Golborne Vintage Radio

A nice idea with the valve bases Amie. I have seen it done before but a tidy solution.

One option that I have seen mentioned for AGC but I have not seen a working example of is to use a 1j37b two x g1 valve and use one of the grids for gain control. I have done a little bit of testing and it does seem a reasonable possibility although I have only tried it at low frequencies / DC as yet. I don't fully understand the process but it seems more like the control of gain on one grid is more like a tap on a fixed bias than shifting the bias on the signal grid which is the usual method (giving distortion / clipping near cutoff). Please let me know if I have that wrong. My take is that the input range may well be a problem with any of them.

At least with initial work I plan to use agc on my RF stage only and make it switchable with optional manual gain. With a more traditional RF valve I do feel I can avoid an overload of issues to check on. It will mean lots of thinking about how to handle the agc control though. Your delayed amplified AGC sounds about right Kalee. I'm not sure if I am being ridiculous or not but I even had thoughts of using an op amp in the AGC line to give ready control of voltage levels (some of my thinking does get 'over the top' at times). Further down the line ideas may well present themselves for effective use of rod pentodes, but a good stable mixer / IF / audio setup will help to categorise problem areas. This guides me to avoid total rod pentodes at first and splitting stages into modules, risking poor efficiency, so that I can isolate problem areas. Something I have learnt over the years is that easy maintenance / faultfinding is a must, especially if the paperwork is not perfect, as, say, in two years time when something fails a mass of electronics seems just like a mass of electronics and not readily fixable.

Another point I have seen little information on though is noise performance. My suspicion is that rod tubes should be fairly quiet as they should have far less / no problems with partition noise (as the grids are not grids and not in the electron beam). There may be advantages in terms of lower voltages too? This is one reason why I have built a 1j37b mixer (another being that my pantry transmitter uses a very simular modulator). Although it is obviously different I am reminded of a beam deflection mixer which is apparently quiet. Although I have bought a couple of noise diodes, some while ago, I have not got round to assembling a test setup yet, never mind how to use one to determine mixer noise.

It is interesting how we setup these ideas to improve reception Amie. I have my two aerials, one loop and one large ferrite rod, working together. I found this by accident as the inter-react. For a station like Caroline they are easy to use though as they are readily tuned and aligned but they would be more fiddly for DX work.

Tracy

Hi.
On my 1J42A pantry transmitter if I varied the control grid voltage of the one that wasn't used for the oscillator feedback, the result was the resultant output varied as per the DC applied. This is logical as I used this grid for the AF input. The 1j37b of course is bi control grid as well. I'd say that using these Bi grid valves that AGC could be applied to one grid.
One project I started was using a 1J42A as a frequency changer, my issue was the lack of coils etc but I did use one grid for local osc and the other for RF input, there was a resultant IF at the anode. It wasn't very good and I got fed up with it and abandoned the whole thing. I'd say the 1j37b would be a better proposition though. May go back to it one day using the higher voltage valves.

As I see it, putting AC on the two grid rods varies the current just as normal, but having the two rods at a different DC voltage bends the electron stream so that it partially misses the anode. (What happens to the electrons? Presumably they get caught by the screen-grid rods?) Thus, gain to anode is varied.

I do have a dual control rod valve but haven't given that option a try. I did start playing with an idea that showed some promise for gain control of the likes of IF amplifiers, that used a FET as a voltage controlled resistor in series with a coupling capacitor grounding the cathode circuit signal, of course, being directly heated valves that would mean in a practical circuit the use of a common mode choke which I haven't got round to yet, but it certainly works with an elevated filament battery, (I think that's the term?).

I would agree that the rod pentodes seem a lower noise option to wound grid types.

Loop aerial...
I read about a type of loop aerial that fascinated me so that I made one. It is tuned by the connected radios own tuning capacitor with a switched RCA phono socket fitted to the set to switch out it's own internal aerial. It uses a toroidal inductance matching transformer. The set I use with it most is an ER Sky King that is an ideal candidate because it has a single coil loop with a loading inductance for lonwave reception... The loop makes the old ER into a very sensitive set indeed.

