AJAX "Leader" marine TX/RX info

[Keith]

I have recently aquired an example the above which I would like to put into operational use on 80 metres. The set has unfortunately been heavily "got at" by a previous owner with various looms cut and attempted modifications to the TX. The output valves are missing (TT21s or TT22s I think) and there seem to be rather worryingly empty spaces on the chassis. The only other identification on the set apart form the name "Leader" is KH50-12V. I wonder if anyone out there has any information on this equipment? So far all I have managed to obtain is a partial circuit of the receiver.

[PJC58-Hythe]

Keith,

Is this the set you are describing?

http://www.g0kvz.talktalk.net/vars/misc/marine.htm

If so contacting him may be a good starting point.

Peter

[Keith]

Hi Peter,

Thanks for the reply. Yes, that is the set in question. I understand that Mike (G4BQF) has passed on some information to another amateur with whom I am in contact. Unfortunately, he has mislaid his CD containing a scanned copy. I shall contact Mike directly as you suggest.

Regards,

Keith

[Keith]

The good news is that I have now obtained a circuit of the transmitter (albeit a slight variant). However the bad news is that it confirms that a section of the TX has been removed by a previous owner, namely the output stage of the modulator (2 transistors and transformer). The transistors are not an issue - I'm sure good old 2N3055s could be pressed into service. The problem is how to replicate a mod transformer to match these to a pair of TT22s or similar?

[PJC58-Hythe]

Shouldn't be that difficult , just need a transformer that can swing 2 x the plate voltage (peak) at the appropriate power. I can't see from the thumbnail what the plate voltage is, but you may well get away with a standard mains toroid in reverse as you're not looking for large AF bandwidth.

Can't see from the thumbnail what the low voltage supply is to the 2 x power transistors, but I'm guessing it's not more than 24V. You could always replace the transistors connect a suitable mains toroid with approx voltage / power rating and see what you get.....?

I presume you have seen the other thread on the forum..

https://www.vintage-radio.net/forum/...ad.php?t=34746

Peter

[Keith]

Hi Peter,

Thanks for the response. Yes, I did wonder about using a mains transformer in reverse. As you say, I'm not looking for hi-fi! The design plate voltage is 700V but could be lower. A centre-tapped 6.3V winding as primary should give me about 500V across the mains winding. Or or if i can find something with , say, a 450-0-450 secondary at suitable curent rating that I could series up to give me a bit more swing. Worth a try.

Regards,
Keith

[Radio Wrangler]

Peter's right, it ought to be quite easy and you have a couple of other factors on your side.

TT21 pair is going to be 50W anode dissipation each, and the 813 at 125W disiaption is the classic 100W carrier AM bottle, so let's guess at 80W carrier and 80*(1.8)^2 = 260W PEP at 80% modulation. So the transistor amplifier needs to be able to do 60 to 80 watts.

We all can't read the voltages on that small thumb nail diagram. So you need to look at the PA HT voltage and reckon that the PA supply after the mod transformer is going to swing down to 20% and up to 180% of its static value. By the wonders of transformer action, you'll probably want the collectors of the transistors in the modulator to be doing something similar, swinging between 20 and 80% of their DC power supply. The ratios of these two give you your turns ratio... essentially the ratio of the DC feed into the transformer to go to the transistors to the DC feed into the transformer to go to the TT21s.

Simples!

The next nice bit is that for comms grade speech, you don't need to go down to 50Hz, in fact a 300Hz cutoff is usual and mandated in some applications. This means that your transformer needs only one sixth as much magnetising inductance as it needs for that voltage at 50Hz, or it means you can run windings to six times their labelled 50Hz voltage

So a mains transformer with a 240v primary winding (2 * 240 * root(2) = 680v swing (peak-to peak) could be run at 600Hz to 4kV swing. Inevitably, this is too good to be true. The insulation cannae handle it, Captain. Also the lamination thickness you can get away with for 50Hz is a bit cruder than what you'd like for decent efficiency at audio frequencies.

Now for the big big problem:

The DC components to the two push-pull modulation transistors mostly cancel (which is good) But those big transmitting bottles take current in only one direction, and it hits some quite big peaks. Unless the modulation transformer is carefully designed, it is going to saturate. The usual ploy is to gap the core to bring down the flux. It also brings down the inductance, quite dramatically.

