see below,
*********** REPLY SEPARATOR ***********
On 4/22/05 at 9:58 AM Dave Haupt wrote:
>Will,
>
>If you know anything about the Elecraft people, you
>would know that they would never "get by" with
>anything. Everything they do is top notch.
>
>It is possible to parallel FETs, but you're asking for
>a whole lot more work in the process of matching them.
> Even if your curve tracer shows that they're
>electrically matched, if they're not thermally
>matched, then once the amps starts delivering power,
>one device can begin to hog the current, and you see
>IMD rise precipitously.
>
>Directly parallelling FETs works marvelously well if
>they're on the same die, or packaged on the same
>thermal header (e.g. MRF154/MRF157), but if you care
>about IMD, it's not advisable to directly parallel
>separate FETs which may or may not have come from the
>same die lot.
I totally understand this as matched pairs, etc., have been used for years.
What I'm getting at is some published circuits I've seen doing this. What they
done was to bias each FET individually, and used matched FET's out of the same
lot. Each FET had it's own regulator if I remember. Also, I'd imagine that
these ought to be set up with some sort of circuit to prevent thermal run-away
similar to using transistors.
I would think one could parallel several FET's, and just use a tank circuit in
place of push pull too for power multiplication.
Transistors, and especially FET's for RF amps, I have to admit I do not have
too much experience with, but tubes are a different matter. One thing, what
I've read, FET's are about as close to tube operation one can get using
semiconductor devices. I'm still not sold on the idea that semis will ever
replace tubes for high power applications where using one or two devices comes
into play. What I was taught years ago about RF power transistors is that the
transistor itself is not just one transistor but many paralleled on the same
substrate. All with emitter resistors being tied together at ground. What
limited work I have done with them, I have encountered the thermal run-away
problem using MRF-454's. I have also encountered the problem of using two
devices not out of the same lot where one will surely fail prematurely due to a
mis-match in internal resistances. Plus there's the distortions created.
>
>They are using the Advanced Power Devices. Head to
>that company's webpage, and read up the app notes.
>K4XU is the chief apps engineer there, and Dick is a
>very sharp guy. Think of him as the next Helge
>Granberg, except that he's also earned his stripes in
>contesting, as well as having a great reputation in
>both the commercial (e.g. medical) amplifier field and
>broadcast amplifier field.
>
>73,
>
>Dave W8NF
>
>
>Will wrote:
>
>>Vic,
>
>>How many devices and or push pull pairs did they use
>>for the power
>>involved? I
>>do know using FET's you can get by with one push-pull
>>circuit and
>>paralell
>>FET's on both sides. This is a big advantage over
>>using transistors and
>>having
>>multiple push-pull circuits with splitters and
>>combiners.
>
>>Best,
>
>>Will
>
>
>
>
>
>__________________________________________________
>Do You Yahoo!?
>Tired of spam? Yahoo! Mail has the best spam protection around
>http://mail.yahoo.com
>_______________________________________________
>Amps mailing list
>Amps@contesting.com
>http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|