Lash up Direct Conversion Receiver

[Skywave]

It's not harmonics of 1 MHz that you need: it's harmonics of 500 kHz: your regenerative detector tunes over a 500 kHz range.
I'll explain that in depth . . . .

Example:
To receive signals in the freq. range 7 to 8 MHz with your regenerative detector tuning 2.5 to 3.0 MHz, for the input range of 7.5 to 7.0 MHz, the local osc. will need to be 10.0 MHz (subtractive mixing): 10.0 minus 7.5 = 2.5; 10.0 minus 7.0 = 3.0.
For an input range of 8.0 to 7.5 MHz, the local osc. will need to be 10.5 MHz (subtractive mixing): 10.5 minus 8.0 = 2.5; 10.5 minus 7.5 = 3.0.
Therefore the local osc. needs to be 10.0 MHz and then 10.5 MHz . . . and nothing in between.

Now extend that concept over the entire signal input range: 5.0 to 15.0 MHz.
On the basis that the oscillator is on the high side of the signal, we have [osc. freq.] minus [RF freq.] = [regen. tuning freq.]

So, 8.0 MHz minus 5.0 MHz= 3.0 MHz; 8.0 MHz minus 5.5 MHz = 2.5 MHz: regen. tunes 3.0 to 2.5 MHz
And: 8.5 MHz minus 5.5 MHz = 3.0 MHz; 8.5 MHz minus 6.0 MHz = 2.5 MHz: regen. tunes from 3.0 to 2.5 MHz
and so on, up to:
17.0 MHz minus 14.0 MHz = 3.0 MHz; 17.0 MHz minus 14.5 MHz = 2.5 MHz.
17.5 MHz minus 14.5 MHz = 3.0 MHz; 17.5 minus 15.0 MHz = 2.5 MHz
This covers the entire R.F. input range from 5.0 to 15.0 MHz.

Thus, since the R.F. stage tunes from 5.0 MHz to 15.0 MHz, the local osc. needs to be in 500 kHz steps from 8.0 MHz to 17.5 MHz.

Now if the regen. detector tuned over a 1 MHz bandwidth, then the local osc. would be in 1 MHz steps . . . in essence, what happens in the Racal RA-17 using a 1 MHz crystal osc. and a harmonic generator: the "kHz tuning" covers a continuous 1 MHz bandwidth; the "MHz tuning", in effect, selecting the appropriate harmonic from the 1 MHz crystal, and thus acting as a 'harmonic selector switch'.

Finally, it is obvious that if only harmonics of 1 MHz were required, rather than harmonics of 500 kHz, the total number of osc. freqs. required would be halved. I suspect that this, in turn, would lead to an easier method of producing the required local osc. freqs.

Over to you, Lawrence!

Al.

[pwdrive]

Thanks Al for your interest, I wish I could put what's in my head into text via the keyboard like what you do,
I have always had a problem with this ever since I can remember so thanks for that.
I need now to explain the main reasons why I initially chose the IF frequency and the IF bandwidth.
The IF center frequency of 2.75 Mhz was chosen as the best all round frequency taking into consideration what is required from the design.
As you will know it's all a trade off between image ratio and frequency stability when choosing the frequency of a tunable IF. From the outset I planned a double tuned RF stage to allow me to reduce the IF for image ratio purposes, the other benefit of a lower IF is that it makes the tunable regen det a lot easier to design and build when taking in to account such things as stability/drift/smoothness of regeneration etc.
The frequency selected has also made it easer to design the regen det tuning coil for a lower than normal L/C ratio whilst still maintaning it's ability to go in and out of regeneration across the whole of the tuned circuits bandwidth, the low L/C ratio coupled with a lower than usual grid leak capacitor value masks the effect of G1 interelectrode capacitances when the regen is brought into oscillation and beyond, this results in an almost drift free oscillation as the regen is advanced, a prime requirement for SSB/CW tuning and listening, the other benefit is that by increasing the amplitude of the oscillation even further the effects of strong signal overload are much reduced all this can now be done with almost no drift.

Reason for initial choice of a 500Khz tunable bandwidth:

When tuning all my old comms receivers the one that stands out for me is the tuning mechanism used in the National HRO, for simplicity and resetability it takes some beating considering how long ago it was designed, the HRO when fitted with it's banspread coils tunes approx 500 Khz on a reduction drive of 20 to 1 (10 turns) this gives a scale increment of 1 Khz, this system feels right when operating it even when using the xtal single signal tuning the turns to frequency ratio feels about right.
My initial idea was to use an old HRO drive mech. if possible.

However none of this is set in stone yet, a 1 Mhz bandwidth could still be employed but a compomise has to be made, in order to obtain full regeneration across a 1 Mhz bandwidth while maintaining a low L/C ratio the ratio of the cathode tap to main winding has to be reduced, this tends to lead to a sharper ( less linear) regen action characteristic rather than a softer one (more linear) a way to reduce this is to alter the pentodes operating characteristic by reducing the anode load however this results in less overall AF gain in that stage but this could be compensated for with a better final selection of af amp/output valve (higher gain) another way would be to use an reverse log pot for the regen control.