That 6 times easier factor for magnetising inductance that was won from opting for a 300Hz cut-off just got spent, and then some more.

Take a look at classic mod transformers like the Woden UM1, UM3 etc and they are big brutes when you consider the audio power to be put through them... ESPECIALLY when you consider they don't have to go down to low audio frequencies.

Toroids are the ultimate low-gap cores. the exact opposite of what you're forced into.

You may be better thinking of the modulation transformer as a choke in the HT feed to your PA which just happens to a handy added winding for you to pour some audio power into. But then you also get the ampere turns of one of the modulation transistors adding its magnetomotive force to the problem on the peaks.

No doubt Ed can put this a lot better than I have, but a modulation transformer has to be a specialised beast. It's easy to calculate and design, but there isn't an easy way to make one out of anything common.

I went through all this in designing a new AM transmitter... it's just passed the approvals testing for fitment to certified aircraft, and manufacture is revving up. Ten years ago if you'd told me I'd be designing AM transmitters in the 21st century you could have pushed me over with a feather. And no, it doesn't have a modulation transformer in it. There isn't room, or weight budget.

David

[PJC58-Hythe]

David is quite right!

I missed the fact the transformer will saturate with the standing DC current of the PA (note to self - read the circuit diagram properly!). In which case if you could use the modified Heising method and put a choke in place of the mod transformer in the PA then couple the output of the toroid via a capacitor. Of course you would need a large choke say 20-30H capable of handling the current drawn by the 2 x TT21's. But unless you can find a suitable mod transformer this may well be a cheap solution.

Peter

[Aub]

Hi Keith,

I've built several AM transmitters, over the years, using mains transformers for modulation. The transmitter I currently use on 80M AM uses a custom made mod tx. from Sowter but, while I was testing it, I used a Radio Spares "250v Economy Mains Transformer" and that was OK. Of course, you'd need a different ratio for your transistor pair. The PA current runs directly through the tranny secondary and I've never had a saturation problem, although I'm only running about 70w input ( the modulator is 2 x EL34s producing around 30w )

As others have said, its great fun!

73s Keith

Aub, G4KQL

[AlanC]

I have a scrap Pye F30AM transmitter chassis sculling around somewhere. It has a modulator using 2N3055's to modulate a QQV06-40A PA, running at 25W in normal PMR service, but capable of somewhat more than this. I know that's a lot lower power than the Ajax Tx, but the transformer is a big brute as I remember. If you'd be interested in it I'll see if I can look it out this weekend and take some pics.
Cheers, Alan
G0NFY

[Keith]

Thank you, gentleman, for your posts, especially your detailed suggestions, David.

However, it seems I may not have a problem after all. Having seen some pictures of my set, another owner tells me that the section that is missing is not the modulator but the power inverter used to obtain the PA voltages from the 12V battery supply. I need to check in detail but it looks like the all the modulator components are still present. I'm not so concerned about the inverter as I had a mind to power the rig from a separate mains PSU in any case.

Very interesting comments nonetheless. I still have the first valve TX that I built when I first got my licence. It uses a 6BW6 modulated by an EL84. Being a poor student, I also used a mains transformer as the modulator. The PA and modulator valves drew about the same current (~50mA) and I always got good reports of the mod. Re-examining the TX recently, I had fortuitously arranged the primary and secondary in anti-phase, thus largely cancelling the standing flux and avoiding core saturation. Pure luck, I'm sure, as my knowledge of such things was pretty minimal at the time. For higher power I agree that the Heising system is a good solution - there's a very good VMARS design using this to modulate a pair of TT21s from a MOSFET amplifier.

Interested to hear that AM designs are still called for in aviation, David. I wonder if it uses low level mod followed by linear amplification or possibly series modulation of a class E PA.

Best regards,

Keith

[G6Tanuki]

If you do want to use a mains-transformer for modulation, don't use a toroid.

Rather, get a non-varnish-impregnated traditional-type transformer with the required turns-ratio. Then split it.

Take out the E- and I-laminations from the core, then reassemble them so they're not interleaved. Stack all the "E"s together and fit them to the transformer-bobbin; then put the "I"s against the ends of the "E"s - you may want to introduce a very thin piece of waxed-paper between the Es and the Is to control the gap.

This will help control the maximum flux-level and stop the transformer being magnetically-saturated by the DC component of the PA anode-current.

