Strange Reaction

[kalee20]

It's got top be stray capacitances that causes the feedback.

Although the HT battery isn't a battery but a power supply, this may not make much difference. Whereas with a battery, the whole HT battery potential may up and down (at RF) with respect to chassis, with a power supply (with a fairly signifiocant capacitance to earth, the supply may stay steady and the chassis move down and up in potential. The end result is the same!

As for the test by MichaelR, does it oscillate with the DL96 removed? I'd bet it doesn't. Of course, as you point out, it'll disable the headphones. But you can detect for oscillation with another radio nearby, if you get heterodyne whistles.

[YT2095UK]

a thought occured when you said how Physically close the 2 valves are, you may be getting RF radiating from the anode plate itself, try cutting the end off a tin can, Earth it and then place it over the detector valve, see if it still happens then, I`v noticed on mine that I can alter the rxn setting just by putting my hand close to the #76 triode.
also, what results did the RF sniffer show?

[Variometer]

The glass envelopes are one inch apart. I placed an earthed screening can over the detector valve and the only difference was that I had to turn the reaction up slightly more to get to oscillation. Same thing when I placed it over the output valve. I haven't made sniffer yet, but I am going to. I don't have a uA meter, but I expect it will indicate on my little GW Instek oscilloscope.
Bob

[Variometer]

I have now discovered beyond dispute what has happened. I am at a loss to explain why it works this way and I would never normally have dreamed of doing what I did – it was a simple mistake!

I take my capacitors from a 224 Ceramic pack obtained from Maplin. For some obscure reason, the manufacturers do not print the value on the capacitors, but put a code number on instead.

I have just made the chassis for the neat build and I was double checking the components on the working prototype that has just been the subject of this discussion. The code number on the grid capacitor of V1 was 223 and I remembered that for a 100pf, it should be 101. I looked up the 223 code and found to my amazement that I had inadvertently put a 22nF (22,000pf) in instead! Maybe my mind had flipped back to pF days when 223pF would have been an acceptable value! Never did like the nF denomination and wonder why they ever introduced it!

I removed it and put a correct 100pF one in and although I received signals, varying the screen grid pot did not produce any reaction at all and the performance was only a shadow of what it had been.

I put the 22,000 pF back and it is again working perfectly!

I would never knowingly have used a value so high for a grid capacitor, but it does work extremely well.
I am not exceeding the valve’s specs either. I put the meter in the anode circuit of the DAF96 and the anode current is a mere .009mA. The total drain of both valves on anode current is less than 4mA, so I am not exceeding the DL96 specs either!

Bob

[MichaelR]

Bob,

Have you tried your scope on the circuit yet. This should clear everything up as Kalee suggested , I am assuming the bandwidth of the scope is wide enough.

Mike

[Variometer]

Mike,
There never was any problem with it working. As I said above, a mistake in the value of the grid capacitor was what made it work with the screen pot as a reaction control and I am delighted with the outcome. I am confident that I, or anyone else, can duplicate the cirucit with good results.
Why it works with a ridiculousy high value of grid capacitor is beyond my brain to fathom out. Maybe someone here will understand why.
Bob

[Variometer]

Further to my last couple of entries above, I have wired up the duplicate chassis using a fresh lot of components and it works just as well as the first one. My final task is to transfer it to the polished baseboard with brass terminals etc, join up the tuning capacitor, reaction potentiometer and matching transformer and it will be complete. I don't know why the 22,000pF grid capacitor makes it work, but I am certain in my own mind that it does, and that anyone can now duplicate the circuit with excellent results. I would be interested to hear any theories as to why it works, but I don't intend to do any more experimenting on it as I don't see how it can be improved and any further soldering or unsoldering would destroy its neatness.
When it is on its new base and up and running, I will display the photographs.

Thanks for all the suggestions and I will soon be making a "sniffer" for use on future projects.

Bob

[YT2095UK]

coolness

I will Try your extra large economy sized cap on my one valver also as I have plenty of messing about to do with it in the moro anyway (A different problem) and see if it does the same this end too, I have noticed that when I keep the earphones on, and unplug it from the main, it goes quiet after a few secs, and then comes back on again and gets quite loud before it fades entirely, this also has intrigued me.

also, if you want to make the sniffer really sensitive, use 2 germanium diodes (one instead of the 10K) wire them A to K, and then the ends to your meter, and use the A/K join as the sensor part, you can add a cap from the A/K join as well, stops any DC getting in if you accidentaly touch the meter terminals whilst "sniffing" (LOL, that sounds well!).
here`s a rough diag:

[PJL]

I think we proved earlier that the source of the feedback is via the 560 ohm, decoupling of the resistor stopped the feedback. The more interesting feature is why the capacitor improves the performance. The feedback is onto the anode of V1 and I am wondering if it's capacitive coupling to the grid - bit of a mystery.

