Construction Begins - Class-E RF deck - Updated 2/7/2005

[Brent]

More construction

https://home.comcast.net/~w1ia/


Brent(Tina)

[frank carcia]

Brent,
The secondary leads on the driver transformers are quite long. Consider flipping them over to eliminate the long leads. This could degrare the gate drive wave form. Looks like serious 3890 strappage. fc

[ke1gf]

B-TINA, I took a look at your pictures and I see that you've got the grounds of your input transformers connected to the heatsink with long leads and not to your source bus...

This is a bad idea, the grounds should be soldered directly to the source bus so that the transfomer is directly connected between the FET's sources and the Gates....

Once other thing, I believe that you wound your input transformers with 4 turns on the primarys, this is the way that I initially did mine, But I changed to 6:1 turns ratio with outstanding results. When I was running 4:1 my twisted pair would get hot and the modules were a pain in the butt to balence.

Sincerely
-Bill 'GF

[Brent]

B-TINA, I took a look at your pictures and I see that you've got the grounds of your input transformers connected to the heatsink with long leads and not to your source bus...

This is a bad idea, the grounds should be soldered directly to the source bus so that the transfomer is directly connected between the FET's sources and the Gates....

Once other thing, I believe that you wound your input transformers with 4 turns on the primarys, this is the way that I initially did mine, But I changed to 6:1 turns ratio with outstanding results. When I was running 4:1 my twisted pair would get hot and the modules were a pain in the butt to balence.

Sincerely
-Bill 'GF

Thanks Bill...yup I allready planned on the change for 6 turns after talking with steve...I could run the gate tranformer ground directly to the source bus...I thought about it...I should have given myself a little more room on the layout.

Brent(tina)

[frank carcia]

Brent Why don't you move the output transformer or driver transformer to the other side of the heat sink? This way you will have more room. fc

[ke1gf]

Tina... I've been following your pictures, keep up the good work!

Another idea that has crossed my mind. The filter that you've built has a first inductor value that will only give you maybe 100% +- 10% modulation percentage if correctly loaded... You'll notice that your positive peaks will flat-top. When you get rig up and running you might consider pulling a turn or two off of the first inductor But watch for negative peak crapout. It's a tradeoff between negative/positive peak attainability. This is one of the shortcommings of this particular implimentation of PDM.

-Bill 'GF

[Brent]

Tina... I've been following your pictures, keep up the good work!

Another idea that has crossed my mind. The filter that you've built has a first inductor value that will only give you maybe 100% +- 10% modulation percentage if correctly loaded... You'll notice that your positive peaks will flat-top. When you get rig up and running you might consider pulling a turn or two off of the first inductor But watch for negative peak crapout. It's a tradeoff between negative/positive peak attainability. This is one of the shortcommings of this particular implimentation of PDM.

-Bill 'GF

I will check when finished....long way off...filter almost finished today...the stand-off's for the inductors are going to use 10-24 rod...was thinking this was a little light...is that what you used to support the coils?

Did you change that value on the first inductor?

Brent(Tina)

p.s. new pics on the web

[frank carcia]

Brent,
I notice you only have 2 screws on your source busses. The coating on the heat sink is an insulator so this places a huge inductance in your source circuit. I suggest you drill and tap a hole near each source lead to eliminate this loop. fc

[Brent]

Brent,
I notice you only have 2 screws on your source busses. The coating on the heat sink is an insulator so this places a huge inductance in your source circuit. I suggest you drill and tap a hole near each source lead to eliminate this loop. fc

There is actually 3, but I may change them to the zinc cap screws instead of the black coated ones....why I bought those is beyond me?

BT

[ke1gf]

Did you change that value on the first inductor?

which value is that Tina?

I counted your turns

-Bill 'GF

[Brent]

Did you change that value on the first inductor?

which value is that Tina?

I counted your turns

-Bill 'GF

The first inductor in the filter....the one you were refering to.

[frank carcia]

Brent,
It is very important to have a low Z to ground on the sources. I was going to mount 22 ground lugs at one time but it would have taken hours to pull my final apart with another 22 screws. That was why I used the plane on the circuit board. Your layout would be perfect for a ground lug at each FET lead making the connection lead almost zero inductance.
I did that on my 160 meter rig modulator sources.
My point is the coating on the heat sink is an insulator. fc

[W1DAN]

Hi Brent.

Your deck looks fantastic! I am waiting to see how far you peg my s-meter.

On my wimpy and ratty PW little RF deck, I used a sheet of PC board copper as the base of the FET's, caps and transformer grounding. I soldered all grounds to this. I then mounted the assembly to the heatsink. When possible I attempted to have the power and output transformer grounds near the center and symmetrical.

I had to pre-drill the heatsink, mount the FET's and solder to the copper plane before the copper was screwed down to the heatsink. The heatsink will sink the heat from the soldering gun and causes poor solder joints.

For the gate and drain busses I used strips of copper cut from the same sheet. For the drain strips I left an oval to fit over the drain caps.

But I know you will do great on your deck.

A picture is available at:

http://www.dbaudio.net/class_e.html

BTW, I made 400 watts of modulated carrier with the 6 FET in push pull. Did not stay there long as my PS would not live that long at that power level. No FET's popped though!

Hope this helps.
73,
Dan
W1DAN