Questions on 160/80 metre AM transmitters.

[Ian - G4JQT]

Sorry I'm late to the party, but here's an article I wrote for PW a few years ago. Amongst other things it describes some simple AT5 mods and corrects an error in the cct of the manual.

I tried to upload the actual pdf of the article as published, but apparently it's too big.

Regards,

Ian

[Bazz4CQJ]

Originally Posted by G6Tanuki

Pragmatically, a sensibly-designed free-running VFO shouldn't suffer enough drift to be a problem when used with 1.8/3.5MHz AM.

Keep the heaters on all the time, keep the VFO running all the time, and use a diode-switched capacitor to shove the VFO up a few tens/hundreds of KHz when on receive.

FET and transistor VFOs - in the absence of the always-on-running mitigations - have low thermal-inertia and so just swap slow frequency-drift for fast-drift.

Klaas Spaargaren PA0KSB's "Huff and Puff" stabiliser from the 1970s is an interesting fix for the problem; it will happily lock to a VFO-tuned frequency and hold it to within 50Hz.

http://www.hanssummers.com/huffpuff.html

Last point first, I was talking to someone a while ago who absolutely did not want to add a FET to his vintage receiver even though it could have been the way to have given it a feature he wanted. If someone is keen to stay true to a vintage chassis or a design and won't use a single FET, the idea that Huff and Puff will be acceptable seems unlikley.

As to the other comments about making a good valve VFO; those all seem very sensible, but it may well be possible that the use of a battery valve could just make it that bit better. Seems unlikely that it would make it worse, but I've never seen it tried.

[CambridgeWorks]

I remember that the Marconi VHF/ UHF signal generator TF1064B/5M used a DCC90 twin triode oscillator then other typical 6.3v heater valves. Stability was reasonable for the frequency coverage.
Rob

[Bazz4CQJ]

And the DCC90 is a nominal 1.4V filament I think?

B

[Cruisin Marine]

Re: post #81, I guess the modification by

"Pat, G3IKR suggested a modification to me which can slightly increase output on
80m. The L3/C9 combination is only roughly tuned and is not adjustable as it
stands. Pat suggested replacing C9 with a 60pF trimmer which is adjusted for
maximum output on 80m. I peaked mine in the phone section and increased RF
output by about 1.5 watts."

would put in to doubt that L3/C9 is a series acceptor tuned cct. to eradicate 1.9 MHz signal when switched to 80 Metre? or maybe not?
Perhaps peaking the rejection/leakage from 1.9 MHz increases the o/p on 80metres as well as doing its own job too by cleaning up the drive energy to be just concentrated on 80m.

[Radio Wrangler]

There is the mystery of what capacitance does L2 resonate with, or is it just an RFC on the band where L3 is switched out? The capacitor to the PA grid ends in a high impedance. I'd need to mess around with some simulation to be sure, but I think the driver anode thinks it's aperiodic with RFC bias. But when the capacitor and L3 gets switched in, several things change.

David

[G6Tanuki]

Be aware that the little B7g battery valves like the
1T4 and DF91/96 can be microphonic. You may be swapping thermal stability for mechanical stability. I had a problem with microphony in my R209 receiver whose local oscillator was picking up the vibrator's vibration coupled through the chassis, and had to try about half a dozen 1T4 to find the least microphonic of the bunch.

[G3PIJpeter]

Originally Posted by G6Tanuki

As to the other comments about making a good valve VFO ... it may well be possible that the use of a battery valve could just make it that bit better. Seems unlikely that it would make it worse, but I've never seen it tried.

Have a look at the 'glowbug' article at http://g4oep.epizy.com/xfy/xfy.htm
Using a wire-ended miniature valve might be the way forward. The results claimed by this article are certainly impressive.

Peter G3PIJ (hiding from doing the sprouts)

[Bazz4CQJ]

Wow, PETER, thank you for brining that article to the party!! That is interesting.
Do you know the author personally - I guess he's not on VRRR?. Of course, we have at least one member north of the boarder who's a great devotee of wire-ended valves. It may not be well known but there was an Acorn valve with a 2V filament meant for use in VHF "walkie-talkies" and I bet that would be another very rugged little device.

@Tanuki - yes indeed, the B7G are quite microphonic. In my early days I built a TRF with two DF91's, and your adage that some amplifiers want to be oscillators could be extended to saying that and some want to be microphones.

B

[G3PIJpeter]

Originally Posted by Bazz4CQJ

Do you know the author personally - I guess he's not on VRRR?. B

No, I don't know G4OEP personally - I chanced upon his website a few years ago and have used various ideas from it ever since. Many of his glowbug designs have a certain 'steampunk' look but the underlying circuitry always seems to include a fresh approach to the old problems.

I've been given a new (40 watt) soldering iron for Christmas, so who knows where that will lead . . .

Peter G3PIJ

[Bazz4CQJ]

Looking at his website, G4EOP seems to be a man who packs a lot more in to a 24 hour day than I do.

I was under the impression that Russian wire-ended valves were cheap as chips on eBay, but somebody seems to have bought them all .

B

[AdrianH]

Back to the Codar AT5 circuit, another question for the more knowledgable.

