Strange Problem

[Variometer]

Feeling the need to get back to radio again, I began my short-wave 2 design. Starting with the first valve, I assembled it & it didn't work. I found that very strange because the valve (DL96) was OK, I checked it in another receiver. Got a loud buzz (as expected) when I touched the grid pin. Felt that my home-made coil & capacitor may be wrong, so replaced them with two 100uH chokes in series & 500pf tuning capacitor. Still didn't work. Reversed reaction choke, still nothing. Removed reaction circuit altogether - still nothing. Needless to say, I had checked the grid leak capacitor & resistor earlier.
Then I removed the grid capacitor & leak & substituted grid capacitor with a germanium diode. The set then worked OK as a crystal set followed by one valve amplifier.
So - aerial & earth systems, tuning circuit, valve & HT/LT supplies proven, but still won't work in grid leak configuation! I can't think of any logical explanation for this. I know a misplaced lead in a reaction circuit can stop a set working, but it should be OK (but not very loud) with reaction circuit removed. Spent a whole day on it with no results & am currently fed up with it & have temporarily put it aside. I will return to it shortly after giving some more thought to it. Have never come across this before.
Bob

[Variometer]

As I was starting this thread, my eye fell upon a miniature Hivac wire-ended triode that has been lying on my computer desk for some weeks. I went & removed the DL96 and soldered the Hivac triode in its place. Lo & behold, it worked! I next connected up the reaction circuit & was pleased to find that I then had a smooth reaction with greatly increased volume. Without altering the tuning, I unplugged HT & unsoldered the Hivac and replaced the DL96. I could then hear the station very faintly. So, it looks like it is something to do with the valve (that I know is OK, because it works in another slightly diferent one valve circuit). I would imagine it is something to do with the screen grid voltage. I have the screen grid pin connected to the anode pin at the moment!
Anyway, I am confident that I will soon sort it out now, but will again put it aside for a few days.
Bob

[AVO_VCM163]

Interesting!
The DL96 is not exactly well-known for good SW performance!!!! The total capacitance, particularly triode connected may be a little heavy.
Perhaps a DF96 may be more lively?
Regards, R

[Variometer]

Not only did it not work on SW, it didn't work on MW either when I tried with with the proven tuning circuit of 200uH & 500pF capacitor. I think I will try a different voltage on the screen & see what happens. Once I get it going OK on MW, I can change it to SW & then add the 2nd valve.
Bob

[Aerodyne]

The DL is an audio output valve, surely. Better results would be obtained at RF by the use of a more suitable valve (DF... series, perhaps). Use the DL as an audio output stage to headphones.
-Tony

[Variometer]

Trouble is, in the past I have built numerous small sets using output valves & they have mostly worked fine. I have just tried changing the voltage on the screen grid, but no effect. It may just be the compact layout that is stopping it working. Still, it is this sort of thing that make it interesting, if they worked every time, I would soon lose interest. I will keep you posted on progress.
Bob

[Skywave]

You say "So, it looks like it is something to do with the valve (that I know is OK, because it works in another slightly different one-valve circuit)".

[My italics for emphasis].

So perhaps a good starting point would be to compare the two circuits - there must be a difference between them somewhere to give you a clue.

Following on from that, and noting that you have an AF output valve configured as a triode in RF service, I wonder if what you are experiencing here is Miller Effect, and that either the input signal is being shunted by the consequent low input impedance or if the stage has become unstable, depending of the phase of the signal as the signal freq. changes.

Al / Skywave.

[Peter.N.]

If I remember correctly, a DL96 has two filaments, 1.4v in parallel for battery operation and 2.8v for series operation. I suppose you didn't have 1.4v across it in the series mode?

Peter.N.

[Variometer]

Al
I think it is most likely that it has become unstable, possibly as you say, due to Miller Effect.
The circuit is identical to the one that the valve works OK in. The only difference is their physical layout, but they are about the same size (2.5" x 2")

Peter
The two filaments are connected OK. The end pins joined together & going to LT+ & the centre tap down to earth, with 1.5V supply.

