开云体育

ctrl + shift + ? for shortcuts
© 2025 开云体育

10W HF Linear PA kit


 

The 10W HF Linear PA kit from QRP-Labs was soldered together. It is a revision 2.00.

Here are my measurements on my finished copy. Supply volage 13.8Vdc

Below you see the

Gain (measured with a 30dB attenuator), Input impedance and? Output impedance.

?

73

Guido? on7ch

Gain

Input impedance



Output Impedance


 

I built one and after operating successfully on the bench for several days it started self-oscillating at high power and heated up the 10 meter LPF, even with the exciter turned off.? I'm still scratching my head over that one.


 

Hi Jerry,

I have built two of these amplifiers and they work flawlessly.
The only thing I can think of is a possible bad contact due to a cold solder joint.

Check the bias currents of the IRF510.

73

Guido? on7ch


 

I built 3 of V1 type 10W linear amplifier to use with ?my own design carrier board and am very happy with it. I am still using them for my sota/vota activities.Great product!



--
?

73, Barb


 

I'm on my 5th build.? All work and the latest partial SMT and works the same as the others.
My curves are different than Guidos.? I never measure output impedance,? However the?
gain I get is flatter to 30mhz.? I used a 50mW into an attenuator so the source sees
a flatter load.?

Fixture line up used was a Rigol815T, Qbit block amp, step attenuator(HP set to 10db),
DUT(the amp), 30DB Narda?power attenuator, Kaye step attenuator to Rigol input.??

The Qbit amp is a class A video amp flat from .5Mhz to 100Mhz with max undistorted
output of 27dbm.??

Test is done at 9W to avoid saturation . for the amps I have and built 1db compression?
varies from 12W to 13.5W at 13.8V at it terminals.?

The amp will oscillate if the wiring is poor that being th einput coax and output coax
to LPF and load.? Your dealing with 26-28db of gain.

--
Allison
------------------
Post online only,?
direct email will go to a bit bucket.


 

?

In this URL you can find the method used to measure the output impedance.




 

I'm aware of that.? Have done it as well.

Also it depends on bias setting and the winding of the transformer.
By that I mean how tight and careful both windings are done.

I look at it as interesting but, what use is it beyond diagnostic
if ian amp is not developing full power as then the transformer can
be ruled out or not.

I have done the reverse on a board without finals loaded and the
Drain to Source path with R/C to simulate active devices to test
if the transformation of the transformer is valid.? I use that more
at VHF and up amp designs.? Same for input transformer or
matching save for the simulated load will match the gate to source.

As to input impedance, its nice to know but it was never meant to
be 50 ohms.? Again it can check the transformer.? FYI the QSX design
the driver was power limited but not voltage swing limited so higher
input impedance was needed.? That was the design source (QSX).

The test fixture I use verifies the assembly is flat and has known?
(measured loss) as that is a full two port test then the DUT(amp or other)
is added.? Since it build it up and measure the rare case is I find a
bad attenuator or cable early in the process.

--
Allison
------------------
Post online only,?
direct email will go to a bit bucket.


 

I agree with your approach regarding the input and output impedance measurements.

?

During the construction and measurements, I must admit that this 10W PA is a well-thought-out design

Wel done Hans.

Other designs on 2 x IRF510 perform a lot worse.

?

?


 

Guido,

There are many designs that are notably bad... and even more copies of them.

There is an Italian design that also works well.? Secrets are layout that treats
the IRF510 as a VHF RF device. Those that don't tend to have VHF parasitics
or outright oscillate.? Device lead lengths are also critical.? Feedback is also
critical.? The operating voltage is also important.? At 13.8 the device is actually
voltage starved so clean power is limited, go to 24V and performance becomes
far better.

Same issue as using 28V D-mos power mosfets at 13V and also bipolar
devices?behave this way.??

--
Allison
------------------
Post online only,?
direct email will go to a bit bucket.