--G6Tanuku

[G6Tanuki]

I built quite a few similar transmitters - some using an 807 PA/807 Modulator in the same way you describe. The trick was to use as a modulation-transformer a traditional type push-pull output-transformer specified for the same kind of valves/power-level you were using. The anode-impedances of a Class-c PA and a class-A modulator being similar, and the current for the modulator and the current for the PA being in opposite directions, they provided no standing magnetisation to the transformer core.

Back then my traditional design was ECL86 as mic-amp/driver, 807, 6L6 or EL34 as class-A modulator; on the RF side again an ECL86 as xtal-osc/driver and 807 PA.

--G6Tanuki.

Originally Posted by Keith

Very interesting comments nonetheless. I still have the first valve TX that I built when I first got my licence. It uses a 6BW6 modulated by an EL84. Being a poor student, I also used a mains transformer as the modulator. The PA and modulator valves drew about the same current (~50mA) and I always got good reports of the mod. Re-examining the TX recently, I had fortuitously arranged the primary and secondary in anti-phase, thus largely cancelling the standing flux and avoiding core saturation.
Keith

[Radio Wrangler]

Hi Keith,

Ships were reasonably well served with MF/HF radio for a long time and only later got into VHF, but for planes, the size of VHF antennae was a bug attraction and aviation went for VHF mostly quite early on, and AM was the current technology. There are so many planes and so many ground stations, it can't really change/

I'm doing high level modulation of a MOS PA. The efficiency figures it produces are astounding and needed to be carefully checked. It is all under closed-loop control to get low distortion and to meet a fearsome spectrum template.

David

[Keith]

Hi akk,

Thanks for your additional comments and advice (and also your kind offer, Alan). Clearly mains tranformers can be used in high level modulators provided that precautions against saturation are taken.

I haven't made any further progress with Ajax TX as yet but I have got the receiver working to some extent. A previous owner has disconnected most of the power rail wiring and it's taking a while to sort it out. Also a couple of relays have been removed so I shall have to reconfigure the RX/TX changeover.

Thanks for the info on the aircraft TX, David. I've often wondered why AM was retained for airborne comms - I thought possibly it might be to avoid the "capture effect" masking a weak signal but, as you say, the cost of changing a worldwide infrastructure would be massive. I'm currently playing around with a pulse width modulated class E MOSFET TX designed by an amateur in North Wales. As you say the efficiencies obtainable are very impressive.

Keith

[Keith]

Well, having said that the receiver was working "to some extent" I can now announce that further progress has been made in this area. Initially the audio sounded very unstable with a sort of "ringing" effect. This was appearing at the output of first audio stage. The receiver includes a narrow peak filter for listening to CW signals. This is switched in using a relay when the BFO is selected. It turned out that a systoflex-sleeved wire had been routed under the relay and was preventing it from returning to it's de-energised position. Consequently the peak filter was permanently in circuit. It may have been like this since original manufacture - possibly the sleeving hardened over time and caused the problem. Anyway, re-routing the wire solved the problem.

I'm not entirely happy that the AGC is functiong adequately. Some distortion is noticeable on strong MW signals when a decent aerial is connected. This can be improved by backing of the manual RF gain control. I was surprised to discover that the (amplified) AGC appears to bias the controlled IF and RF stages harder on for strong signals. At first I thought this was a fault but looking at the circuit it appears to designed in this way. Although it does reduce the gain, I haven't seen this configuration before. I wonder if anyone has any thoughts on the arrangement.

[G6Tanuki]

It's called "Forward AGC" and was standard practice in transistor TV receivers in the 1970s/1980s. There are transistors specifically designed for operation in this mode.

It has the advantage that the transistor's signal-handling capacity *increases* as the AGC action takes hold - unlike the traditional AGC where reducing the gain also pushed the transistor into the non-linear part of its operating-curve (with all the risk of generating more intermodulation products - just what you don't want when faced with the strong signals the AGC is trying to tame!).

Originally Posted by Keith

I was surprised to discover that the (amplified) AGC appears to bias the controlled IF and RF stages harder on for strong signals. At first I thought this was a fault but looking at the circuit it appears to designed in this way. Although it does reduce the gain, I haven't seen this configuration before. I wonder if anyone has any thoughts on the arrangement.

[Keith]

Thanks for that, G6Tanuki. It makes sense, of course, but not something I'd come across. Most of my experience with TV IFs involved EF80s!