Lash up Direct Conversion Receiver

[pwdrive]

I have been looking a bit more into the differnt coupling methods from an RF amp anode to a mixer grid.
This was done on some of the receivers I have hear, the methods fall into the following types:

A) Marconi CR300, RF amp with tuned grid and tuned anode (load) anode then capacitively coupled to an untuned mixer grid.

B) Hallicrafters and HRO's, RF amp(s) with tuned grid and a high inductance self resonating anode primary (anode load) inductively coupled to a secondary tuned mixer grid winding.

C) Eddystone 358X, RF amp with standard tuned grid input with low inductance anode primary winding (anode load)
inductively coupled to a secondary tuned mixer grid winding.

D) Hammarlund HQ150, RF amp with standard tuned grid and tuned anode (anode load) capacitively coupled to an untuned mixer grid, the anode connecton point however is a tapping up from the tuned circuits signal cold end (the HT end)

I know from experience that the method used in A) needs a very carefull and precise layout of the components concerned in order to prevent self oscillation at higher frequencies so I am discounting this method for the Regenerodyne.
Method B) works very well in the receivers concerned however high inductance primaries would need wave winding and I don't really want to get into that, so this method is discounted.
Method C) is the more coventional method, on the coils I checked the primary to secondary ratio is approx 1 to 4, this method is feasable for the project.
Method D) Is simple so far as coil consruction is concerned, so is also a contender.

A couple of questions now to members, if method A) or B) was used say in a cascode amp can anyone forsee any problems apart from one of the tuning gangs stators being at HT?
If either cascode or single a pentode are used would isolating the HT from the tuning gang by the use of a high value silver mica capacitor affect the overall Q of the circuit to any serious extent?
Any comments/ideas welcome.

Lawrence.

[pwdrive]

Quote, pw_drive: "it is probably due to direct pickup by the regen det tuning coil itself".

Certainly sounds like that's the reason. After all it only takes a few pF of 'hidden' coupling to produce that symptom.

Incidentally, have you retained that R.F. choke in the anode of the R.F.Amp., or did you implement my suggestion of replacing it with an R.F. transformer with a variable-capacity tuned-secondary?

Al.

Hello Al, yes I'm pretty sure it's just stray pickup, it does not make to much of a problem, I feel sure when fully screened that it will be reduced by a lot or maybe even eliminated. I would be to embarrassed to post a picture of this mess but I gues you can imagine what form it takes.... coils hanging on wires everywhere...!
I still have the psuedo RF choke in the anode of the RF amp this is then capacitor coupled to the primary winding of the tuned circuit feeding G3 of the mixer valve.
The reason for this at the moment is that the coils that I found and that I am using all have a common pri/sec cold end ie: 3 connections therefore I cannot feed the primary with HT. I have a spare former for one of these coils so I might have a go at winding one with an isolated pri/sec ie: four connections and then connect the primary as the anode load in the RF amp. Eventually I will wind all new coils as the ones I am using (RF grid tuning & Mixer grid tuning differ a bit in their inductance values, I have managed to get them tracking "reasonable" but you have to remember that this whole test receiver is built from junk bits and I only have a limited number of resistor and capacitor values. I have the full coil winding details for all the B40 and also all the HRO coils which I can use as a guide when designing my own coil set as well as numerouse web based tuned circuit calculators (remember log/antilog tables?)
In all that I am pleased with the results so far as I am trying to design this thing so the controls do what they are supposed to do with the right operating feel to them which as you know is important for SWL especially on the Amature Bands.
Interestingly the Murphy B40D I have just found out uses a similar coupling arrangement from it's 2nd RF amp to the Mixer grid tuned circuit, they do not use a choke however but two carbon resistors in parallel, I think the parallel arrangement is just for dissipation purposes and I am guessing that a resistive load is better than a choke considering that the B40D has to tune from MW up to 30 Mhz. Interstingly the HRO uses high inductance wave wound primaries designed to self resonate just outside the low end of the band being tuned in order to provide gain compensation at that end of the selected band, later on Hallicrafters hit on the same idea, the primaries in the B40 are only a few turns though.

Lawrence.

