To: rsgb_lf_group@blacksheep.org
References: <4ffc7958-810a-e4d2-7745-258a58338934@n1bug.com>
 <5A3AE81B.9010205@posteo.de> <bad0d594-0b94-7b2d-2ada-eeb60b2eb0af@n1bug.com>
 <5A3B8FB1.1020403@posteo.de> <7ee833a8-04b4-4712-d927-7eccf37eec14@n1bug.com>
 <5A3D9BC1.1040609@posteo.de>
From: N1BUG <paul@n1bug.com>
Message-ID: <6fa587ef-d5b0-6a67-b604-ba94144444f7@n1bug.com>
Date: Sat, 23 Dec 2017 08:13:47 -0500
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101
 Thunderbird/52.5.0
MIME-Version: 1.0
In-Reply-To: <5A3D9BC1.1040609@posteo.de>
Content-Language: en-US
Subject: Re: LF: TXing 2200m WSPR
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org

Hi Stefan,

> The C on the gate is not a good idea i think. Better try a 10...22 Ohm
> resitor or a ferrite bead in front of the gate.

I removed the C on the gate and added a 15 ohm resistor in front of 
the gate (in series with RF input to the gate). No change in 
performance.

> Is the bias voltage stable? It may be worth to spend 1000 uF parallel to
> the Z diode. And a normal diode in front of the z diode, i.e. into the
> yellow wire...

I didn't have any 1000 uF but I put 470 uF in parallel to the Z 
diode, added a 1N4007 in series at the end of the yellow wire. No 
change in performance.

At this point I became curious about the 1 uF C bypassing the +V 
lead of the blue core to ground. It seemed strange to have it 
bypassed with 1 uF there, then a series choke, then 2200 uF bypass. 
I could not find any published circuits like that.

As a test (not having any idea what I was doing!) I removed the 1 uF 
bypass. The amp went completely crazy.

So I decided to try some more C in parallel to the 1 uF. All I have 
are electrolytic caps, but I tried 10 uF. Stable output increased 
from 12W to 16W.

What about even more C? I removed the 10 uF and put 100 uF in 
parallel to the 1 uF. No change from the 10 uF.

It still does very strange things if I try to adjust bias for more 
than 16W.

I don't know what this experiment tells me...

> The blue core, maybe it saturates at higher power? Could be worth to try
> 2 more turns and better use abt 0.5 mm wire for higher power.

I will try that next if I find some suitable wire here. I also 
thought about trying a 10 turn bifilar winding instead of the 
trifilar. Again I have no idea what I am doing, but the worst that 
can happen is another dead FET. :) I have more.

> But the design is not good for > 30 W or so. Don't expect to much from
> an IRF530 at 13.8 V.

I would be happy if I could get 25W stable from this amp. I can get 
25W but it is not stable at that level.

73,
Paul N1BUG