Bob

[Skywave]

Instability: obviously, if you have a 'scope to hand, you'd have used it by now to investigate this possibility. So, here's an idea.

Assuming that the grid bias is provided by a resistor (1 Meg., 470K ?) from control grid to chassis direct - as opposed to this resistor in parallel with the grid condenser - lift the earthy end of the grid leak from chassis and connect a 1 k-ohm resistor in series with it to ground. Wire a 0.01uF cap. across this new R as an RF by-pass. Now you can monitor the grid current with a volt-meter across this new R and thus calculate the bias voltage on the control grid. Under the conditions of zero reaction applied & no aerial connected, I would expect the bias voltage to be very small (contact potential) as the signal tuning cap. is swung through its range. If the monitored voltage is / becomes large, this can only come about by grid rectification of a large signal at the grid - which, in turn, can only be produced by the stage oscillating (unwantedly).

Alternatively - and I appreciate that this might not be convenient - exchanging the LC components in the grid circuit between the two circuits might narrow the search area.

Al / Skywave.

[Variometer]

Al,
Thanks. I do have a 'scope, but it never crossed my mind to use it in this case.
Bob

[Variometer]

Further to the above, I have solved the problem without even getting the 'scope out.
The problem was with the screen grid connection. On the DL96, pin 2 is the anode & pin 3 is the screen grid. I had simply linked these two pins & taken the lead to the bottom of the anode choke. Last night, when I put the Hivac triode back in, I separated these two pins, but left a "flying lead" on the screen grid so I could experiment with different voltages above the board when I put the DL96 back. Anyway, different voltages made no difference, but when I touched the flying lead from the screen grid, to the bottom end of the anode choke (that is connected to the anode pin), the whole thing worked very well, but if I link the pins on the valveholder (amounting to the same thing electrically), it doesn't work!
I have come across similar things before in reaction receivers where the physical position of a single lead can stop the whole thing working, but it has usually been in the reaction circuit.

I can now move on again. Thanks for all suggestions.

Bob

[boiss]

Just a note to the above;-
Back in the sixties the were many circuits around for one and two vave superregens and they all appeared to use audio O/P bottles. One even used 2 6V6s which I had built and the RX section was definitely better on MW with a 6V6 instead of an EF39 which I had tried.
As for audio bottles at higher frequencies I remember a Radio control transmitter that used 2x DL92 as a PP oscillator /output circuit.
cheers,
Mike

[Variometer]

I also remember short-wave sets being built using HL2s & they worked OK. I have not done anything further of late because of computer problems. (got a virus in it & had to have hard drive scrubbed & Windows XP reinstalled). Took me ages to reinstall software & transfer saved files back from external hard drive.
Bob

[squier]

output valves will work just as well as RF valves in those type of receivers. If you look on the web there are numerous circuits for simple SW TRFs using an 807 (or 1625) transmitting valve using 12 volts HT apparently there's enough output to drive a small loudspeaker.

I also saw in the latest SPRAT (G-QRP mag) a circuit for a little SW TRF that uses 3 x EL84's, one as detector, that the other two as 1st and 2nd audio

[Skywave]

Rumour has it that the ubiquitous 6V6 can put out quite a few watts up to several MHz !

Al.

[Variometer]

Yes, I have found that a lot of mains valves work very well on extremely low HT voltages. Have not done anything radio of late, but will probably be coming back to it soon. I have really given up the high voltage mains stuff. Not that I have had a bad shock or anything, just realised that they will work quite well on 100 volts or less.
Bob

[Al (astral highway)]

Originally Posted by Variometer

, just realised that they will work quite well on 100 volts or less.
Bob

yes, it's useful to bear in mind that you can build RF circuits (=low power) especially using a much lower HT+ than full-wave rectified mains. But of course you will have to change grid bias R values