Extract parameters from very high power RF transistor. How?

[sinatra]

I have some very high power RF transistors.
They should be able to provide more than 200 watts at 400MHz.
I would like to use them to build an amplifier but I have no datasheet.
I would like to extract the s-parameters of those transistors.
I think I should extract the parammeters at the desired output power since those are hevily non linear devices.
I have a network analyzer, couplers, amplifiers to provide the driving power, etc...
I'm not sure about it but I had seen on the net that people use "tunners" for this.
Any body can help me to show me the way????? :'(
Is there an easy way to do that????How to proceed????
Thanks a lot to all you.
Greetings
S.

 

[vfone]

Re: Extract parameters from very high power RF transistor. H

Maybe this link can help finding something useful about large–signal analysis.
Is bad that you don’t have the datasheet, because you have to know at least the maximum supply voltage and maximum bias currents, otherwise you could kill the transistor before take any measurements.
**broken link removed**

 

[sinatra]

Re: Extract parameters from very high power RF transistor. H

Yes vfone. I agree with you, without the datasheets the things are difficult.
I believe it is one of those ldmos that uses 28V power supply.
I tryed to find the datasheet but was not successful.
I wanted to profit on this situation and try to learn how to extract aproximate parameter of a transistor like that and then biuld my own design, this would be a complete experience. :') (maybe mazoquism).

For the power this transistor should deliver (I have an amplifier that uses this transistor) 200 to 300W I expect that the real-part of the output impedance of this transistor (in reality the conjugate of what the transitor should see from the load/matching structure) should be something arround 3 ohms.

I was thinking about building a kind of "tuner" that would allow me to adjust the tunning in both input and output in such a way that I could use a kind of brute force method and make the device work (for me the load pull method is nothing but a brute force method.... and I tought about going the same way....but maybe making the tunners kind of home made tuners that would allow me to do that.....

Have you even seen some thing like that......???
S.
ps:I'm not yet talking about the power supply that will feed this best :') But I have it :')
I also have the loads that can withstand up to 1KW.... :')

 

[vfone]

Re: Extract parameters from very high power RF transistor. H

You can use tuners for input and output, but you have to use Bias-Tees to be able to provide the input and output bias.
May find here a datasheet close to the one that you are looking for. Only to give an idea.
**broken link removed**

 

[sinatra]

Re: Extract parameters from very high power RF transistor. H

Yes vfone, I believe this is the type of transistor I probably have.
By chance, I found a quite nice application note in the site you indicated.
www.st.com/stonline/books/ascii/docs/2904.htm

There is a good description of a balanced topology.
Similar to what I want to make.

Anyway, it is not an easy job to characterize a power transistor. Looks like every body is using those devices and noone ever characteriyed one.....who came first the egg or the chicken :')
S.

 

[toonafishy]

Re: Extract parameters from very high power RF transistor. H

You need to "load pull" the transistor. You present varing output impedances to the device and plot the results on a Smith chart. You will generate circles of constant IM3, Pout, gain, similar to the noise circles of a LNA transistor. You can do this with three stub tuners if you can find ones that can handle the power. It is very tedious to do this by hand which is why you don't see the information very often. Maury Microwave makes an automated load pull setup $$$$.

 

[molloy]

Re: Extract parameters from very high power RF transistor. H

Hi,
try to give a look at this article: "Impedance measurements for high power RF transistors using the TRL method" you can find it at the site of MIcrowave Journal October 1999.
I think it is very difficult without any other informations: you should know at least if the device is a push pull device for instance and the bandwidth (it is i think prematched and postmatched). Maybe you should know even the output power you have to match the device

stefano

 

[sinatra]

Re: Extract parameters from very high power RF transistor. H

Hello all.
Molloy: Thank you very much for this article. This is exactly what I was already preparing. It is good to see the article. I see it is worth making such a system, and good to see that somebody already did it and even published....makes me feel more sure....

The transistors I have are old replacement parts from an amplifier I have in operation. They are, I believe, Ldmos. And I know they are push-pull because I had opened this amplifier once to see how it was designed (always good to see how other people work...).

Toonafishy: I think with the system proposed in the article from Molloy I can make a kind of home made tuner system, but in this case the tunning is done using microstrip resonators.

vfone:Thanks also for the hints.

I still have to read more carefuly the article from Molloy to see the tricks that can help me.....But it looks like that is the cheap way to go :')

Another idea would be to run the amplifier in pulsed mode to limit the dissiüated power......

Let's see......
Greetings to all!!!!!
S.
Ps:If anyone has more ideas just tell, good ideas are always welcome. :')

 

[jpsmith123]

Re: Extract parameters from very high power RF transistor. H

Hello Sinatra,

I'd like to see that article also...would you mind posting it or maybe pm it to me? Thanks.

Regards,
Joe