[AMPS] Biasing MOSFET Power Transistors

[Jon Ogden]

>
>There appears to be about enough parts there, but no schematic...  The
>bias circuitry is regulated, but does not appear to have any temperature
>compensation.  Don't know how important that is with MOSFETs.  I
>understand they don't run away as easily as bipolars.

Ouch.  Well, my first priority would be to trace out the schematic.  
You're probably right regarding temp compenstation on MOSFETs.
>

>
>No current limiting, other than the limits of the power supply.  I
>thought on MOSFETs what was important was bias voltage, since the
>gate current is zero.
>
True.  In a FET, voltage is the more important parameter.  
>
>
>I examined the 1KW amp that was in Feb 90 QEX.  Those are 60A
>transistors, and Helge set the idle current at 800mA.  I tried this
>ratio with my input devices, and I now get power at all input levels,
>whereas before I had to dump in about 30 watts to see any output.  I had
>to change the bias divider to stiffen the bias voltage to do this.
>Tonight I'll work on the output transistors.

Well, the higher you run your bias current, the closer you are running to 
class A.  You'll get better linearity.  However, efficiency will suffer.  
My thought would be to use just enough current so that when you apply 
your lowest RF drive level, the fets go from the cutoff to active region. 
 So if it comes out to 1.6A that works for you, then fine.  Just be 
prepared to do extra heatsinking or add a cooling fan.

That's another thing I fogot to mention.  A class AB amp is going to 
produce much more heat than a class C amp.  Depending on how large your 
heatsink is, you might need to add additional cooling (a fan) or even put 
the amp on a larger heatsink.  Many class C heatsinks I have seen would 
be totally unacceptable for class AB.
>
>   One last thing:  The large signal impedance parameters of the devices are 
>   quite possibly going to change depending on the bias conditions.  So 
>   while the matching circutis are all designed to match 50 Ohms to a class 
>   C amp, they may not be correct for matching 50 Ohms to a class AB 
>   amplifier.  You may have to retune the input and output matches.
>
>I believe you are absolutely correct here.  The amplifier is rated for
>600W output, but I'm only seeing 450W at saturation.  Probably the
>result of impedance mismatch.

Well, if you are operating the amp linearly (which is why I assume you 
want to got to AB instead of C), then you should really not be into 
saturation.  If you are over driving the amp, then yes, the gain will 
begin to go down as you increase your input.  The standard for maximum 
gain compression to use in a linear amp is 1 dB.  So check your signal 
gain with at roughly half your maximum desired input power.  Then check 
your gain at your desired maximum power.  If the gain has compressed by 
more than 1 dB, you will need to decrease the input power to a level 
where the gain has compressed 1 dB or less.  So, it's possible you are 
compressed.

Doing some calculations I just ran: 600 Watts = 57.78 dBm and 450 Watts = 
56.53 dBm.  That's only 1.25 dB difference, so you are possibly over 
driving things a little.  In order to get maximum gain you may need to 
reduce your input power.

Of course, it could just be impedance mismatches as well.
>
>The other problem I'm seeing is the feedback resistors on the MRF150s
>are dissipating way too much power.  They're going to unsolder
>themselves.  I'll have to find resistors I can heatsink.

Hmm...Possible yes.  Send me a note away from the reflector and I might 
be able to help you out.
>
>   I'll see if I can dig around in a text book of mine on solid state amps 
>   and see if I can find anything else to help you.
>
>Jon, thanks for your help and advice.

No problem.  I am not the most knowledgable out there, but I try to tell 
what I know.  Hopefully others can help you too.

73,

Jon
KE9NA



--------------------------------------------------------------------------
Jon Ogden

jono@enteract.com
www.qsl.net/ke9na

"A life lived in fear is a life half lived."


--
FAQ on WWW:               http://www.contesting.com/ampfaq.html
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com
Search:                   http://www.contesting.com/km9p/search.htm