From: "Clemens Paul" <cpaul@gmx.net>
To: <rsgb_lf_group@blacksheep.org>
References: <1608d073056.marcocadeddu@tin.it> <d877cf2d-4e75-b372-9bd2-302791f416b0@n1bug.com> <4556666E6F7849679451F95BD8C43FAA@Clemens0811> <e0a831c8-5a4b-1734-335e-58d995fa004a@n1bug.com>
In-Reply-To: <e0a831c8-5a4b-1734-335e-58d995fa004a@n1bug.com>
Date: Thu, 28 Dec 2017 22:33:42 +0100
Message-ID: <6F51339F764245AA8F6F22951FB90EFD@Clemens0811>
MIME-Version: 1.0
Thread-Index: AdOAGXnO3Y4hjvZVQhahuJONIaIG+AAAxyOw
Subject: RE: R: Re: LF: TXing 2200m WSPR
Content-Type: text/plain;
	charset="iso-8859-1"
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Content-Transfer-Encoding: 8bit

Hi Paul,

>My Ultimate 3S does have a LPF before the feed to the PA. I 
>understand some PAs prefer a square wave. Should I remove the LPF?

No!
It helps reducing harmonics output of the PA which in case are reflected
by the LPF to the FET and MAY cause instability and IMD.
That's why *sometimes* diplexer filters are used.
They swamp harmonics in the 50 Ohm resistor of the parallel HPF instead of reflecting them.
 
>I understand some PAs prefer a square wave. 

These are push pull class D or E PAs.
Yours is single ended and Class AB,B or C depending on bias.

>When I isolate the 
>grounds, the problem is worse! What does this tell us?

This supports my assumption that your problem is caused
rather by the RF field than by the LPF.
It would be also a good idea to put the Ultimate and even more the PA in shielded metal boxes.

>Could I have some mistakes in my isolation transformer? I have 
>primary turns spread around the whole core, with the secondary turns 
>around almost the whole core placed between primary turns. But the 
>primary and secondary windings do not start at the same position on 
>the core. I thought this wouldn't matter. Maybe it does?

No,it doesn't, even on HF.

>I have some FT-144-JC cores. Any chance these could be 
>used for a test?

I guess there is a typo:
Size 144 does not exist,either it's 114 or 140.
J material is very fine for this purpose,JC is the same as J but with coating.
If you can manage to wind *at least* 10 turns on one core
- more turns and/or more cores would be better - with thin coax you'r in business.
The minimum goal is to achieve a choking impedance of 1K which would
reduce common mode currents on the coax shield by 20dB.

73
Clemens
DL4RAJ  


>-----Original Message-----
>From: owner-rsgb_lf_group@blacksheep.org 
>[mailto:owner-rsgb_lf_group@blacksheep.org] On Behalf Of N1BUG
>Sent: Thursday, December 28, 2017 9:20 PM
>To: rsgb_lf_group@blacksheep.org
>Subject: Re: R: Re: LF: TXing 2200m WSPR
>
>Hi Clemens,
>
>Thank you for taking the time to run these simulations!
>
>My Ultimate 3S does have a LPF before the feed to the PA. I 
>understand some PAs prefer a square wave. Should I remove the LPF? I 
>have a function generator which can make the same power as the 
>Ultimate 3S with either sine or square wave. I could try that as a 
>test...
>
>Today I made some more tests. I can run a little more power before 
>trouble begins when I have the primary (coax) and secondary 
>(antenna) grounds tied together at the antenna! When I isolate the 
>grounds, the problem is worse! What does this tell us?
>
>I'm not sure how to make an effective common mode choke for this 
>frequency. I can try another transformer at the shack if I can find 
>a low power version that uses an inexpensive core. At this point I 
>don't want to buy 3 x FT-240-77 cores for a <20W test. Any 
>suggestions? I have some FT-144-JC cores. Any chance these could be 
>used for a test?
>
>Could I have some mistakes in my isolation transformer? I have 
>primary turns spread around the whole core, with the secondary turns 
>around almost the whole core placed between primary turns. But the 
>primary and secondary windings do not start at the same position on 
>the core. I thought this wouldn't matter. Maybe it does?
>
>73,
>Paul
>
>
>
>On 12/28/2017 08:00 AM, Clemens Paul wrote:
>> Hi Paul,
>> 
>>> I already have plans to build the big T type LPF.
>> 
>> Out of curiosity I've done some spice simulations of your PA.
>> They indicate that a T type LPF won't help at all.
>> This was to be expected because your PA is not a class D voltage mode
>> push pull PA like W1VD's or IK1HSS's or 2E0ILY 's.
>> Indeed the T type LPF will cause considerably higher voltage 
>spikes at the FET'S drain.
>> In the attached screenshot you can see three 
>simulations,each with a different output filter.
>> A shunt C LPF,a series L LPF and even a diplexer filter.
>> The red traces represent voltage at the drain,green is DS current.
>> With the diplexer voltage spikes are lowest.
>> But I obviously the filter is not your problem.
>> It seems that the near field of the antenna is causing the trouble.
>> I would suggest to try a second isolation transformer or a good
>> common mode choke *at the shack* end of your coax run 
>instead of building a T-type LPF.
>> 
>> P.S. Has your Ultimate 3S a LPF installed before you feed it 
>into the PA?
>> 
>> 73
>> Clemens
>> DL4RAJ
>> 
>>> -----Original Message-----
>>> From: owner-rsgb_lf_group@blacksheep.org
>>> [mailto:owner-rsgb_lf_group@blacksheep.org] On Behalf Of N1BUG
>>> Sent: Monday, December 25, 2017 1:15 PM
>>> To: rsgb_lf_group@blacksheep.org
>>> Subject: Re: R: Re: LF: TXing 2200m WSPR
>>>
>>> Hi Marco,
>>>
>>> I don't detect any heat in the LPF but at this low power level it
>>> could be difficult to detect it.
>>>
>>> I already have plans to build the big T type LPF. I have the
>>> T-225A-2 cores and metal to make a compartmented box. Capacitors are
>>> on the way, could arrive around 29 December. I still need to order
>>> the wire and some kind of material for mounting the big inductors.
>>>
>>> I'm making that LPF because eventually I will also build the big
>>> amplifier. I'm not sure when that will happen, probably a bit late
>>> for this LF season.
>>>
>>> The most curious thing is what seems to be some measurement error or
>>> impedance transformation but the cause remains elusive. I hope to
>>> find the reason so I can be more confident about measurements when I
>>> go go high power.
>>>
>>> 73,
>>> Paul N1BUG
>
>
>---
>Diese E-Mail wurde von AVG auf Viren geprüft.
>http://www.avg.com
>