The AM Forum

For the past few days, I've been using a new RF amplifier I built - a 24 MOSFET unit that covers 75 and 160 meters.  The amplifier is organized into 4  6 MOSFET modules in single ended push pull.  Finally got it on the air!

(http://www.classeradio.com/finished.jpg)

The digital drive came right up - IXDD414 drivers, one driver for every 2 MOSFETs.  The output of the homebrew VFO drives the driver ICs directly - 6 ICs per phase, 3 per 6 MOSFET module.

The bandswitching is accomplished very simply:  For 160 meters, the VFO output is set to the 160 meter band, and generates the two phases.

There are 4 Radio Shack 30 amp relays, one for each module, to switch in additional shunt capacitance when on 160 meters.  There is a contactor that is used to short out or not (depending on the band), the 160 meter coil.  On 160 meters, the coil is not shorted out and the relays connecting more shunt capcitance to the drain busses are energized.

When using 75 meters, the VFO output is set to the 75 meter band; the relays that bring in additional shunt capacitance for 160 are not energized, and the contactor shorting out the 160 meter tank coil is energized, shorting out the coil and effectively removing it from the circuit.

The modulator is a pulse width modulator, using the latest PWM board set.

Anyway, that's pretty much it !!  The transmitter came up nicely on both bands.  The 160 meter switching circuits and components were added after the fact, however there was so much room in the RF cabinet, the additional components were no trouble at all.

I have documented some of the major steps in building this transmitter.  Here is a list.  Click on any step to view!

http://www.classeradio.com/step1.jpg (http://www.classeradio.com/step1.jpg)
http://www.classeradio.com/step2.jpg (http://www.classeradio.com/step2.jpg)
http://www.classeradio.com/step3.jpg (http://www.classeradio.com/step3.jpg)
http://www.classeradio.com/step4.jpg (http://www.classeradio.com/step4.jpg)
http://www.classeradio.com/step5.jpg (http://www.classeradio.com/step5.jpg)
http://www.classeradio.com/step6.jpg (http://www.classeradio.com/step6.jpg)
http://www.classeradio.com/step7.jpg (http://www.classeradio.com/step7.jpg)
http://www.classeradio.com/step8.jpg (http://www.classeradio.com/step8.jpg)
http://www.classeradio.com/step9.jpg (http://www.classeradio.com/step9.jpg)
http://www.classeradio.com/step10.jpg (http://www.classeradio.com/step10.jpg)
http://www.classeradio.com/step11.jpg (http://www.classeradio.com/step11.jpg)
http://www.classeradio.com/step11a.jpg (http://www.classeradio.com/step11a.jpg)
http://www.classeradio.com/step12.jpg (http://www.classeradio.com/step12.jpg)
http://www.classeradio.com/step13.jpg (http://www.classeradio.com/step13.jpg)
http://www.classeradio.com/finished.jpg (http://www.classeradio.com/finished.jpg)
[/list]

I'm going to provide some written supporting documentation when I get time!

Regards,

Steve

WTG Steve, How much power can the run on AM?
Thanks for all the pics, It will be a big help to me
when I get around to building one..

73, Chuck

Steve,
Sounded good on 160M. What value shunt cap is on each module for 80M and what did you switch in for 160 in parallel with it.
Now imagine how easy it would have been if you built the universal driver board....

I use 1500pF of capacitance per module on 75 meters.  Nice waveform!  On 160 meters, I switch in another 1000pF per module, HOWEVER, it *should* be another 1500pF.  I didn't have enough 1500pF multilayer ceramic caps, so I used the 1000pF units for now.  I am getting more 1500pF capacitors.  With the 1000pF capacitors, the peak RF voltage at the highest percentage of modulation just reaches 500 volts.  With the proper capacitors, it reaches 450 volts, which is proper.  Of course, I'm using 900 volts MOSFETs (with 540 volt Transzorbs across them), so I think I'm well in the safety zone, but I would like to use the correct values just because...

Those multilayer ceramic capacitors from ATC are just great !! Low impedance, no heat, SMALL, *INEXPENSIVE* and easy to use!

On the driver board - yeah, when are you going to lay that thing out ??????????  ;D :D :) ;)

Regards,

Steve

Steve:

Congratulations!!

It'll be a great antenna ice melter!

I look forward to hearing it on the air.

Thanks for posting your work and ideas.

73,
Dan
W1DAN

A transorb on the drain will add some C. cool.

WOW,  Looks great Steve !

FB OM,  Will be listening on 75 M.   73    Vic

I donated the design hoping for a few cheap boards

Yeah, but its got to be layed out first   ;)  Make a layout in ExpressPCB, and next time I (or someone) has some boards made, whomever can piggy-back that layout onto the order.  It's generally just incremental in cost to do that, since the setup charge and the per-board charge are the big numbers.  I usually put at least 3 different boards on one larger board, and then cut each of the individual boards out of the one larger one.  I leave about 1/8 inch of space between individual circuit layouts.

Regards,

Steve

TO INFINITY AND BEYOND!!!!
Nice work, Steve.
DAM no clip leads...........
Beats tubes with handles, I guess

Fred

steve deserves a lot of credit for just going his own way and not following teh crowd. In doing so, he got a neworz crowd.  8)

All them black soldiers laying on their backs on the heatsink must put out at least legal limit.

Fred

barely Fred, Steve likes to derate parts so they never blow.

As a philosophy, I like to use larger numbers of small (read: inexpensive) parts in favor of a small number of expensive parts.  In the unlikely event of a device failure, you're looking at very small money.

This transmitter uses very cheap (under $3.00) power supply MOSFETs in the RF and modulator applications.  So, each module has about $17.00 worth of MOSFETs (6 MOSFETs) - compared to a single device that could do the same thing, which would cost probably over $100.00 - and maybe a WHOLE LOT more   8)

I have not had a MOSFET failure in a class E RF amplifier since 2002.  The TransZorbs work wonders in keeping any voltage spikes under control.  MOSFET failures are almost universally caused by a poor design, failing shunt capacitors or poor layout.  These problems have been mitigated by digital gate drive, MultiLayer Ceramic shunt capacitors, and a standard layout.  The combination makes for a very reproducable, stable solid state transmitter  ;)

(http://www.classeradio.com/topview.jpg)

Hi Steve,

Heard you on last week and a few days ago here on the West Coast.   Great signal !  Sounds very nice,  indeed.

73  Thanks  Vic

Steve,
Can I borrow the coil winder u used to wind those coils ;)

Just kidding Steve..........as usual the rig sounds great.......matter of fact I think I sent u a recording.

Bill

Steve were you running thi last night on 3885?? 40 over in Ga.fine audio...

Hi Jerry,  That was the rig!  I enjoyed your 'net last night.  Heard pretty much everyone who checked in.  The band was very good for most of it - the locals were there, but so were the distant stations.

Regards,

Steve

What MOSFET are you using in the RF amp? That is a beastly big radio there.

Frankly I'm disappointed...not enough clip leads ::)

Sounded great on air as I expected of course.
~ps