The Harmonic Generator:

I have a feeling that the harmonic generator/amp is where it will succeed or fail it will depend on how much drive I can get out of the Xtal osc and how selective the tuned anode tank is in the harmonic generator/amp and wether or not it will supply a minimum of 200 mv mixer injection on all the selected harmonics while rejecting the unwanted harmonics down to a level that's insufficient to drive the mixer, bit of a tall order but you never know. I will start off experimenting with a 1 Mhz xtal or possibly use the sig. gen. output and overdrive it.

I managed to tap into the 500 Khz xtal calibrator in the B40D this afternoon and via a lashup tuned circuit I managed to extract the 20th harmonic (10 Mhz) and succesfully tuned the 40 meter amature band I have no idea what the final voltage of this harmonic was but the volume control was only one point down as compared to using the 200 mv from my sig. gen. I tried for the 34th harmonic (17 Mhz) to cover the 20 meter band but no joy, wether that's down to the xtal osc or my harmonic extraction lashup I am not sure....probably a combination of both. It also worked on RAF and Shannon Volmet on 5.4/5.5 Mhz. There were some funny hetrodyning issues but not had chance to look into that yet.

Lawrence.

[Skywave]

O.K. Lawrence, read & understood. My last post: thanks for the compliment. At the time you wrote it you were not aware that it took me about two hours to get that post technically correct and presented in a style that both made sense and one that I was happy with.

I've a busy non-radio day today, so must go. Will talk further about your project again soon.
TTFN,

Al.

[AlanBeckett]

Lawrence,
I've just nipped down to the garage to check that the 17MHz Calibration marker on my B40C is there OK and it's stopped working, no audio at all
It was OK yesterday when I was listening to Radio Stoke. I'll have to get it on to the bench.
Alan

[pwdrive]

Lawrence,
I've just nipped down to the garage to check that the 17MHz Calibration marker on my B40C is there OK and it's stopped working, no audio at all
It was OK yesterday when I was listening to Radio Stoke. I'll have to get it on to the bench.
Alan

Happy lifting Alan, at least the audio is all on the main bottom deck, presumably you have a manual for it, if not I have the full BR1617 if you need anything copying give us a shout.
Mine is also tuned to our local radio station it's a permanent fixture on the bench and is more or less in daily use.

Lawrence.

[AlanBeckett]

Lawrence,
I have the Manual, thanks.
Radio Stoke on 1503 kHz puts out 1kW 30 miles away and points in the wrong direction. However, I can listen to the football OK.
It's on a shelf to the side, above and at a right angle to the bench. Getting it on to the bench will not be too bad. Getting it back will be another matter.
Alan

[pwdrive]

Good luck with it Alan, let us know what the problem was when you find it.

Lawrence.

[pwdrive]

O.K. Lawrence, read & understood. My last post: thanks for the compliment. At the time you wrote it you were not aware that it took me about two hours to get that post technically correct and presented in a style that both made sense and one that I was happy with.

I've a busy non-radio day today, so must go. Will talk further about your project again soon.
TTFN,

Al.

I have now extended the regen det tuning range to 1 Mhz in a downwards direction so it now tunes from 3 to 2 Mhz. I did a rough and ready test for image rejection at 7 & 8 Mhz, the threshold for the audiability of a cw wave injected into the aerial terminals at the two spot image freqencies of 12 & 13 Mhz from my spare sig. gen.was about the same after as it was before the range was extended.

I have been scanning the web for Harmonic Generator ideas, the only one using valves that turns up is our old freind the RA17/RA117 so I will have to use that as a model to base mine on.
The harmonic mixer they use employs an interesting valve the 6AS6, wouldn't mind getting hold of one of these as they seem to be available and see how it performs with a view to incorporating it as the mixer in the final design.

Lawrence.

[AlanBeckett]

Lawrence,
The B40 one claims to produce harmonics up to 30MHz. Of course the output is not tuned in any way.
Alan

[pwdrive]

Lawrence,
The B40 one claims to produce harmonics up to 30MHz. Of course the optput is not tuned in any way.
Alan

They claim that and indeed it does Alan. it's an unusual circuit if you look at it, mine had trouble booting up when I first got the receiver, removing the xtal and cleaning it and the plates seemed to cure the problem, the screen volts are quite critical to it's operation, a lower value screen feed resistor is switched in when switching the calibrator on in order to increase it's harmonic output. I tried extracting the harmonics from it to inject into my regenerodyne project, only up to the 20th harmonic did it have enough output to drive the mixer, having said that it was designed as a calibrator/marker and 'nowt else.

Lawrence.