PWM

[Stretchyman]

Modulating your Tx with a mod tranny is fine for moderate power levels but the size, weight, & power loss not to mention frequency related problems tend to get a bit too much with moderate power levels.

ALL that and you have to wind the damn thing too!

I looked at PWM a few years back and looked rather complicated but then I noticed a design based around a single chip.

I've tweaked this and ended up with a neat solution.

I've attached a picture of my current MiNi PWM.

SMT or thru' hole components and can pass 200W with a TO-220 FET and probably 1KW with a TO-247 FET.

Needs a filter on the O/P but that's just 2 coils n 2 caps.

You can set the carrier to ANY level and mod upto Vmax, so 150% is easy and sounds LOUD!

The ONLY downside of this is as it's a (lossless) 'SERIES' modulator you have to use TWICE the voltage as you don't get the voltage doubling action of the missing mod tranny.

Let me know if you want any more info.

Stretchy.

[redhat]

I take it the black block is your DC/DC converter to give you the floating supply for your fet driver?  What are you using for the PWM generator, a TL494?

+-RH

[Stretchyman]

TL494 NO Good!

It's the UCC35702 and will work down to 10V so fine on a draining battery.

you wont get low mod % with a TL494.

YES! it's a DC to DC to float the supply for High side drive, could be done much easier with low side drive but don't fancy a floating RF earth!

This style of circuit is the ONLY one to use, done bother with the linear one, it's class A and ONLY 20% eff, PWM is 90%+

Having a PCB negates any complication issues, if you can solder you can build it!

2 chips, a FET & a diode, why use anything else??

OK have added my orig 2 FET design with audio LPF, will pass 1/2 KW, YES 500W, crazy!

[redhat]

I found similar issues with the TL494.  Most power management IC's have built in dead time in their oscillator circuits that prevent cross conduction in PSU power stages.  This is all good and fine, but it limits your minimum duty cycle to around 10-20% which will limit your negative modulation.  I wound up using a discrete design based around a 1 MHz oscillator, followed by a divide-by-10 circuit, and finally an integrator to get the linear triangle waveform.  From there it's just a comparator and a fet driver.  I echo the difficulties of a floating RF ground, but when you get above a certain power level, its much easier to rectify the mains, apply power factor correction, and transformer isolate the RF output.  If the drive signals are close to 50%, you can transformer couple them to the RF stage, or use photo-couplers.

Neat stuff.

+-RH

[Stretchyman]

OK, Yes can get a bit overly complex hence 'High Side Drive', you can keep the RF stage 'Normal'

The best implementation of 'Low side' I've see that keeps it simple is with the 'Super Senior'

I guess you're familiar with that design?

Stretchy.

[redhat]

Yup, I've seen it, although he doesn't have a 7 MHz version yet

+-RH

[Stretchyman]

Just copy the modulator, your there then, even simpler than mine!

 

 

[redhat]

They say imitation is the sincerest form of flattery, but I do like to try it myself at least once

+-RH

[John Poet]

PWM = 'Pulse Width Modulation'

I had look that up...

I am extremely jealous of you guys... meantime, I think my brain is gonna explode after reading all that...

Maybe I'll just go wind an antenna coil.  That I can do
(after the online calculator tells me exactly what I need!)

[redhat]

For those out there that are still not sure what we're talking about, a good primer comes from WA1QIX's website http://www.classeradio.com/pdm_article_solid_state.html and also the FAT-MAX PDM modulator from the UK http://www.shortwaveradio.co.uk/puwma%20draft4.pdf

+-RH