AM Micro Transmitter - Page 3 - UK Vintage Radio Repair and Restoration Discussion Forum

MichaelR · 25th Feb 2008, 12:20 am

Hi Dom,

No problem, fascinated by the thread , thank you.

Regards
Mike

kalee20 · 25th Feb 2008, 2:02 pm

I've been thinking about the automatic compensation for changing load. Ultimately, the aerial or feeded will provide a constant load, so an automatic circuit won't be needed.

I always regard it as good practice to get everything working well without such tricks as negative feedback, automatic level control, etc. Then, they can be added at the end as 'the icing on the cake.' You can optimise your modulator by replacing the PA with a dummy-load resistor from cathode-follower down to deck, and then play with the PCL82 triode cathode resistor for best output swing. With such a low HT voltage, you may need to raise the anode load. But don't make it too high, or else it won't be passing enough current to charge and discharge the various capacitances, and you'll get slew-rate limited distortion at high audio frequencies (a good reason for having HF roll-off as next paragraph).

If you are getting the AF from a CD player, I can see that some sort of volume compression may be needed if the quiet passages are not to be lost in the noise. Much as I dislike the idea, something like a low-pass filter with 10kHz roll-off, a compressor such as a vari-mu valve with bias derived from rectified AF (not RF), followed by an instantaneous peak clipper for if all else fails.

Valveholders are in the post!

dominicbeesley · 25th Feb 2008, 2:29 pm

Thanks Kalee, I'll be pulling the old setup apart tonight and tidy up, then I'll concentrate on getting the AF section spot on (or as near as I can get before adding back in the RF sections, when I get the valve holders).

Hopefully with about 7 times the voltage to play with I should get a bit plenty of power - I just hope I don't radiate too much!

Cheers

Dom

PJL · 26th Feb 2008, 2:34 pm

Fantastic piece of work and great reading. I don't know much about the RF side so have a couple of questions:

As it is very easy to do with your circuit, did you try screen grid modulation?

The feedback must be loading the tank coil. I wonder to what extent this effects the radiated power? The clamp diode might increase the feedback but I am wondering if the increased loading needs to be considered?

Managing the harmonics seems a nightmare, the drive voltage, aerial loading and non-linearity must all come into play. With near 100% plate modulation surely harmonics must be all over the place?

dominicbeesley · 27th Feb 2008, 3:29 pm

I tried screen modulation earlier but this didn't give much modulation, though the modulation it did give sounded ok.

Hopefully its not loading it too much, the input is through a 100k, the suck it and see test was to see if the trace altered noticeably on the scope. With the output of the feedback grounded and connecting it in and out of the RF out made no noticeable difference so I passed it as OK. I'll try it on a separate winding when I get back to the RF side of things, this should isolate it somewhat and allow me to simplify it a bit (get rid of clamping and use full wave rectification instead?)

The harmonics are present but seem to die off pretty quickly. I'll put in a low pass filter before any antenna on the next iteration and hopefully that will improve matters a little.

Kalee20 very kindly sent me some more valve bases and a few bits and bobs today so will be back on it over the next few days and will report back with progress

Cheers

Dom

kalee20 · 27th Feb 2008, 6:27 pm

I tried screen modulation many years ago, and although it worked, it wasn't particularly impressive. Linearity was a big issue if I remember OK.

The combined anode and screen modulation looks much better to me, everything is proportional so as long as there's enough drive, the anode current pulses should vary linearly with the modulating voltage. It takes more power from the modulator (anode and screen current rather than just screen), but is worth it.

The idea of a pi filter on the output to me is the way forward, to reduce output harmonics. Reminds me of work I once did with resonant switch-mode power supplies. I've got some ideas on how to calculate the values, but they're not ready for sharing yet! And I don't think that changing the output circuit will invalidate the rest of the good work that Dom's doing anyway.

The loading imposed by the feedback detector should indeed be negligible given the high values of load resistor. (Although wouldn't a 6AL5 double diode be nicer than those nasty PN junctions!!)

mickjjo · 27th Feb 2008, 6:45 pm

There are some interesting circuits on the 'net for AM micro power transmitters like this one also using diodes in the feedback circuit:-

http://www.dogstar.dantimax.dk/tubestuf/amtx-3.htm

This one uses a 6AL5/EB91 in the feedback circuit:-

http://www.dogstar.dantimax.dk/tubestuf/amtx-4.htm

Interesting output stage too.

Regards, Mick.

Ian - G4JQT · 27th Feb 2008, 6:53 pm

The above valve circuits seem mighty complex for just a few tens of milliwatts...

