A couple of things here. I've used these for class AB amps, and they can be
made pseudo linear... but only by running them at reduced voltage. Like all
MOSFETs they have a negative temp. coef. The q. drain current therefore
increases as the temperature of the junction increases. What typically
happens is, at full voltage (temperature directly related to voltage)... the
heat conduction properties of the substrate overwhelm the properties of the
device case... and therefore the device is unstable... and, of course, leads
to non-linearity. By operating the device at 30-50% of the maximum voltage,
this problem is solved... at least the gate shift can be compensated. One
other thing I remember... the leakage reactance of the device is much higher
than other MOSFETs we deal with... of order 18 uh.... making input matching
a bit of a challenge for anything other than narrow bandwidths.
Sincerely,
Dr. William J. Schmidt, II K9HZ
Trustee of the North American QRO - Central Division Club - K9ZC
Email: bill@wjschmidt.com
WebPage: www.wjschmidt.com
"If computers get too powerful, we can organize them into a committee; that
will do them in." -- Bradley's Bromide
----- Original Message -----
From: "Steve Thompson" <g8gsq@ic24.net>
To: <amps@contesting.com>
Sent: Saturday, May 21, 2005 3:00 AM
Subject: Re: [Amps] ARF1505
> On Saturday 21 May 2005 04:44, Will Matney wrote:
>> What the spec sheet says is this;
>>
>> "The ARF1505 is an RF power transistor designed for off-line 300V
>> operation
>> in very high power scientific, commercial, medical and industrial RF
>> power
>> generator and amplifier applications up to 40 MHz."
>>
>> On the webpage it says they are for;
>>
>> "INDUSTRIAL, SCIENTIFIC, MEDICAL (ISM) & HF COMMUNICATIONS"
>>
>> Notice it saying amplifier applications. On the linear portion, that
>> would
>> really be where you set the bias at. It should run fine at class A or AB
>> service. Also, being for HF communications would have to be
>> linear........
>
> It might not be quite that simple. If I remember the app. note for the
> 50MHz
> 'linear', the IMD is always poor because the devices don't have a linear
> transfer characterisitic - the power in/out curve is banana shaped,
> without a
> straight line anywhere.
>
> I've also seen instability where devices intended to be driven hard into
> saturation for 'brute power' delivery are biassed for linear operation.
>
> Steve
> _______________________________________________
> 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
|