3-valve reflex 1925 kit?
 

Picked up from a local carboot for £10 this looks like it was a kit as there are professional looking markings on the board. I would guess around 1925, anyone recognise the kit? It is built into a basic box made from pine T&G that is now a little worm eaten and needs a rebuild. The intervalve transformer still measures OK but I will have to look out for some plugin coils.

Re: 3-valve reflex 1925 kit?
 

The Dubilier grid leak resistor is toast... unusually all connections have been soldered but it uses the square wire typical of the period.

Re: 3-valve reflex 1925 kit?
 

Probably not a kit but a typical home constructed set of the time. Sets of transfers of the type on the panel were available for amateur construction at this date. Panel quite neatly laid out.

Re: 3-valve reflex 1925 kit?
 

It actually says '2-valve reflex set' and the middle socket is for an RF transformer. There is a publication here https://www.6v6.co.uk/vcomp/manuals/...ers%201924.pdf and it is similar to 'Fig. 33'.

V1 is the RF amplifier with a tuned aerial circuit, it drives V2 detector via a tuned RF transformer, the detector drives an regen coil back to the aerial coil and an AF transformer which passes the audio back to V1 which now serves as the audio output.

This is looking interesting but I need to read up a bit so I can work out how to wind the RF transformer.

Re: 3-valve reflex 1925 kit?
 

These early Ormond variable condensors are not logarithmic and of pretty low capacitance, about 120pF fully meshed. I am looking at a scrap IF transformer as a possible for the RF transformer. That just leaves the plug-in coils but I think I have some but doubt they will be the correct values.

Other than regeneration the only method of controlling the gain would be by using the filament rheostats. These are very small and I suspect they may simply be a carbon block with a screw down terminal. Have to see if that works with PM series valves and varying V1 filament is also going to reduce the audio output power.

Re: 3-valve reflex 1925 kit?
 

The rather delicate 'rheostats' have not appreciated being soldered to and one of the screw terminals has come away leaving an opening to the inside but this has answered the mystery, they are filled with carbon dust which is compressed by the screw down adjuster. The good one appears to work pretty well with gradual decrease down to about 12 ohms. It should be possible to fix the broken terminal with conductive paint and then add some strength with some araldite. They will be bypassed for the 2V PM series valves in any case.

The rest of the radio has been dismantled, stripped of solder, cleaned and the brass polished and is in the process of reassembly using solder tags rather than the original weld everything together approach.

Re: 3-valve reflex 1925 kit?
 

I went back to the carboot this morning and got lucky and now have the parts that were found in the loft with the radio.

There are a bunch of plugin coils including a fascinating Blue Spot one that is tapped, as well as what I assume are the original valves with intact filaments, a Nelson Multiple in original box and a Marconi DER in a later box.

Re: 3-valve reflex 1925 kit?
 

Reassembly is largely complete but unfortunately the RF transformer I received, used to couple V1 to V2 and has a tuned secondary, measures as 1mH and according to my calculations that places it in a band between UK LW and MW stations. Should get me going but I will still need to make a higher frequency version.

Re: 3-valve reflex 1925 kit?
 

It's reassembled with 2V valves and doing something but mostly squealing and considerably more so with no aerial. I have received a station but I think it was down in the 700KHz area so probably need to do more work on the RF intervalve transformer.

The RF Transformer that came with the set is for a low frequency (between the MW and LW) and seems to be made of identical windings wound in a bifilar manner but the one I am using is from a crystal filter coupling transformer and is not going to be as closely coupled so offering gain and selectivity I do not need. With the original bright emitter valve I could have fiddled with the filament current to adjust the gain but I am not planning on experimenting with the pristine 1924/5 Nelson Multi.

Next stage is to stick a scope on it and see exactly what it's oscillating at then strip some of the wire off the RF transformer secondary.

Re: 3-valve reflex 1925 kit?
 

I played a bit more and got a couple of stations but instability remains a problem and a signal generator shows it to be good up to 1MHz so I won't mess with the coils yet. The reflex design means more opportunity for stray RF feedback so I will focus on getting it stable without an aerial. There are some suspect original decoupling capacitors I will look at and also make sure my PSU connections are all locally decoupled.