Still struggling to make a "simple" 5Watt PA for the 1.8 - 50MHz range
The QRP-LABS IRF510 5W PA  will not deliver more then 1W at 28MHz
not to speak of 50 MHz there it is an attenuator.
The U3S LPF needs to be terminated with real load to keep the signal a bit decent.
Using the RD15HVF1 Mitsubishi did bring just a little more.
3-4 W at 28  and 1.5 W at 50MHz
The QRP-LABS PCB has extra pads for the  RD15HVF1 that have crossed drain and source.
The RD15HVF1  source the middle pin and cooling flange direct screwed a heathsink.
The cheap eBay RD15HVF1  are fake  IRF510  stamped as RD15HVF1
The real RD15  has low Gate capacity  approx 83pF   while the IRF510 has 400pF or more and the source and drain are crossed.
eBay is poisoned with  fake transistor sellers.
Like  5 pcs RD15HVF1 5$  good quality ...

The specs of the ( REAL) RD15 look perfect, too good  to be true!
FEATURES High power and High Gain:
Pout>15W, Gp>14dB @Vds=12.5V,f=175MHz
High Efficiency: 60%typ. on VHF Band
Integrated gate protection diode.

So it must be a piece of cake to make 5W at 28 and 50MHz  ...

But  i have been trying and struggling  long now with very poor results.
Yes got 5W at 28MHz  at 13.5V   but not for long
Sudden the output level began to move and dropped to zero the RD15  was dead.
Overheated ? it was not extreme hot dc power input was normal  0.8A 12.5V
to high drive ?  bad thermal joint to heathsink  could be.
Never reached more then 5W at 28  an 1.5W at 50MHz
  even when self oscillating the output never exceeded 3Watts
A bad sign, since that's what most amplifiers are good in.

When leaving out  the power envelope modulator ment for CW key in off smoothing.
 A simple PA 4 part circuit remains. 
The RD15  driven over 100nF at the gate
Output is taken from a 1:9  transformer via 100nF
Bias voltage at the gate with a pot-meter
12V  power is applied over drain impedance transformator.
The PCB is old school FET and heathsink mounted vertical.
Picture shows a PCB enabling the RD15 to be screwed convenient to the bottom of a housing as heath sink.
I tested several input and output 1:9 transformers and power regulation at the input.
but  with the gate drive voltage 10V tt at 28MHz the output is near zero  no gain!
failed  to reach the easy 5W level at 12.5V
Is my RD15   fake too  or is this device only usable for professional RF designers?

My latest test i found that the U3S output was dropped to 20mW
After resetting and re-configuring, the 28MHz output was at normal level 200mW
200mW !  a very reasonable  drive level I would say.
And then the RD15 PA  with some attenuation  and 56 OHms termination over the gate and a 100 OHm  potentiometer  the output reached 3.5W , increasing drive did not bring more.
 reasonable gain and output at 12.5V  but way below specs.
3.5W would be acceptable, but placed in the U3S and followed by the LPF board only 1.5W was left
This output drop  could be because of the very high IMD products from this U3S + PA
 A spectrum analyser view would make that clear. 

My attempts resulted in 3 destroyed RD15HVF1
It did run a month at 3W  28MHz   1W at 50MHz using the original 5W QRPLABS PCB
This is not what i had in mind for a 5W PA
At frequencies  10MHz  and below live is good  8W to 10W out.
It seems most U3S  users are 14 MHz  or below since no-one is complaining?
The RD15HVF1 should need only 30mW drive for 10W at 175MHz
But U3S output level can be as high as 500mW  as in my first build U3 the BS170 can bring 800mW @5V !
Far to high level for the RD15 .

b.t.w.  Elecraft K3 uses a pair of RD15HVF1  in its  LP-PA   8W 6m   12W HF
Had to replace 2 times a broken RD15HVF1   so how reliable is this RD15HVF1 ?

Any help or hints for "simple 5W at HF + 6m" are more then welcome !
please help me  hihi

the QRP-LABS PA with optional power envelope modulator

Only the part below is used as PA where Q9 idle signal blocker can be a bridge
  as long as the 12V is switched on  TX only using the U3S key signal to a IRF9540 P-FET



PA2RF said…
Misschien zonder IRF, maar met gewone transistoren proberen:


Weet niet wat er op 28 / 50 MHz gebeurt, maar misschien het proberen waard.

Jaap van oosten said…
Ron bedankt voor reageren
is zeker het proberen waard
Probleem bij de IRF510 types is juist op de hogere frequenties
op 80 160 40 30 gaat het prima
maar het idee om gewone transistoren te gebruiken is erg aantrekkelijk
en waarom niet, een IRF510 is ook niet voor RF ontworpen hoewel de naam dat wel insinueert

PA2RF said…
Ja precies,
Eigenlijk zijn dit soort IRF mosfets ontworpen for LF schakeldoeleinden. IRF is de naam van de fabrikant en heeft niks met RF te maken :-)
Heb zelf gemengde ervaringen met IRF in de eindtrap. Soms ging het uitstekend op HF maar niet altijd. Bias erg kritisch en niet vergevingsgezind.
De 'ouderwetse ' transistoren (2N2219 ofzo) halen met gemak 50 MHz.
Succes met JBOT, Just a Bunch of Transistors.
73 Ron
Jaap van oosten said…
op 630M 475kHz daar werkt het nu goed met de IRFP4227 100W RF out 110W DC

fool proof zonder antenne blijft het heel

tot nu toe slechts.. 1 stuk door fout

de IRF540 die ik eerst gebruikte sneuvelden allemaal soms direct soms na 1 week..

een slagveld van IRF540 - IRL540 en andere versies

Franta Město said…
Hello, Please, How You Wrapped That Output Transformer to be 1: 9 I would like to test this design as a driver for my final stage. Thank you for your explanation. 73! František OK2WF