Ian

MichaelR · 27th Feb 2008, 8:01 pm

maybe but there is a lot of education going on here Ian, personaly I am loving it

mike

kalee20 · 27th Feb 2008, 8:13 pm

The web pages that Mick has given us links to, generally use a frequency-changer valve as a modulator. After all, the valve is a multiplier which is exactly what an AM modulator needs to be. Then, the modulated RF is fed to what looks like a linear output amplifier (working in Class A).

It's a bit complicated, but does work! Dom's work on the other hand is heading more towards a very low power version of a full-scale transmitter.

Either way is fascinating.

dominicbeesley · 27th Feb 2008, 11:19 pm

Quote:

Originally Posted by kalee20

The web pages that Mick has given us links to, generally use a frequency-changer valve as a modulator. After all, the valve is a multiplier which is exactly what an AM modulator needs to be. Then, the modulated RF is fed to what looks like a linear output amplifier (working in Class A).

It's a bit complicated, but does work! Dom's work on the other hand is heading more towards a very low power version of a full-scale transmitter.

Either way is fascinating.

Thanks for the circuits above guys. As kalee20 points out they use pentagrids for the modulation. I saw these a while back and originally tried getting something to work using an ECH42 to work, but this had a remote cut-off and sounded pretty awful when I tried it out (had to feed AF into g1). Indeed the circuit for the oscillator is cribbed direct from the Philips application note for the ECH42! I was quite surprised when it transferred without requiring any changes to the PCL82 triode.

This could have been my fault though or I might have been able to get round it somehow, however try as I might I couldn't get a decent level of modulation this way so gave up and investigated my junk-box favourites the PCL8* valves.

As Kalee20 points out above I got the idea for the second attempt by looking at some plans for commecial transmitters. And also an excellent book I found on Google books that I can now not find again

Just spent tonight winding a mains Tx with a 48V, for 4 PCL82s and a hopefully 100ish volts for the HT. Not tried it out yet....

PS: Kalee20 you have a PM.

Cheers

Dom

dominicbeesley · 4th Mar 2008, 12:20 pm

Hello,

I've been working away over the weekend to make a higher voltage version, running from a mains Tx and get the thing into a nice box.

This is proving to be quite difficult with lots of stability problems being introduced by running from an non-stableised supply. I've got the whole thing wired up with what should be a nice star ground etc.

I managed to get the AF amplifier and cathode follower working nicely into a dummy load and also into a speaker through a large HiFi Tx and sounding good. But when I connect up to the actual PA valve, set up just as before, it sounds absolutely terrible! I've not even tried the feedback loop yet, with what I suspect to be HF instability of some sort I suspect all I'd get is a nasty squeal!

The circuit is pretty much as before except that the first stage is decoupled via a 47k resistor and 1uF to ground, the other two AF stages have their own 0.1uF decouplers and the oscillator has a 2200uH choke and a decoupler added...

Any ideas where to attack first?

Its quite frustrating that something that worked great when it was sprawled randomly and messily on the bench then misbehaves when I've tried to lay it out nicely with short leads and proper earthing etc!

Oh well....

Dom

kalee20 · 4th Mar 2008, 2:25 pm

Hi Dom, good to see you back on the project! Did the transformer arrive OK?

Posting up the latest circuit diagram would probably help a lot. I suspect you've got excess AF gain now - I've always found layout is critical when you have lots of gain. Dunno why it should be OK with a dummy load...

Decoupling with 0.1uF does sound rather light. Hopefully, with the diagram up, it will all be clear!

Isn't it annoying when a side issue grabs all the attention and has to be dealt with before the main drive can be restarted??!!!

dominicbeesley · 7th Mar 2008, 12:56 am

Hello all,

I got a bit further this evening, after a rebuild. Found out that the reason it sounds so bad is because its overloading all the sets in the vacinity. It sounds fine at the end of the street! All this with no aerial to speak of (about 5' of wire), I can probably lose a bit of the power/voltage somewhere in the PA stage!

It now sound pretty nice from the Goblin Timespot in the attic but a bit ropey from any transistor set in the ground floor (overloading RF/IF stages?)

I'll hopefully now be able to make the PA valve sit somewhere above ground (DC operating point with a big cap to hold it there) and so get 100% modulation without over-driving the AF stages.

One thing I can't work out quite is why the two circuits in the attached pictures don't behave the same at RF. Am I missing something. I want to use the second (rightmost) as the load of the PA valve as I've already got a dual gang variable capacitor, one gang of which I'm using for the oscillator section.

Dom

paulsherwin · 7th Mar 2008, 1:00 am

Quote:

Originally Posted by dominicbeesley

Got a bit further this evening, after a rebuild. Found out that the reason it sounds so bad is because its overloading all the sets in the vacinity. It sounds fine at the end of the street! All this with no aerial to speak of, I can probably lose a bit of the power somewhere!