Hello again Al, when the power came back on this afternoon I rewound the primary of the coil as has been mentioned so that the primary and secondary windings were fully isolated ie: 4 connections as opposed to the original 3. pri. connected as anode load, sec tuned and connected to the mixer grid. The test receiver works fine with this arrangement, the gain would now appear to be a bit higher at 14 mhz than it was with the choke capacitor coupling (the coil tunes from 5 to 15 Mhz) I suspect that if I had a matching set of coils for freqencies higher than 15 Mhz then the gain would increase proportionatly compared to the choke and capacitor method.

Lawrence.

[Skywave]

Hi Lawrence,
This is all getting a bit complicated, isn't it? "Complicated", that is, when only the written word is used to explain these things: hence, the value of a drawing, then with a few supplementary comments thrown in.
So let me try to consolidate what you have there now and where you want to take this project; this also gives me the chance to verify my understanding of those two things.

At the moment, you have an R.F. amp. with an L/C tuned grid (transformer coupled from the aerial). Its anode cct. has an untuned transformer primary. Its secondary comprises the 'L' of an L/C tuned cct. feeding the grid of the mixer. Good: all very conventional. The two variable 'C's are not ganged: debatably unconventional.

If you wish to replace the R.F. amp. valve with a triode cascode pair, I would choose the series cascode connection, as opposed to the shunt cascode (former is easier to implement). Then the anode of the 'top' triode can be connected to the primary of the original anode load transformer as before. The secondary winding is isolated from the HT rail: no need for HT isolating capacitors.
And all of that is effectively what the 'front end' of my modified Eddystone 888A receiver looks like - except for the ganging of the variable 'C's, the application of A.G.C. and an R.F. gain control. The valve is an ECC85, incidentally.

You're making good progress: perhaps it's time for an updated drawing!
Keep at it,
Al.

[pwdrive]

Thanks for the comments Al, just to verify that the RF grid tuning and the mixer grid tuning are ganged (did that the other day) the two coils that the ganged capacitor tunes (RF grid & Mixer grid) are not identical but they were all I had for test purposes, I managed to lash up a slug and trimmer on one of them in order to get them to track, they track quite well. The tunable regen detector capacitor however is not ganged to anything and is not designed to be. The proceedure at present is to set my sig. gen. to a frequency 3 Mhz above the lowest signal frequency I wish to tune (10 Mhz for 7 Mhz) peak up the ganged RF/Mixer capacitor for max noise, there will be two peaks Fo plus signal & Fo minus signal so therefore the Fo minus peak is selected, the regen det tunes from 3.0 Mhz down to 2.5 Mhz therefore by tuning the regen det from 3.0 down to 2.5 Mhz the signal frequencies are tuned from 7.0 Mhz up to 7.5 Mhz the apparent frequency reversal being due to subtractive product of the mixing process, if the sig. gen was set to the lower difference frequency of 4.5 Mhz then by tuning the regen det from 3.0 Mhz down to 2.5 Mhz the signal frequecies of 7.5 Mhz down to 7.0 Mhz are being tuned due to the addative product of the mixing process, moving from the 7 Mhz amature band to the 14 Mhz amature band is quick and easy.
I know that Al and others will already know how all this mixing stuff works, I have wrote some of the above for the benifit of any newcomers to the hobby, working out the mixing process with a fixed IF freqency is fairly straight forward but some times when using a tunable IF (which is what the tunable regen det is) it sometimes can get a bit confusing.
As for the next phase I have not yet decided, I would still like to try the harmonic generator idea for the local osc. I have been looking at the RA17 harmonic generator circuit again I am pretty sure that the 0 to 32 Mhz LPF between the harmonic generator and the mixer it feeds does not have any bearing on the harmonic generation or selection of the harmonics itself only insofar as to suppress harmonics above 32 Mhz which of course could play havoc with the 1st VFO mixing process in that receiver, if any one knows any different could you please let me know.
In order to mess about with this Harmonic generator idea I will have to order up some stuff so it might be a while before I can try it.
Another word of note, I have tried tunable IF's from around 10 mhz down to 2 mhz the best one so far, taking into account the consistancy of regeneration across any tuned IF coupled with virtually zero pulling from non to full regeneration, direct signal breakthrough etc seems to be a tunable IF of 2.5 to 3.0 Mhz so that means tuning 500 Khz chunks of the spectrum at a time, getting it to tune 1 Mhz chunks is possible but some compromises have to be made as is always the case with these things, I have not discounted 1 Mhz in the final design it all depends if I can get the harmonic generator off the ground as setting the band set in 1 Mhz increments would be easier than 500 Khz so far as band setting and dial/scale calibration is concerned.
If I have time I will post yet another updated circuit showing things as they are at the moment, no doubt there will be many more revisions! The final design is more likely to employ B9A/B7G valves rather than Octals so various valve voltage parameters wiil no ndoubt have to be changed but this is not a problem in itself.