Here is a link to KR1S webpage description of his Hoop loop.
http://kr1s.kearman.com/html/hooploop.html
Picture of the actual loop.
[attachment=20929]

(01-09-2021, 09:33 PM)Amie Wrote: [ -> ]I would agree that the rod pentodes seem a lower noise option to wound grid types.

I'm getting to think that, too.

The screen-grid current in the rod pentodes is a small fraction of anode current, well under 1/10, which contrasts with a traditional, helical-grid pentode where screen-grid current is often around 1/6 of the total current.

That being so, partition noise can be expected to be rather smaller in the rod pentode.

I do have an intention to try building a magnetic pick-up preamplifier and equaliser using rod pentodes, but that'll have to wait another day!

Hi.
Did that with 1J24Bs in my black box. Initially running a Sonotone 9TA now a shure M78 with excellent results. HT derived from the internal amp as well as filament supplies that are regulated.

Here is the set working driving a speaker.
https://youtu.be/b_gs7CDwSRs

That's pretty impressive, Trevor!

Loudspeaker volume from the 1j42a, too, from 9V HT!

I like it.

(Any chance of posting the circuit diagram of the phono pre-amp?)

(01-09-2021, 08:57 PM)Murphyv310 Wrote: [ -> ]Hi.
On my 1J42A pantry transmitter if I varied the control grid voltage of the one that wasn't used for the oscillator feedback, the result was the resultant output varied as per the DC applied. This is logical as I used this grid for the AF input. The 1j37b of course is bi control grid as well. I'd say that using these Bi grid valves that AGC could be applied to one grid.
One project I started was using a 1J42A as a frequency changer, my issue was the lack of coils etc but I did use one grid for local osc and the other for RF input, there was a resultant IF at the anode. It wasn't very good and I got fed up with it and abandoned the whole thing. I'd say the 1j37b would be a better proposition though. May go back to it one day using the higher voltage valves.

I love the video Trevor. You have got them very well and your radio works well too!

If I may I would just like to give the (probably rather rough) figures I gathered from my 1j37b mixer setup today. I know that your mixer / oscillator setup is different but the figures I give may be of some value. I didn't mention in my earlier post (because I completely forgot) that I had tested it earlier with a FET oscillator I have been working on. I put this together inspired by an LTSpice example circuit, just adding a buffer amplifier. with a view to possibly building a valve version at some point, although it may be more sensible to stay with the FET for stability, drift etc. Using this for local oscillator and my signal generator for signal input. As previously mentioned the 1j37b mixer is followed by an emitter follower matching to a 455kHz ceramic filter then to a two stage 1j24b IF amp. All the valves have single coil tuned circuits, currently tuned to 455kHz. With the setup I provided a capacitor feeding a 47kOhm resistor as a nominal load. During testing all coils, when loaded by the oscilloscope probe, were retuned as required. All testing / measurements were done using the oscilloscope so frequency measurement, particularly, is nominal. My signal generator, and the measurement equipment, are limited so the minimum setting on output is used which is somewhat higher than ideal. This may well mean limitations in accuracy etc but I think the gist of things is correct. Due to things being a bit of a lash up I am limited. The results (all voltages p to p). HT supply 60V (except oscillator):

Signal input 12mV 1.16 MHz to 1j37b g1a

Oscillator drive 1.5V 1.61 MHz to 1j37b g1b

Frequency changer (ceramic filter) output 120mV 455 kHz (noisy waveform but may be effect of probe or setup).

First IF amp 1j24b output on second IF amp grid resistor (100kOhm) 5.4V

IF amplifier output (across 47K load) 58V

As yet I have only used unmodulated signals although the signal generator output is messy. I think the figures show that this is a goer but careful signal handling will be needed. The circuits I use tend to follow the Russian examples, the actual mixer here being closely influenced by that of the R-326 military radio.

If anyone has any questions please don't hesitate to ask.

Tracy

(02-09-2021, 03:54 PM)Kalee20 Wrote: [ -> ](Any chance of posting the circuit diagram of the phono pre-amp?)

Indeed I will. I need to see if I've still got it on the PC and it's not been lost after an HDD failure a few years back. If not I'll redraw it over the next few days.