeks.
Thanks Amie and you keep up the good work too.

[Kalee20]

The audio stages share a common screen feed and decoupling

With partial decoupling, you could get yourself some extra AF gain as there would then be positive feedback between the valves!

I second Amie's comments, it's a good build method.

(Good to know about 1j42a rarity. I have a few: I'll take care of them.)

[ppppenguin]

That sort of ground plane construction is good for stability and consistency, especially at higher frequencies. Even in low frequency designs, the active devices don't know they're only meant to be working at LF. Probably more of a problem with transistors than valves. Even modest devices like the BC548 have some gain up to 300MHz.

[Amie]

...You can get caught out with RF and high currents causing signal gradients in the copper, I always find its worth bearing in mind that a ground plane is not always what it seems.

Another thing I found was one of my go to methods of anchors for connection nodes is punched chads of copper laminate board, stuck to the ground plane with CA glue, these can have significant capacitance and can affect HF circuits noticeably.

I can also vouch for the BC 5** series transistor range being able to take off at VHF.

Those 1J24P power power rod pentodes are even more amazing Amie

So they are to be sure! Ive seen your tv line output experiment utube video! However in my circuit as it was to be powered by batteries the 200mA heater current is a bit spendy, and the extra power not needed. Peters (Kalee20) current project has had its audio stage designed with two of the 1j24p valves giving 2 Watts.

I think the pair of you should self promote a little more if you've the time, there is an audience for these types of projects, if only to inspire others.

Steering back to topic,(almost), I have recently investigated a commercial 50's Superregen and have an intention to try to replicate the design for a pair of 1J29b to see how it compares

Amongst other projects of course

[ppppenguin]

Amie, indeed you can get caught out, even with a ground plane. In my professional design work I've occasionally had to segment a ground plane to prevent circulating currents and induced noise. However it's nearly always a lot better than simple point-to-point wiring.

A conventional metal chassis, as commonly used for valve equipment, is a pretty good ground plane. In this respect, the move to PCBs, usually single sided in older domestic equipment, was a backwards step.

[BusyBee]

Brilliant work Trevor. I like your circuit. I do like your construction too. Good for stability.

With a recent circuit of mine I used Manhattan style too but with a slightly different slant. I have built a 1j37b mixer driving a 455kHz ceramic filter via a transistor.  With this I had most of the 1j37b components on a copper board using surace mount components with all the tube base components connected into that. The copper board was connected by a short lead to the tube shield that I usually use. The only connection seperate from this was the anode connection which which was to a small piece of stripboard complete with anode tuning. The transistor / ceramic circuit was on stripboard, somewhat alongside but screened by the valve screen. I am working toward superhet ideas but not specifically pure rod pentode. I have tested this with just a 455kHz signal and through a 2 x 1j24b IF strip which I put together some time ago. The setup performed well with no sign of instability even though there was pretty high gain. It's early days though and somewhat experimental. The mixer circuit box is 3" x 2". I am hoping that the use of 'Manhattan Style' will much reduce the need for fully screened boxes but I have looked into solutions if required.

I do look on ebay from time to time for these 1J42A's but have yet to see any. It's perhaps a good thing though as I am pretty much fixed to 60V for a supply, generally varying this between 24V and 75V for testing. My preference is to avoid falling into the 12V and lower HT as, apart from some of these somewhat special valves, traditional types are getting to require somewhat specialist techniques to operate well whereas, at the voltages I use I can reasonably intermix transistors and valves well although battery power is not really sensible in the normal sense (without invertors etc).

   

[Kalee20]

Those 1J24P power power rod pentodes are even more amazing Amie.

Erm... would that be 1p24b? 

They are, to be sure... though they take a lot of drive voltage and a lot of bias: they're not very sensitive.

The devil in me says, build up an audio amplifier, immerse it in Fluorinert liquid (I've got a couple of litres!), screw up the HT to 200V and the Ia to 50mA each, and see how long they can blast-out 10W of high-quality audio. (After all, if they can push out 800W pulses of RF...   )

But I just can't bring myself to treat bottles as I can, encapsulated impure sand...

[Amie]

Those 1J24P power power rod pentodes are even more amazing Amie.

Erm... would that be 1p24b? 

Trevor knows I'm a bit dyslexic so he typed that so I'd understand