Lawrence.

[pwdrive]

Latest circuit diagram as promised.

Lawrence.

[Skywave]

Thanks Lawrence. That cct. now looks like something that I would expect to work - and work well: performance being largely dictated by the Q of the tuned ccts. and the general level and nature of the mechanical assembly - which, at the moment, is not yet complete of course. So the next stage is to get the local osc. section designed I suppose. It'll be interesting to see how that bit develops. Presumably, having got the freq. ranges correct, frequency stability will be the next most important factor.
I'll be watching . . . .

Al.

[pwdrive]

Thanks Lawrence. That cct. now looks like something that I would expect to work - and work well: performance being largely dictated by the Q of the tuned ccts. and the general level and nature of the mechanical assembly - which, at the moment, is not yet complete of course. So the next stage is to get the local osc. section designed I suppose. It'll be interesting to see how that bit develops. Presumably, having got the freq. ranges correct, frequency stability will be the next most important factor.
I'll be watching . . . .

Al.

Thanks Al, yes, harmonic generator to try next, if I had more room on the table I would take the cover of the B40D and tap in to the Xtal marker as the one employed in it seems to give plenty of output within the receiver itself, trouble is in order to try a pick out harminics selectively I need another tuning capacitor and that's one thing I havn't got, time to do some Ebaying I think.
If and when I start the harmonic generator idea I will be sure to let the forum know as to wether it stands a chance of being a go or a no go.

Lawrence

[Skywave]

O.K. - inspire me! Let's have a brief spec. of the H.G. requirement. I'm sure some of us here will come up with a few ideas. Now there's a challenge for the collective membership!
Whoops! Gotta go: just had a shout from the Domestic Engineering Dep't.: :ready for launch:

Al.

[pwdrive]

O.K. - inspire me! Let's have a brief spec. of the H.G. requirement. I'm sure some of us here will come up with a few ideas. Now there's a challenge for the collective membership!
Whoops! Gotta go: just had a shout from the Domestic Engineering Dep't.: :ready for launch:

Al.

Yes that would be good, talking of which I have just wound a slugged coil with a fixed capacitor across it, I have got it to resonate at 17 Mhz, this happens to be the osc frequency required so that the tunable regen det covers the 20 meter amature band, over this coil I wound anothe 3 turns for a coupling winding,(primary) I connected the sig. gen. to this primary and set the sig. gens frequency to 2.83 Mhz the 6th harmonic of which is very close to the 17 Mhz normally required, I peaked the secondary to 17 Mhz and tried tuning the regen det and lo and behold it worked, amature signals a bit weak (to be expected) but it basically works, I then fiddled around with another coil and capacitor and got it to work on the 4th and 2nd harmonic, obviousely the transmissions came though stronger on these lower harmonics the strongest reception was by using the 2nd harmonic.
Tried for odd order harmonics but my sig. gen. does not do them (to be expected)

My intial thoughts on a harmonic generator is on the first stage ie: the Xtal osc it will need to be heavily into Class C I think....maybe a Tri-tet? I am supecting though that a seperate harmonic amp will be required in order to obtain enough mixer injection voltage from it.
I will have to get some bits and have a go.

Lawrence.

[pwdrive]

I forgot the basic spec. not quite sure yet wether to go for 500 Khz or 1 mhz fundamental Xtal osc..... let's say 1 Mhz to make it easier, so basicaly a 1 Mhz Xtal osc but must be rich in harmonics followed by (if needed) a harmonic amplifier stage the harmonics need to be selected by no more than 3 switched resonant circuits using a common tuning capacitor, with 3 switched coils I can wind the three to cover most of the shortwave spectrum. The selected (tuned) harmonic output needs to be 200 millivolts minimum. Any Ideas? I have the details for the Racal RA17 harmonic generator but I would be good for me/you to come up with a suitable design....valves only, max two envelopes the ideal goal would be one envelope.We are talking of harmonics in 1 Mhz increments so both odd and even will be required.

Lawrence.