10-05-2018, 05:49 PM
Hi everyone,
I have been doing other things, with various Russian valves. I did a little with a regenerative receiver with a rod pentode after being inspired by Trevor. I need to tidy it at some point, but have received a couple of stations with a lashup (Medium Wave).
My interest with radio is more toward a superhet though. Elsewhere I did come across someone suggesting using a Franklin oscillator, which looked interesting. I put together a test circuit using an ECC82 (easily done as my (homebrew) valve test equipment is modular). I got the shown skeleton circuit to work, (details in attached working sketch). Then I progressed to an 6N16b. This is a Russian miniature double triode. Each section seems identical, and cannot easily be determined as far as I can see as there is only a diagonal in the glass base to indicate pin location. This works similarly with both sections. In a similar manner to someone online who has used an IC socket for these I soldered direct as per the sub-diagram on my sketch. I had previously tried making a module using the 6N16b the same as with my rod pentodes, but got hopelessly confused with connections. The IC socket type layout works well for me and seems tidy (see attached photo). I didn't use a socket as I am not sure of reliability with wired connections. The circuit in the photo is intended for a local oscillator for use with a MW superhet (with and addition tuning capacitor). This, I have just used for testing at the moment, but I got tracking over the correct range for 455 Khz IF (just the oscillator, no RF circuitry). I plan to continue with this at some point.
Now to the point – I decided to try using two 1J24b's in a Franklin circuit. I made a new module using two of these (see photo). I did have a few issues, I think due to the low gain of the 1J24b's. I have noticed this before, and have used 1J29b's in other circuits with better results. I did find that, in using this type of circuit with a double valve board, I had g1 connection issues, using long wiring runs. This, I think, added additional capacitance etc and tuning was a bit out. The circuit, as made, works reliably, but only to a bit over 1MHz. It was stable, but I think (especially in comparison with the 6N16b or even ECC82 which I found work much better in this circuit), I will not continue at this point as I can see redesign of things would really be required and I have other fish to fry.
On another point, I have been thinking of an amplifier specifically to feed my record deck into my Pantry Tx. I had the thought of using a 1J37b as a mixer, feeding each channel into a seperate grid.
As yet it is in early stages, but one thing relevant I found is that the gain from each g1 (the other being grounded) does not seem to be half of the combined total (both grids driven), as found by Joe Sousa (Radiomuseum). For RF it may be of less relevance, but for an audio mixer it would be more important. I have not done formal testing, but on more than one of these tubes I have found a difference. I would suggest that anyone intending using one of these valves in a critical application check. This afternoon I set up a circuit to vary the bias hence vary gain. I got a rough balance when g1a was biased to -1.5V, and g1b at Ov (ground).
Tracy
I have been doing other things, with various Russian valves. I did a little with a regenerative receiver with a rod pentode after being inspired by Trevor. I need to tidy it at some point, but have received a couple of stations with a lashup (Medium Wave).
My interest with radio is more toward a superhet though. Elsewhere I did come across someone suggesting using a Franklin oscillator, which looked interesting. I put together a test circuit using an ECC82 (easily done as my (homebrew) valve test equipment is modular). I got the shown skeleton circuit to work, (details in attached working sketch). Then I progressed to an 6N16b. This is a Russian miniature double triode. Each section seems identical, and cannot easily be determined as far as I can see as there is only a diagonal in the glass base to indicate pin location. This works similarly with both sections. In a similar manner to someone online who has used an IC socket for these I soldered direct as per the sub-diagram on my sketch. I had previously tried making a module using the 6N16b the same as with my rod pentodes, but got hopelessly confused with connections. The IC socket type layout works well for me and seems tidy (see attached photo). I didn't use a socket as I am not sure of reliability with wired connections. The circuit in the photo is intended for a local oscillator for use with a MW superhet (with and addition tuning capacitor). This, I have just used for testing at the moment, but I got tracking over the correct range for 455 Khz IF (just the oscillator, no RF circuitry). I plan to continue with this at some point.
Now to the point – I decided to try using two 1J24b's in a Franklin circuit. I made a new module using two of these (see photo). I did have a few issues, I think due to the low gain of the 1J24b's. I have noticed this before, and have used 1J29b's in other circuits with better results. I did find that, in using this type of circuit with a double valve board, I had g1 connection issues, using long wiring runs. This, I think, added additional capacitance etc and tuning was a bit out. The circuit, as made, works reliably, but only to a bit over 1MHz. It was stable, but I think (especially in comparison with the 6N16b or even ECC82 which I found work much better in this circuit), I will not continue at this point as I can see redesign of things would really be required and I have other fish to fry.
On another point, I have been thinking of an amplifier specifically to feed my record deck into my Pantry Tx. I had the thought of using a 1J37b as a mixer, feeding each channel into a seperate grid.
As yet it is in early stages, but one thing relevant I found is that the gain from each g1 (the other being grounded) does not seem to be half of the combined total (both grids driven), as found by Joe Sousa (Radiomuseum). For RF it may be of less relevance, but for an audio mixer it would be more important. I have not done formal testing, but on more than one of these tubes I have found a difference. I would suggest that anyone intending using one of these valves in a critical application check. This afternoon I set up a circuit to vary the bias hence vary gain. I got a rough balance when g1a was biased to -1.5V, and g1b at Ov (ground).
Tracy