Yes, please concentrate on reducing the output Dominic, as otherwise the transmitter will violate forum rules and can't be discussed here.

It's very easy to radiate a lot of signal with a valve microtransmitter without meaning to do it. Even a one valve design based on a frequency changer (classic American 'phono oscillator') can produce enough output to cause problems.

Paul

Sean Williams · 7th Mar 2008, 8:37 am

Dom,

What HT voltage and current do you have on the PA valve?

My pantry transmitter has a similar coverage, and that uses an EF80 as the "PA" - has about 150v on the anode at around 4 or 5 ma

The problems you are experiencing now are quite normal - happens here all the time just part of the fun when homebrewing stuff!

dominicbeesley · 7th Mar 2008, 1:18 pm

Quote:

Originally Posted by paulsherwin

Yes, please concentrate on reducing the output Dominic, as otherwise the transmitter will violate forum rules and can't be discussed here.

It's very easy to radiate a lot of signal with a valve microtransmitter without meaning to do it. Even a one valve design based on a frequency changer (classic American 'phono oscillator') can produce enough output to cause problems.

Paul

Don't worry Paul, I've removed the aerial altogether for now and will wind a new PA load coil to step it down to a 75ohm impedence and run it round the house using the roll of coax I've just bought. I don't want to pollute AM as reception in the bottom of a deep valley is bad enough as it is.

I've no chance of competing with the crappy inverters used on some of the roof-top windmills round here though, they've effectively got rid of about half of MW and a couple of stretches of LW never mind SW.

Dom

dominicbeesley · 7th Mar 2008, 1:22 pm

Quote:

Originally Posted by Sean Williams

Dom,

What HT voltage and current do you have on the PA valve?

My pantry transmitter has a similar coverage, and that uses an EF80 as the "PA" - has about 150v on the anode at around 4 or 5 ma

The problems you are experiencing now are quite normal - happens here all the time just part of the fun when homebrewing stuff!

No even that high the "HT" is 115V so the quiescent (no AF signal) voltage of the PA valve is about 65ish, not got round to measuring the current yet.

I just really want to get the PA tuned using the other half of the grounded variable capacitor and then sort out the aerial properly. I still don't understand why the two circuits in the earlier post produce different results, I would have thught either 0V or +HT would "appear" the same to RF?

Dom

Aub · 7th Mar 2008, 1:48 pm

Quote:

Originally Posted by dominicbeesley

I've no chance of competing with the crappy inverters used on some of the roof-top windmills round here though,

Dom

Hi Dom,
I'm intrigued by this. I know it's off topic, but are you saying that wind generated electricity is widespread where you are, or do you mean something else?

regards

Aub

dominicbeesley · 7th Mar 2008, 2:41 pm

There are a few higher up on the hill with blades about 3' long. The green ones where the blades point away from the wind (don't know the make) seem to create a real racket on the AM bands, I'm guessing its the inverters, though it could be the actual generators themselves. A bit like putting a radio next to a computer but lasts a few hundred yards blocking out R4 on LW.

The big "proper" turbines, there are a lot round here in various groupings, are well behaved.

25th Feb 2008, 12:20 am	#41
MichaelR Retired Dormant Member Join Date: Mar 2006 Location: Sheffield, South Yorkshire, UK. Posts: 1,587	Re: AM Micro Transmitter Hi Dom, No problem, fascinated by the thread , thank you. Regards Mike

25th Feb 2008, 2:02 pm	#42
kalee20 Dekatron Join Date: Feb 2007 Location: Lynton, N. Devon, UK. Posts: 7,081	Re: AM Micro Transmitter I've been thinking about the automatic compensation for changing load. Ultimately, the aerial or feeded will provide a constant load, so an automatic circuit won't be needed. I always regard it as good practice to get everything working well without such tricks as negative feedback, automatic level control, etc. Then, they can be added at the end as 'the icing on the cake.' You can optimise your modulator by replacing the PA with a dummy-load resistor from cathode-follower down to deck, and then play with the PCL82 triode cathode resistor for best output swing. With such a low HT voltage, you may need to raise the anode load. But don't make it too high, or else it won't be passing enough current to charge and discharge the various capacitances, and you'll get slew-rate limited distortion at high audio frequencies (a good reason for having HF roll-off as next paragraph). If you are getting the AF from a CD player, I can see that some sort of volume compression may be needed if the quiet passages are not to be lost in the noise. Much as I dislike the idea, something like a low-pass filter with 10kHz roll-off, a compressor such as a vari-mu valve with bias derived from rectified AF (not RF), followed by an instantaneous peak clipper for if all else fails. Valveholders are in the post!

