MOSFET RF power amplifier design

[Buddhafollower]

Good day,
Can someone recommend me a website or some documents to read so I can have an idea as to how I can build/design an RF power amplifier with MOSFET's?
I am interested in the subject. I like it but it's all a mystery to me for the moment.
I imagine somewhere on the world wide web there must be a website dedicated to this subject. I really wish I would design my own amplifiers. I'am interested in broad band RF power amplifiers with amplification extending from 630m to 40m. So between 472kHz to 7.3MHz. I think the best source will be a website created by a radioamateur.
Thank you,
Michael

[spacecase0]

I have been playing with amateur radio for about 30 years now.
RF power amplifiers do some odd things,
you really need filtering to make sure that you are not transmitting out of frequency range.
so, usually an amplifier that works on 472KHz will not be the same one that works on the 40M band...
though there are some that will do it, you might want to build 2 separate amplifiers to get that kind of range, or at least have selectable filtering for what frequency you are using at the time.
my next question is how much power out are you looking for ?
I have been collecting schematics for this sort of thing for many years, so I likely have schematics of what you want.

[Buddhafollower]

Good day,
I'am looking for a few amplifiers made with common N-channel IRFP transistors. Amplifiers that work from 250kHz up to 7MHz. No problem, I can build more than one to cover the band.
Power output, at least 50W or even 20W.
Thank you,

[spacecase0]

that should not be to hard.
here is one example.
filters and setup/tuning are also needed to complete this, but I want to know if this makes sense to you so far.
and you can change the mosfets for ones you like if you are running lower power.
what kind of source are you driving this with ? (power level and impedance would be handy info)
this setup is assuming that you are needing it to turn off to receive. if you are transmitting only then it can be simpler.

I got this off of the web somewhere I forgot, so sorry for not identifying my source.
MOS-PA-scematics.jpg
150w-mos-pa-1.jpg

MOS-PA-layout.jpg
MOS-PA-PCB.jpg

output-transformer.jpg

[Buddhafollower]

I know this design. I have it. It does make sense but I would take the relays out and the two transistors leaving only the IC. I hate relays and transistors in these designs. It complicates the schematic and I'am not planning to become a radioamateur.
To drive this I have a signal generator with a maximum of 20Vpp. I will build an amplifier from a russian schematic. This amplifier gives few watts (3-5W).
Here is the schematic diagram. I'am referring only to the first half. It's power transistor is a 2T920A or KT920A. 5W Should be enough for many amplifiers. The one in my image can be supplied with a voltage of 0.1 to 0.3Vpp. It is very sensitive. I just tested it on the oscilloscope and it sure can amplify all the way to 24MHz, only that my oscilloscope can go as far as 20MHz so the measurement is quite unreliable to say the least. Clean sine wave, that is my point.

[Buddhafollower]

I have a question, how can I find out the input impedance of a Mosfet? I can see all these input transformers and they all have a certain primary to secondary turns ratio. I don't know how to find out the input impedance nor the output impedance. Someone said that my answer is in the transistor's datasheet but i have not a clue where to look at. They said that the Zout is dependent on frequency. Well... that complicated things even more. From Lone Pine Writings I know that impedance is Electromotive force divided by Displacement current.

[Buddhafollower]

I did some calculations. See if they are correct.

[spacecase0]

the input impedance of a Mosfet is going to change with frequency.
they act like a tiny capacitor on the input.
so, look up the input capacitance and then calculate using your frequency to get current.
on lower frequencies you can assume the input impedance is infinate

so, now that I know you are transmitting only, and powering with a signal generator,
mosfets are a voltage controlled device,so, assuming you have the current out to drive the input (you likely do) and if you have DC offset on your signal generator, you might not have to have much of the circuity you see.
you can set your DC offset to get your idle current on the mosfet (yes you would need a separate channel out for each of the 2 mosfets)
and run into the output of the mosfets to the transformer the same as the schematic you posted (or the one I posted)

Edit:
the schematic and transistor you posted is a regular transistor and not a mosfet,
the impeadance of them is much harder to figure out, and the circuitry to drive them is more complex.
and you need multiple stages of amplification to get to your final power.

[Buddhafollower]

Ok but what mathematical formula do I use to calculate the Input impedance of mosfets using frequency?

I'am asking this because I can use the formula for reactance but I know not how to use the -j operator.... imaginary number, i never understood them....
Excuse my stupidity.

[Buddhafollower]

I have used the formula for Capacitive reactance. The calculations make sense indeed. For a certain amplifier, the output transformer ratio is good. It is broadband really. I have already built it. Let me show you:
http://f4eoh.free.fr/amplificateur_h...fp250_1110.htm
Tomorrow I shall take a look at the input ratio as well.
Thank you,

[Buddhafollower]

The calculations with the reactive capacitance formula works well enough for me. No complex number required so far. I don't know how to use that anyway.
This is what I got.
Concerning the above design, the ratios for the input and output transformers are as specified in the design papers.
Zin= 1/2pi*14,350,000Hz*0.000.000.0028F= 1/0.252= 3.97Ohm| sqr(50Ohm/3.97Ohm)= 3.55 so 4:1 ratio
Zout= 1/2pi*14,350,000Hz*0.000.000.000.78F= 1/0.07= 14.28| sqr(50Ohm/14.28)= 1.87 so 1:2 ratio
This is because the amplifier was design to work in the 20m band.

[spacecase0]

yes, Capacitive reactance,
so seems like you are on the correct path to me.
one more thing,
most mosfets have a gate voltage of 20V before they break,
so, make sure you don't go over that limit.

I am here if you have any more questions. (like setting your bias voltages)

[Buddhafollower]

Yes please. Help me understand how I can calculate the bias voltage and current for mosfets. Thank you for your help.

[spacecase0]

Yes please. Help me understand how I can calculate the bias voltage and current for mosfets. Thank you for your help.

the bias voltage is adjustable, (see the variable resistors in the schematic), you set them to zero volts when you first power up,
ground the signal in, you want no input, just the bias.
measure the current in the mosfet. then turn up the bias voltage till the current starts to go up.
you want to set the current approprately for the mosfet, most of the ones I have set up are about 50mA or something,
but you need to look at the waveform out with an input to figure out the setting if you don't know it to start with.
this setting has to be set separate for each of the mosfets.

[Buddhafollower]

Yes, that's how I know to do also. I imagined there were certain calculations. For the amplifier in the link I gave you, per mosfet you ought to have not more than 50mA. Even this is a lot, so the designer specified a current of at most 30mA per mosfet.