The PA Choke simply annotated as RFC1 does not seem to have a value, considering I have only got the small resistor type chokes of values between 0.56, 1, 2.2, 3.3, 4 and 4.7mH I was considering doing two values in series say a 1 and a 3.3 mH, giving 4.3 mH, the chokes are 1 watt but the resistance of them is low enough that I should be able to pass 230 mA before I hit the limit for the 3.3 mH choke. Can anyone see a problem with using these in this way?

I have done an APC for the anode connection, and I just thought that having two different value chokes may be better then one choke on it's own.

Adrian

[Cruisin Marine]

You will only need about 2.2 mH value needed throughout for all RFC's.

That will have a high enough reactance at both 160 and 80 metres to act as a choke or stopper and will produce a decent load for the valves.

One way to tell if an RFC is not big enough is when they get hot, if they stay cool, then they are big enough in value.

230 mA should be fine too, even when the PA is off tune it should be enough.
You can only hit trouble if the chokes have any self resonances, that will be obvious if you have a speccy analyser or if you get some strange resonances.

I would doubt you will have that problem though.
Merry Xmas

[G6Tanuki]

There's much to be said in favour of putting the choke on the downstream side of the pi-network - see below.

RFC8 being the 'anode' choke.

This decouples any silly and unpredictable choke self-resonances from the PA valve and means the choke is working at a low RF impedance.

This approach was used in a number of US WWII-era military transmitters, but seems to never have taken hold this side of the pond.

[AdrianH]

OK thanks for the two comments, taking the second first, I thought the idea of the end choke was to provide a DC path to ground should the Anode blocking cap fail providing a DC path enough to blow a HT fuse, it could also help reduce static build up on long wires?

The thoughts on multiple chokes and reactance's was my looking at the old style choke, (which I did not get) was that they looked like small coils/chokes in series rather than one lumped coil. A typical 2.5mH being of 5 sections each of 0.5mH. It would have a potential lower stray capacitance and therefore have less tendency to be resonant at the frequencies it was likely to block then one coil of the same total value.

These small lumped coils I am not sure off, they probably have low stray capacitance as I think they are ferrite cored, but was not sure.

Adrian

[Cruisin Marine]

There's much to be said in favour of putting the choke on the downstream side of the pi-network - see below.

That is enlightening G6Tanuki.

I had heard of this, but never seen a tried cct. before.
It may be worth pointing out that in that cct. RFC 6 and 7 would be TVI parasitic stoppers consisting of say 6 turns wrapped around a 56-100 ohm resistors fed from two anode's. RFC 8 Is the HT feed RFC and RFC 5 is a protection RFC, it is there to provide a path to earth in case the cap C87 goes short cct. It will blow the HT+ fuse if that happens.
The other advantage of that cct. is that RFC 8 could be a lower value as it is placed at a lower impedance point.

This leads on to Adrian's comment posted about the same time as mine.

"I thought the idea of the second was to provide a DC path to ground should the Anode blocking cap should fail and provide a DC path enough to blow a HT fuse, it could also help reduce static build up on long wires?"

Absolutely spot on mate (I didn't realise this was what you were getting at, sorry).

I used to use smaller lumped 2.5 mH RFC's way back in the 70's, these were single winds (probably made from Litz wire looking back, I used to get them from H.L. Smith in Edgware Road), I never had any trouble with them at all. I only used them up to say a 6BW6 or 5763 level type valve. On an 807 or so a 'normal' 5 layer 2.5mH would be used.

[AdrianH]

OK I have done a compromise for myself, two 1 mH and 1 .5 mH in series with the 1 mH to the anode. I have done a parasitic stopper as well.

I had a thread in the homebrew section I will put up a few pictures of progress in there. only working towards Top Band CW at present.

Adrian

[AdrianH]

Re reading Ian's post #81 and looking at the diagram of the modified circuit, especially around where the 100mA meter has been moved to the anode side of the output valve, I think I can see an issue with the Key 'click' filter.

To my mind it would have worked with the resistance and inductance of the meter in the cathode line, but without the meter I can not see any shaping with KEY DOWN. I can see it with KEY UP, but if the key has a very low resistance I can not see anything to shape the RF/Current rise time. To me some resistance needs to be between the Cathode pin and the Key socket?

Wonder if anyone else is considering building for Top band and 80?

Adrian

[Graham G3ZVT]

I didn't use my homebrew AT5 clone very often on 80m, but on one occasion when I did, my CW QSO was interrupted by an S9 station purporting to be "GPO monitoring station Gxx" ordering me to go QRT.

I imagine it was because it's not just the frequency that gets doubled on 80m, so does the chirp!

I fully expected a visit, or at least a letter, but neither came.

It's been over 50 years now, do you think it's safe to retrieve the TX from under the bed

[Aub]

Originally Posted by rambo1152

I didn't use my homebrew AT5 clone very often on 80m, but on one occasion when I did, my CW QSO was interrupted by an S9 station purporting to be "GPO monitoring station Gxx" ordering me to go QRT.

I imagine it was because it's not just the frequency that gets doubled on 80m, so does the chirp!

I fully expected a visit, or at least a letter, but neither came.

It's been over 50 years now, do you think it's safe to retrieve the TX from under the bed

Blimey, that was scary. I wonder if it was real, or someone trying to wind you up.

73

Aub, G4KQL