25th Feb 2008, 2:29 pm	#43
dominicbeesley Octode Join Date: Nov 2004 Location: Hebden Bridge, West Yorkshire, UK. Posts: 1,885	Re: AM Micro Transmitter Thanks Kalee, I'll be pulling the old setup apart tonight and tidy up, then I'll concentrate on getting the AF section spot on (or as near as I can get before adding back in the RF sections, when I get the valve holders). Hopefully with about 7 times the voltage to play with I should get a bit plenty of power - I just hope I don't radiate too much! Cheers Dom

26th Feb 2008, 2:34 pm	#44
PJL Dekatron Join Date: Sep 2005 Location: Seaford, East Sussex, UK. Posts: 5,997	Re: AM Micro Transmitter Fantastic piece of work and great reading. I don't know much about the RF side so have a couple of questions: As it is very easy to do with your circuit, did you try screen grid modulation? The feedback must be loading the tank coil. I wonder to what extent this effects the radiated power? The clamp diode might increase the feedback but I am wondering if the increased loading needs to be considered? Managing the harmonics seems a nightmare, the drive voltage, aerial loading and non-linearity must all come into play. With near 100% plate modulation surely harmonics must be all over the place?

27th Feb 2008, 3:29 pm	#45
dominicbeesley Octode Join Date: Nov 2004 Location: Hebden Bridge, West Yorkshire, UK. Posts: 1,885	Re: AM Micro Transmitter I tried screen modulation earlier but this didn't give much modulation, though the modulation it did give sounded ok. Hopefully its not loading it too much, the input is through a 100k, the suck it and see test was to see if the trace altered noticeably on the scope. With the output of the feedback grounded and connecting it in and out of the RF out made no noticeable difference so I passed it as OK. I'll try it on a separate winding when I get back to the RF side of things, this should isolate it somewhat and allow me to simplify it a bit (get rid of clamping and use full wave rectification instead?) The harmonics are present but seem to die off pretty quickly. I'll put in a low pass filter before any antenna on the next iteration and hopefully that will improve matters a little. Kalee20 very kindly sent me some more valve bases and a few bits and bobs today so will be back on it over the next few days and will report back with progress Cheers Dom

27th Feb 2008, 6:27 pm	#46
kalee20 Dekatron Join Date: Feb 2007 Location: Lynton, N. Devon, UK. Posts: 7,081	Re: AM Micro Transmitter I tried screen modulation many years ago, and although it worked, it wasn't particularly impressive. Linearity was a big issue if I remember OK. The combined anode and screen modulation looks much better to me, everything is proportional so as long as there's enough drive, the anode current pulses should vary linearly with the modulating voltage. It takes more power from the modulator (anode and screen current rather than just screen), but is worth it. The idea of a pi filter on the output to me is the way forward, to reduce output harmonics. Reminds me of work I once did with resonant switch-mode power supplies. I've got some ideas on how to calculate the values, but they're not ready for sharing yet! And I don't think that changing the output circuit will invalidate the rest of the good work that Dom's doing anyway. The loading imposed by the feedback detector should indeed be negligible given the high values of load resistor. (Although wouldn't a 6AL5 double diode be nicer than those nasty PN junctions!!)

27th Feb 2008, 6:45 pm	#47
mickjjo Rest in Peace Join Date: Jul 2004 Location: Dartford, Kent, UK. Posts: 1,661	Re: AM Micro Transmitter There are some interesting circuits on the 'net for AM micro power transmitters like this one also using diodes in the feedback circuit:- http://www.dogstar.dantimax.dk/tubestuf/amtx-3.htm This one uses a 6AL5/EB91 in the feedback circuit:- http://www.dogstar.dantimax.dk/tubestuf/amtx-4.htm Interesting output stage too. Regards, Mick.

27th Feb 2008, 6:53 pm	#48
Ian - G4JQT Octode Join Date: Jun 2004 Location: Reading/Fakenham, UK. Posts: 1,323	Re: AM Micro Transmitter The above valve circuits seem mighty complex for just a few tens of milliwatts... Ian

27th Feb 2008, 8:01 pm	#49
MichaelR Retired Dormant Member Join Date: Mar 2006 Location: Sheffield, South Yorkshire, UK. Posts: 1,587	Re: AM Micro Transmitter maybe but there is a lot of education going on here Ian, personaly I am loving it mike

27th Feb 2008, 8:13 pm	#50
kalee20 Dekatron Join Date: Feb 2007 Location: Lynton, N. Devon, UK. Posts: 7,081	Re: AM Micro Transmitter The web pages that Mick has given us links to, generally use a frequency-changer valve as a modulator. After all, the valve is a multiplier which is exactly what an AM modulator needs to be. Then, the modulated RF is fed to what looks like a linear output amplifier (working in Class A). It's a bit complicated, but does work! Dom's work on the other hand is heading more towards a very low power version of a full-scale transmitter. Either way is fascinating.

