Analog modulating a 1000 watt carrier class E rig

[ka1tdq]

It's starting to come together. I've started buying all the silicon stuff. I've made a link for the progress of the project here:

http://ka1tdq.radio/current-project/

Jon

[ka1tdq]

I bought all the parts for the heat sink assembly and now it's starting to take shape.

Jon

[ka1tdq]

Here's with all the FETs and drivers mounted. The heat sink assembly is almost done!

Jon

[Opcom]

In a wholesale catalog of car audio gear there were the 4KW amps and all of that along with the 100W/CH types. The big stuff was class D.

The point is that the power ratings for hypothetical identically-rated amps were sometimes given as "4000 Watts MAX" with no other indication, and otherwise given as "2000 WATTS RMS". To add to confusion, some would headline with 4000 WATTS, and then in the bullet points, state 2000 WATTS RMS.

Store/shop sales people a lot of times have no idea what this is about and just try to state the higher rating. The salesperson at the distributor will know the facts and was more than happy to explain that MAX meant driven as hard as possible to square wave clipping, which was for the competition and rap/bass-boom car people, and RMS meant without clipping, which would be the rating for the hi-fi people.

Part of a wholesale catalog page attached to clarify the marketing.

[KB2WIG]

P,


Nothing about Bird watts, is there?


klc

[ka1tdq]

The medieval version of explaining MOSFET amplifiers to the general public.

https://www.youtube.com/watch?v=_rMsgmaBV8g

Jon

[Opcom]

P,


Nothing about Bird watts, is there?


klc

naw.. just the audio frequency kind. Hey the Bird Watts come from the fantastic class E stage!

[ka1tdq]

I have the driver rail supply installed and now need to do the tiny soldering of all those small resistors and caps.

Jon

[ka1tdq]

...and more stuff

Jon

[steve_qix]

Looks good.  The shunt capacitors should be around 1500pF for 6 MOSFETs, unless you're going to run lower power.  It looks as if the shunt is 1000pF.

[ka1tdq]

They are 1000pf. It’s what I could find for ATC capacitors. I’m only going to run 400 watts carrier, so everything should be well within margin.

Soldering tolerances are closer than what I expected on paper. Things are really tight! But, I’m going to check everything as I go. This is too big of a transmitter to do otherwise.

Jon

[ka1tdq]

The heat sink assembly is all done except for testing the gates under drive and installing the ground shield between the drain and drive buses.

Jon

[steve_qix]

The shunt capacitor keeps the drain voltage in check.  It's a good idea to do that - vastly improves reliability (among other things).

I have used these caps (and other similar caps of varying values) with good success in many class E projects:  https://www.ebay.com/itm/560-pF-500-V-5-LOT-OF-1-RUSSIAN-SILVER-MICA-CAPACITOR-SGM-3-3/282488834530?hash=item41c5a589e2:g:lmcAAOSwao1bEkqV

You could put one on each drain bus, which, in addition to the capacitance you already have, would give you a total of slightly over 1500pF.

[ka1tdq]

Thanks! Yeah, I thought about that afterward. Even though I plan to only run 400 watts carrier right now (QRP), later on I might try more and couldn't with that limitation in there. It's better just to do everything right from the beginning.

Jon

[Opcom]

Just noticed something in the latest image. The RF cables are held in place with large eye bolts. Very appropriate engineering measure for this project.

Used to do something similar with an old rat rod - chained the stoked 400 Pontiac engine to the frame because otherwise it would sometimes break the motor mount on launch. The Millenium Rat is missed.

[K8DI]

Maybe it's the picture angle, but it looks like the number of little capacitors is inconsistent -- 4, 5, 5, and 6 per bank. Is this so? Or just my eyes fooling me? If so, why?

Ed

[ka1tdq]

Good catch. Those caps are on the DC supply rail for the IXDD bus. I ordered 20 caps with the intention of using 5 per MOSFET rail. During construction I got solder happy and put one too many on one section. I shorted another section.

It’s no biggie. The whole thing is DC common anyway. It shouldn’t matter. I know this from my last transmitter. I accidentally wired half of them backwards when during construction I didn’t realize they were polarity sensitive. Those that were wired backwards popped like firecrackers. The transmitter still worked fine on half a load.

Jon

[ka1tdq]

These are little more expensive, but I think well worth the cost.

https://www.ebay.com/itm/ATC-PORCELAIN-HIGH-POWER-RF-CAPACITOR-680PF-2500VDC-20-MICROSTRIP-TERMINATION/323824215343?hash=item4b656d652f:g:SnAAAOSwyTtc9GVB

The Russian silver mica caps only have a voltage rating of a little over 500 volts. A combined shunt capacitor value of 1680pf is a little high, but I can stand a little loss in efficiency for an improved safety margin.

On a side note, I'm building this project on a budget of $90/month. It was a challenge put to me by my wife. I cheated a little and sold the ARC-5, but basically it has come this far on a pittance.  All I have left to purchase are the cores, tuning and shunt caps.

Jon

[steve_qix]

Those ATC caps will work, and the 500V Russian caps will also work.

The 500V rating is plenty - the highest peak RF voltage at full modulation, assuming the modulator can provide 130V peak is under 500V - and that 500V rating on the cap is the DC rating, so you're good.  The ATC caps I use are typically 500VDC units.  I also use the Russian 500V silver micas.

I tend to go with lower voltage ratings if possible.  For the same (physical) size cap, the current rating will generally be higher with a lower voltage device.

[ka1tdq]

I actually decided to go with 8 more of the 1000pf value ATC caps. I'll put two in series and parallel them with the existing ones for exactly 1500pf.  It'll work out mechanically as well since I'll bridge the bottom of the second one with another capacitor rather than using a piece of wire to form a "U".  It'll look like it was meant to be there.

I didn't think about size and current/voltage rating stuff. I'll have a parallel path siphoning off some current and all that good stuff. 

[ka1tdq]

It's starting to look like something!

Jon

[ka1tdq]

A little more progress.

Jon

[ka1tdq]

I can't do any more until the vacuum variable and the cores come in at the end of the month. I did get the additional 1000pf ATC caps, but I won't solder those on until I get ready to test the heat sink assembly.

Jon

[ka1tdq]

This picture shows the mounts for the vacuum variable cap almost done. I'll drill a hole for a #10 bolt through the copper to secure the cap mechanically to the plexiglass and it'll serve as a connection point for lugs.

For feng shui, I ordered a clear acrylic 1/4" shaft for the knob. It's also nice to not get zapped while tuning.

Jon

[ka1tdq]

I got the meters wired in. All that's left is to complete the heat sink assembly and connect it to the transformer.  

Jon