Lash up Direct Conversion Receiver

[AlanBeckett]

Lawence,
Mine's a C and it works beautifully. I wonder if it was refurbished before it was flogged off.
Alan

[pwdrive]

It might well have been Alan, mine definately wasn't, it's been through the wars a bit, it works well though except for the backlash.

Lawrence.

[pwdrive]

Ah - Ha!  PW = Precision Worm (drive) . . . . I see; thank you Lawrence. 

When I read your reply, at first it sounded like someone's address:

H.R.O. esq.,
National House,
PW Drive,
Somewhere Stateside.

. . . . or something like that!

Al.

Malden, Massachusetts.

Lawrence.

[pwdrive]

Another update on the regen detector, found a 6J5 triode today so rewired the valve base to take that, it works a lot better than the 6AC7, managed to get the 550 khz tuning range from 2.45 to 3.05 mhz but using quite a low L/C ratio by removing 6 turns off the coil and putting a 270 pf in parallel with it, the 400 pf tuning capacitor in series with a 220 pf is also connected across the coil. Regen can now be obtained across the whole of the above tuning range whilst maintaining a low L/C ratio.
There is no pulling at all when tuning the mixer grid and very little when going in and out of regen on SSB/CW.
Regen starts with an anode voltage of around 22 volts with full oscillation at around 30 volts, it seems to be very drift free once warmed up.
Whilst receiving an SSB transmission there was almost no shift of pitch when altering the receivers HT plus & minus 20 volts above and below it's 200 volts. design voltage. Regen is now obtained by varying the det anode voltage via a pot. divider across the HT. The audio output from this det. is slightly less than the previous arrangement but it's still ample for the triode pentode O/P stage.
I think I have just about gone as far as I can go with the det. arrangements, circuit wise I am happy with it for the moment, ideally though I would prefer a B9A or B7G valve for it if I can find one with similar characteristics as the 6J5.

Lawrence.

[Skywave]

Lawrence: you wrote: " . . . ideally though I would prefer a B9A or B7G valve for it if I can find one with similar characteristics as the 6J5".

Then you might care to consider the EC90 (6C4; L77); B7G base. Its Ra and gm figures are very similar, whereas the inter-electrode capacitances are substantially lower. However I'd be surprised if these differences caused a problem, since if all else fails, you can add on a bit of external capacitance to compensate! And as you and I know, a common problem with doing a 'sub' for a different valve (that is operating at R.F.) is usually having too much inter-electrode capacitance with the 'sub'

You're making good progress; I'm watching with interest; keep at it.

Al.

[pwdrive]

Thanks for that Al, I will keep that in mind, I am guessing the EC90 was designed as an osc. valve.

Lawrence

[pwdrive]

Just trying out my new laptop:

As previously stated the work on the anode feedback regen det stage has more or less come to an end, so time for a go at a Hartley regen det (cathode tap) for this I used a 6SH7 pentode which I had spare, regen is controled by a screen grid pot, the same coupling arrangement from the mixer was used and remained unchanged (choke/capacitor coupling) After removing the previous windings I wound another coil on that same 1.5" former winding 20 turns over a distance of approx 1.5" (approx 10 uh) this spacing enables enough room select and solder on the cathode tap, the best results were obtained at a tap 1 turn from the chassis end.
It tunes the designed 3 to 3.5 mhz frequency range ok, at the maximum frequency the capacitive element of the tuned circuit is approx 170 pf so the L/C ratio is still quite low which is part of the design requirement.
Regeneration is very smooth throught it's advancement, when in full oscillation there is no pulling when tuning the RF in to the mixer and virtually none when advancing the regen control from the onset of oscillation to the point of full oscillation, when the regen control is further advanced there is a small pulling effect which I guesstimate is around 200 to 400 cycles (based on listening to a CW & SSBtransmissions) and a drop in gain which is to be expected.
All in all a good result, out of the two methods the Hartley one has the edge for me.
It is to be noted though that all regen designs suffer from overload when presented with a very strong signal, to that end the final design will have a simple flick switch attenuator in the aerial input circuit.
I gave it a good test on 40 and 20 meters last night and it works very well. (Copied a Ham from Boston US)
If I can get my new laptop to communicate with the scanner/printer etc I will post up a couple of diagrams of the story so far.

Lawrence.

[Skywave]

I'm following your progress with this, Lawrence.

Al. / May 11th. //

[pwdrive]

Managed to get the new laptop all workie workie:

The attached image gives details for the prototype Regenerodyne to date, there will be more additions as and when, probably an RF amp and Local Osc. and a single envelope AF amp & Output stage.
The diagram gives details of the coil winding for a tunable IF of 2.95 to 3.55 mhz.
Not shown on the diagram is a 15 pf variable (fine tune) across the 300 pf regen tuning capacitor.
The HT used was 200 volts, you can use more or less but the potential divider values of R8 & R9 might need to be altered in order to maintain the regeneration action over the full 600 khz tuning range.
Final observations on the circuit as shown:
Tested from MW up to and including the 18 mhz amature band.
Full regen available over the full 600khz.
Regen action is smooth (no plops)
Regen osc stability is very good (checked against Shannon Volmet over a 1 hr period)
Regen tuning rate remains unchanged which ever band is selected (as was expected)
No pulling of the regen when altering the RF tuning.
No pulling of the regen when going in and out of full regen.
+ / - 20 % variation of the HT supply made no significant differnce to the regen osc frequency stability.
Very strong signals can overload the regen det (as they do) so some form of RF attenuation will be needed in the aerial input.
The upper tuning range of the regen det coincides with the lower end of the 80 meter amature band, I thought this might be a problem when tuning the RF input to the mixer for that part of the band and to a certain extent it was, regen det oscillation was breaking through but did not present to much of a problem as offsetting the RF tuning slightly to the higher side did not have to much effect on the signals being received, tuning the regen det further up the band increases the seperation and reduces the effect of regen det osc breakthrough.
Local osc injection source for testing was an Avo CT 378A signal generator, output attenuators set to give 25 mv output.
The gen. is set to Fs + IF, eg: for 7 mhz amature band the gen. is set to 10.5 mhz the regen det (2.95 to 3.55) will then tune from 6.95 to 7.55 mhz the image frequency range at this setting is 13.45 to 14.05 mhz, ie: 2 x IF above Fs.
L4 should be wound with tinned copper wire to facilitate ease of connection and fine tuning of the cathode tap, L3 is enameled coated copper wire.
C20 is not a misprint, it is 4.7 pf.
(C8 A) is fitted and C10 should be omitted...that's my drawing mistake.....
For testing and evaluation purposes L1 & L2 were any old aerial/RF tuning coils I had hanging around.
If using a 6AC7 then note that it has a greedy heater.

Lawrence.

[Skywave]

Hi Lawrence,
O.K., I've briefly read that and stored the cct. diag.; will read your report and study the diag. at a more convenient time when I can give it my full attention that it deserves: at present, my 'in-tray' is rather full! Then I'll come back to you here and talk about it.
Anyway, well done in seeing it through to this stage. I know how daunting and demanding 'home-brew R & D' can be!

TTFN,
Al / May 14