upside down tube transmitter/pwm modulator?

[kb3ouk]

Lately I've been thinking about alternative ways to generate AM, ways that are different than what is usually used (like my recent post on outphasing modulation). Well, here's today's question: I've been looking at the upside down tube circuit. This thing is basically a high level balanced modulator, capable of producing DSB with any amount of carrier depending on the plate voltage. If there is no B+, and the only power being applied is the audio, then there is no carrier, which produces a DSB-SC signal. Now, the circuit shows a mod transformer to couple the audio to the RF stage. What if you replaced the mod iron with a PWM moduator? You would only really need the audio, not the carrier voltage, if you wanted suppressed carrier, but still have the option to have some DC for carrier (like a normal PWM modulator) if you wanted to run carrier. A PWM modulator puts out DC for carrier, how would you set one to put out no DC for no carrier, or be able to adjust the voltage to set the carrier level? In the case I'm describing, the PWM modulator would probably be acting more like a class D audio amp. Has anyone else ever thought of doing this?
Shelby

[WA1GFZ]

The pwm would be more complicated and need to swing above and below zero for the upsidedown tube to conduct.

[kb3ouk]

ok, so out with that idea. but what if you used a direct coupled audio amp, then either fed through some DC for carrier or no DC/pure audio for DSB-SC?

[WA1GFZ]

as long as the audio was AC coupled

[The Slab Bacon]

Just FWIW, I have a PDF of Bill Nagle's original "upside down tube" circuit in my computer at home. (Hand drawn schematic) It is not all that complicated. If anyone is seriously interested, I will post it one evening this week.

It is one of the many "interesting" alternative ways of producing AM that have long been forgotten about. Nowadays, it is either "class E", balanced modulator, or plate modulated. There are many other ways as well to produce AM. the more you research it out, the more fascinating stuff you will find.

Don (KYV) is a wealth of informantion on this subject! ! 

[kb3ouk]

Is it this one? http://amfone.net/Amforum/index.php?topic=19258.0
I think what I might try is an upside down tube RF deck using a big single ended audio transformer fed backwards by a solid state amp for the modulator, then have a power supply running off a variac to adjust the carrier level, and also have a switch on the B+ side of the audio transformer to switch that side between the power supply (for DSB with carrier) or ground (for DSB with no carrier).

[K1JJ]

Shelby,

I have a digitally driven, 24 pill, PDM class E rig here. (With high rise pepperonis and dual anchovies)

Speaking of class E, by using the PDM carrier level adjustment within the PDM modulator, it is easy to set the carrier low enuff so that it will give 300%+ audio peaks. It is not a balanced modulator, nor does it have the classic DSB wavelets between major peaks. But by using the internal negative peak limiter, we are able to keep the carrier from going over -100% negative.  It sounds pretty good and reasonably clean when limiting like that and running big audio.
That is a good way to get high efficiency and big audio, if that's your goal.

I've built several upside-down TUBE plate modulated rigs, including a pair of 4-400A's and a bigger 4-1000A upside down with another that was plate modulated by a pair of 3CX-2500's. Now that rig had some audio capability!  The problem was the usual - the audio peaks were just too much for the average diode detector receiver.

The bottom line is if you want to play with a balanced modulator or true BM audio waveforms, it would be a better and easier idea to do it with a low level DSB rig and linear or even an SDR rig.  That said, these days there are many sync detectors listening so many guys would have no problem with the big audio. Still, many still have older receievers and would object.

I think the best approach, at least what I do for all my AM rigs, is to run a negative peak limiter and let our own natural voice decide how much postive peaks are generated. In my case I find about 130% positive is fine and most guys have no problem. When things get ruff, then 150%++ is easy with the class E rig - but the limit with the plate modulated iron rigs is perhaps 135% positive.

Tom, K1JJ

[kb3ouk]

Actually, I'm just looking to have a rig capable of AM or DSB-SC. On AM, the setup I plan to use would be capable of only about 145% positive peaks. For what i want, I could go the low level DSB and linear route, but what's the fun in that? I was also considering a sollid state diode balanced modulator that could be unbalanced to pass a little bit of carrier, but I like tubes more.