The AM Forum

I decided to make a lin. amp with 2 x EB104 boards in parallel. These boards are purchased at Communication Conxcepts and I am quite happy with the quality. I didn't encounter a single problem during assembly. I had a stock of BLF177 FET's , so used these in stead of the Motorola devices. The input is quite simple, a power divider and two resistive attenuators of approx 3 dB. But the output can be a bit more difficult to combine, one need a 25 to 50 Ohm transformer. These can be purchased , but I considered it more simpel to add 1 turn more at the output transformers of the EB104 boards (use teflon isolated wire!!), so 4 turns instead of 3. That makes the output impedance close to 90 Ohms and one can use a simple twisted line combiner to combine the outputs of both amps. More wideband, less losses and very simple.
The amps are slightly low loaded, but I get 1100 Watts out of 2 boards at 48 Volts without a single problem.
The amp doesn't have any protection, the BLF177 should be quite rugged and indeed they are.
I can transmit full power at an open antenna, a short circuit, switch the wrong low-pass filters (0,1 dB Chebichev L-C-L-C-L, air coils) the antenna or the tuner may arc, but the amplifier doesn't mind.
I use a meanwell 48Volts 1500 Watt switching power supply. The current may peak upto 35 Amps worst case.
For AM, the carrier is 250 Watts. The negative feed-back resistors at the EB104 boards are a little small and may get quite hot. It may be better to take resistors with twice the dissipation.
I am very happy with the overall results, IMD seems close to -35 dB and quite low cost.  I am running the amp in general at 1 kW SSB, my driver set is a TEN-TEC Corsair 2 and doesn't have AM. I tested the am at AM with my homebrew AM EL84 transmitter

Very nice!
I like the big copper heat sink!
(Could not find one piece that big, had to use two)

Just getting started on mine, very similar. using two NXP
1500 watt LDMOS devices and a combiner.

http://www.nxp.com/products/rf/rf-power-transistors/rf-broadcast-and-ism/1500-w-cw-over-1.8-500-mhz-50-v-wideband-rf-power-ldmos-transistor:MRF1K50H?&fsrch=1&sr=2&pageNum=1

Should be very low stress for AM.

That will be a hefty one!, nice project!!. I made something like that with its smaller 1 kW brother for 27 MHz plastic welding.
Take care, there is a theoretical problem with these circuits that NXP nor Freescale don't address.
When you imagine the two FETS in phase, not in balance, the input and output transformers do not couple to the input and output load. That means that the gates and the drains will see an unloaded inductance which can give rise to IN PHASE oscillation. I had an amp that did almost destroy itself by generating hundreds of volts at the drains without any output and with low current drain. NXP and FREESCALE did not consider this and do warn for instabilities at low frequencies, wich is not correct.
The solution is very simple, just load the split secundary of the input transformer with  e.g. 2 x 3,9 Ohm or so to the center of the sec. That prevents that the gate sees an in-phase open inductance. Photo's attached.
Please keep me updated with the progress, I am highly interested!