Amps
[Top] [All Lists]

Re: [Amps] Checking Power FET's (Mike & Becca Krzystyniak)

To: amps@contesting.com
Subject: Re: [Amps] Checking Power FET's (Mike & Becca Krzystyniak)
From: "Dave" <dave@g8kbv.demon.co.uk>
Date: Sat, 28 Jan 2012 21:03:10 -0000
List-post: <amps@contesting.com">mailto:amps@contesting.com>
Hi.

The amp should idealy have some indication of fault status, possibly over 
current, or under current (with no RF).   In an ideal world.

Quick and simple way, so long as you can get physical access to them, not 
even needing test gear.

With the input and output terminated (the output term' should be good for 
the full rated power CW if possible at the highest frequency the amp is 
designed for.)  Power it up into TX mode wait a minute, power down and 
feel for warmth on each FET.   They should be noticably warm, but you'll 
have to be quick if the heatsink is good.

Else, with a DMM, again no RF and in/out term'd.  Power up, and measure 
the Gate volts.   Each device should have about the same bias, but 
probably not all the same.   One that is bad, will have it's gate stuck 
at ground, or more usually VDD.

In some real bad designs, there is only one bias source, and if one 
device fails, it kills all the others too by yanking the overall bias up 
somewhat, couple that with no Drain current monitoring and protection.  
Not good.

Look for burnt devices, other components etc.  Any feedback components 
that look stressed, are a sign of poor design, or overdriving for some 
time.

FET PA's are not complicated, but their support circuitry can have 
mulitple functions.   Parasitics shold not occur in any amp of course 
(tube, transistor or FET) but "things do happen" especially if the load 
was very reactive with narrow band FET amps.

Find the service manual for the amp, even google the FET makers part 
number to get typical opreating conditions, and application notes etc.

Note, that unlike tube amps, a 1000W solid state amp, will only do 1000W 
CW or PEP maximum, there is no overdrive capability, at best you'll push 
them into class C, at worst POP!   Assuming a poor PSU design and no real 
time drain current monitoring and protection.   Normally, you'd under run 
them to keep a clean signal, unless there is envelope pre-distortion as 
in some real high end commercial/military com's kit.

Remember too, Low Voltage, but High Current PSU's can be just as 
dangerous as High Voltage ones when things go wrong and an awful lot of 
energy can be disipated in a small space in a short time, making flames 
etc.

If you get to identify the failed part(s) and replace them.  Take great 
care with the mounting, make sure the mating surfaces between device and 
'sink are surgicaly clean, and use the bare minimum of heatsink compound 
even spread over the entire device's mating surface, as a very thin film.  

Follow the device makers recommendations re torquing down the fixings.
Not enough, bad thermal contact.  Too much also bad thermal contact (many 
devices distort and bow upwards in the middle, result, early failure.)

Also, provide a small kink ~ in the flat tab/leads, so that thermal 
expansion etc will not unduely stress the solder joints.  Another cause 
of eary failure in some solid state amps.

Lastly, cooling is parramount.  You do not want the things to run "Hot", 
keep 'em cool for a long life.  The cooler exhaust air should not be much 
above what the inlet air is, then all the heatsink will be working.  Keep 
vents, filters, fans etc clean.

I work on commercial solid state amps from 50W up to many kW out.  You 
wouldn't believe the failure modes that can happen.  Most due to poor 
workmanship in manufacture, or doubtfull design in the PSU and controls.

Like tube amps.  Making a high power amp is relatively easy.  Making a 
good safe and relaible one, takes some effort, and a serious ammount of 
understanding, of not just the RF aspects...   If anything, those 
servicing and repairing them need the same depth of knowledge a lot of 
the time.

Take care.

Dave G0WBX.
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps

<Prev in Thread] Current Thread [Next in Thread>
  • Re: [Amps] Checking Power FET's (Mike & Becca Krzystyniak), Dave <=