4th Mar 2008, 12:20 pm	#52
dominicbeesley Octode Join Date: Nov 2004 Location: Hebden Bridge, West Yorkshire, UK. Posts: 1,885	Re: AM Micro Transmitter Hello, I've been working away over the weekend to make a higher voltage version, running from a mains Tx and get the thing into a nice box. This is proving to be quite difficult with lots of stability problems being introduced by running from an non-stableised supply. I've got the whole thing wired up with what should be a nice star ground etc. I managed to get the AF amplifier and cathode follower working nicely into a dummy load and also into a speaker through a large HiFi Tx and sounding good. But when I connect up to the actual PA valve, set up just as before, it sounds absolutely terrible! I've not even tried the feedback loop yet, with what I suspect to be HF instability of some sort I suspect all I'd get is a nasty squeal! The circuit is pretty much as before except that the first stage is decoupled via a 47k resistor and 1uF to ground, the other two AF stages have their own 0.1uF decouplers and the oscillator has a 2200uH choke and a decoupler added... Any ideas where to attack first? Its quite frustrating that something that worked great when it was sprawled randomly and messily on the bench then misbehaves when I've tried to lay it out nicely with short leads and proper earthing etc! Oh well.... Dom

4th Mar 2008, 2:25 pm	#53
kalee20 Dekatron Join Date: Feb 2007 Location: Lynton, N. Devon, UK. Posts: 7,081	Re: AM Micro Transmitter Hi Dom, good to see you back on the project! Did the transformer arrive OK? Posting up the latest circuit diagram would probably help a lot. I suspect you've got excess AF gain now - I've always found layout is critical when you have lots of gain. Dunno why it should be OK with a dummy load... Decoupling with 0.1uF does sound rather light. Hopefully, with the diagram up, it will all be clear! Isn't it annoying when a side issue grabs all the attention and has to be dealt with before the main drive can be restarted??!!!

7th Mar 2008, 12:56 am	#54
dominicbeesley Octode Join Date: Nov 2004 Location: Hebden Bridge, West Yorkshire, UK. Posts: 1,885	Re: AM Micro Transmitter Hello all, I got a bit further this evening, after a rebuild. Found out that the reason it sounds so bad is because its overloading all the sets in the vacinity. It sounds fine at the end of the street! All this with no aerial to speak of (about 5' of wire), I can probably lose a bit of the power/voltage somewhere in the PA stage! It now sound pretty nice from the Goblin Timespot in the attic but a bit ropey from any transistor set in the ground floor (overloading RF/IF stages?) I'll hopefully now be able to make the PA valve sit somewhere above ground (DC operating point with a big cap to hold it there) and so get 100% modulation without over-driving the AF stages. One thing I can't work out quite is why the two circuits in the attached pictures don't behave the same at RF. Am I missing something. I want to use the second (rightmost) as the load of the PA valve as I've already got a dual gang variable capacitor, one gang of which I'm using for the oscillator section. Dom Attached Thumbnails Last edited by dominicbeesley; 7th Mar 2008 at 1:05 am.

7th Mar 2008, 8:37 am	#56
Sean Williams Dekatron Join Date: Feb 2005 Location: St.Ippolyts, Hitchin, Hertfordshire QRA IO91UW Posts: 3,517	Re: AM Micro Transmitter Dom, What HT voltage and current do you have on the PA valve? My pantry transmitter has a similar coverage, and that uses an EF80 as the "PA" - has about 150v on the anode at around 4 or 5 ma The problems you are experiencing now are quite normal - happens here all the time just part of the fun when homebrewing stuff! __________________ Engineers make things work and have spare bits when finished

7th Mar 2008, 2:41 pm	#60
dominicbeesley Octode Join Date: Nov 2004 Location: Hebden Bridge, West Yorkshire, UK. Posts: 1,885	Re: AM Micro Transmitter There are a few higher up on the hill with blades about 3' long. The green ones where the blades point away from the wind (don't know the make) seem to create a real racket on the AM bands, I'm guessing its the inverters, though it could be the actual generators themselves. A bit like putting a radio next to a computer but lasts a few hundred yards blocking out R4 on LW. The big "proper" turbines, there are a lot round here in various groupings, are well behaved.