Hello Han,
Pity the frequency used was not mentioned.
Your amp is running at 7.5/300W = 16dB power gain (also achieved in the
application note), which is way too high and could account for the instability.
In the Motorola application note for the MRF151g 50v FET, HO Granberg the
author provides a calculation for the power rating for the feedback resistors
R8 and R8 (48.3 ohms), rounded off to 50 ohms.
In the application note for the similar MRF141g 28v FET these resistors are 25
ohms and being smaller will allow more feedback. This amp runs at 13 dB gain at
150 MHz
Maybe you should try this, or else run at a lower voltage if feasible?
Also he mentions part of the gate bias circuit (for both amps - the layout is
identical) is mounted on the output PCB (!) which might allow RF into the gate
bias circuit - try additional decoupling and bypassing.
He also mentions resistive and/or inductive loading of the gates to reduce gain.
If the amp is mounted in a metal box poor layout might also cause instability -
try gluing graphite sponge material inside the box.
A possibility on the harmonic imbalance problem, an unpleasant one, is that one
FET in the dual package is damaged and not achieving equal gain with the other.
However, the MRF141g application note has harmonic radiation curves (Fig 5) and
the 3rd harmonic is only around 12 dB down (power 19W for 300 W output) below
70 MHz while the 2nd harmonic is worst case 38dB down.
So the LPF needs to handle 20W - thanks for pointing this out.
73,
Ian ZS6BTE
Refs:
AR305 for the MRF151g
AR313 for the MRF141g
________________________________
From: Han Higasa <higasa@plum.ocn.ne.jp>
To: amps@contesting.com
Sent: Monday, August 22, 2011 11:24 AM
Subject: Re: [Amps] MRF141g amp kit offered by CCI
hi Ian
I have bought one MRF151G kit from CCI and made
experiments for long. There are some tips to tell.
It has too much power even on 2nd and 3rd harmonics
such as -16dBc (7.5W @300W out).
You need an absorption type LPF to address these
high power harmonics. Normal pi-LPFs can burn.
It is because of high-gain on unnecessary high band
and it can not be suppressed by feed-back resistors.
Ballancing two FETs in a gemini package by
separate gate bias circuits is better to lower these
harmonics.
And, maybe it is a fatal defect of the kit - input and
output board are separated and the FET tend to oscillate.
You need to short these boards with two pieces of
cupper ribbon near the FETs.
It may be the same on MRF141G kit and you need very
careful experiments to make it work on your desired
band - GL.
de Han JE1BMJ.
_______________________________________________
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
|