You could prove it is via the anode by putting a RC filter in the anode HT path.

[MichaelR]

The larger coupling capacitor is driving the 2nd valve into oscillation which is feeding back into the first valve. the variable resistor is giving some degree of gain control on the first valve, both valves are performing as a 2 valve oscillator IMHO.

Mike

[Variometer]

It isn't the coupling capacitor that is large, it is the grid capacitor in V1, the one with the 1M grid leak resistor across it! Why it works, I don't know, but as I have now duplicated the circuit with the same results, I am happy with the outcome. Placing in on a polished base is a non-technical issue and really more to my liking. I will not be investigating further into it, but may try the large capacitor in an existing 1 valve triode receiver just to see what happens. I really don't involve myself too much with theory these days, as it is only a hobby, I am happier "tinkering" with the unknown where the outcome doesn't matter either way.
Bob

[Dave-N2DS]

Hi Bob,
I think you have a nice start to your radio. As others have mentioned, unwanted oscillations (instability) is quite often caused by not enough bypassing.

I would also add an electrolytic (10uF or so) bypass capacitor across your 560 ohm bias resistor. I can see where some feedback could occur from the anode current changes getting back in to the grid circuit.

One discovery that I found (not original by any means) is how your filament is connected to your LT supply. I would suggest making LT+ connect to earth, as well as the + terminal of each valve filament. The LT- would no longer be connected to earth but only the minus filament terminal on your valve holders.

Before you change, you can test by leaving your HT connected, and removing your LT battery. In the brief time before the filament cools, should your reception should be better, then you should change those connections. I found it make a huge difference in my one valve radio I made with a #30 valve.

Good luck!
Dave, N2DS makearadio.com

[YT2095UK]

I`v seen some valves that have a triode and a pentode in the same envelope, with your resistor sized inductors and some of the Mini variable caps and a few GP23s (12v batts about half the size of a AAA) I`ll bet you could make a Pocket radio with your experience!
might be another nice project for you later

[Variometer]

Yes, I can make small ones, but the effect is generally ruined by the fact that they need and aerial and headphones. Those double valves are usually 6.3 Volt or more heaters and generally require 0.3A for them as well. I do have a large amount of wire ended triodes that are 32mm long and 10mm diameter. They only require 1.25 Volts on the filaments and will work well with 27 Volts HT. Three of them takes up less space than a triode pentode.
I have made a 1 valve VHF set with a 955 acorn valve that didn't require an aerial at all!
Bob

[YT2095UK]

I just found something interesting whilst searching for info on the 6BW6 (I found a box of them), take a look here: https://www.vintage-radio.net/forum/...87&postcount=1
you can see that the regen is also a pot off the screen grid, And, with my problem getting an ECC83 to work with regen, I found that with a 0.1uF grid feed and pulled down with a 150k resistor I can get it to work, whereas with a 100pF cap it won`t work.
so I`m thinking a commonality May be something to do with the Gain of the device used?

[Variometer]

Pleased to hear that it made yours work as well. I looked at the circuit in the link, but found it difficult to understand as it is drawn back to front. I have spent so many years looking at diagrams where the valve line-up goes from left to right, it confuses my brain to see it the other way round.
I prepared the polished base for my set today, so it will soon be complete.
Bob

[Variometer]

Delighted to see that this topic got over 1,000 views. It is completed now and here are the final pictures. It functions best on 54 Volts HT and that gives an HT current of 3.8mA total. It does not require an earth, in fact an earth stops the reaction from working! When I first connected the completed receiver up, the HT current was just over 5mA with 54 Volts HT and I felt that was a bit high. I quickly noticed that I had inadvertantly connected the LT battery the wrong way round and it had upset the bias on V2. Connecting it the correct way lowered the HT current and also increased the volume.

As usual, I was feeling pretty fed up with it by the time it was completed to my satisfaction, so it will probably be several weeks before I build another one. Probably the next will be a three valver.
Bob