Message-ID: <4E45B929.70802@iup.uni-heidelberg.de>
Date: Sat, 13 Aug 2011 01:37:13 +0200
From: =?UTF-8?B?U3RlZmFuIFNjaMOkZmVy?= <Stefan.Schaefer@iup.uni-heidelberg.de>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; de; rv:1.9.1.8) Gecko/20100227 Thunderbird/3.0.3
MIME-Version: 1.0
To: rsgb_lf_group@blacksheep.org
References: <4E4527EF.2000304@iup.uni-heidelberg.de> <52970A62D1224CAF9A966D1743DCBF00@PcMinto> <4E456AB7.6090403@iup.uni-heidelberg.de> <4FB899125601413A9077646E2728C784@PcMinto> <4E45A678.5010908@iup.uni-heidelberg.de> <4E45B0F9.1000007@iup.uni-heidelberg.de> <002101cc5947$0a6ce300$4001a8c0@lark>
In-Reply-To: <002101cc5947$0a6ce300$4001a8c0@lark>
Subject: Re: LF: Re: Faulty FKP-1
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: quoted-printable
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
X-AOL-SCOLL-URL_COUNT: 0  
x-aol-sid: 3039ac1d60ca4e45ba2951c6
X-AOL-IP: 195.171.43.25
X-AOL-SPF: domain : blacksheep.org SPF : none


Hi Alan,

Yes i know, but these caps perform very well in a my PAs so far. Have=20
never had a problem with them. The are rated at 2 kV DC and 700 V AC. I=20
alwyays ignored the 700V AC level. But probably the voltages in the=20
series L-C were much higher than that, probably even higher than 2 kV=20
rms. I will do measurements at lower power soon. So then, the solution=20
would be to make series parallel combinations, e.g. 3x 10 nF in series...
Am 13.08.2011 01:24, schrieb Alan Melia:
> Hi Stefan, The voltage rating alone is not enough. The capacitors in th=
e
> Class E stages must be "pluse rated" They have to pass massive amounts =
of
> current. I tried originally to use old valve caps but they are not rate=
d for
> that level of current and the connection of the individual foil plates =
to
> the lead-out wires kept fusing and the capacitance went down and down. =
I
> would have though the WIMA caps were intended for SMPSUs so should be
> suitably rated.
>   =20
Yes, they are excellent for impulse issues. I had some switching=20
impulses in the H bridge when operating without a load. A 33 nF cap=20
directly between the upper drain and the lower source was a nearby=20
perfect solution. Before i observed up to 3x UDC peaks.
> Because it is in the antenna circuit is it possible it might have been
> damaged by  a nearby lighning surge??
>   =20
No, almost impossible. Even in the case of a direct lightning strike i=20
expect the would survive, in that "position" where they are in the=20
circuit. I have a lightning spark gap across the isolator which has=20
about 4 cm distance. So just a short impulse would be applied to the=20
coil until it gets shortcutted by the gap. The voltage in that=20
arrangement, dorwn transformed to 50 Ohm wouldn't be high enough i assume.

But an interesting point!

73, Stefan

> Alan G3NYK
>
> ----- Original Message -----
> From: "Stefan Sch=C3=A4fer"<Stefan.Schaefer@iup.uni-heidelberg.de>
> To:<rsgb_lf_group@blacksheep.org>
> Sent: Saturday, August 13, 2011 12:02 AM
> Subject: LF: Faulty FKP-1
>
>
> Minto, LF,
>
> Maybe interesting for other homebrewers in the QRO region as well...
>
> Actually the only thing that caused all the faults was a damaged WIMA
> FKP-1-2 kV cap. I never have had problems with these caps before. Never
> expected that. They are even suitable as a shunt cap for a 10 MHz 300W
> class E PA or as the shunt cap in a 1 kW 137 kHz class E... But now one
> looks quiet big here. I even heard it dying but thought the sound is
> coming from the FETs, making some noise when switching, you know. Thus
> the C was much lower and i didn't come on resonance. Probably the
> undefined C (sometimes even an inner shortcut?) caused strong mismatch
> and lastly destroyed the FETs.
>
> That was the weakest link of the chain. QRO optimisations... The C was
> 0.33 nF + 1 nF + 2.2 nF in parallel, this in series to the small
> variometer part in the shack. It allows resonating within the entire ba=
nd.
>
> Now replaced, all works well again...
>
> 73, Stefan/DK7FC
>
>
> Am 13.08.2011 00:17, schrieb Stefan Sch=C3=A4fer:
>   =20
>> Hi Minto.
>>
>> The signal must have been about 6 dB lower than yesterday. Still have
>> to go back to the old system, unfortunately. Difficult to do good
>> measurements here in the "shack" where all the stuff is floating
>> either on mains voltage or on 137 kHz or a mixture of that. Have a
>> battery supplied oscilloscope but anyway.
>>
>> I didn't got more than that what you saw, no higher power was possible
>> even when running the SMPS at 320 VDC. Unfortunately now i saw what
>> the reason was/is: Some of the turns of the loading coil moved a bit
>> and so i was outside the resonance. Since  the Q of the TX antenna is
>> very high, the power goes down quickly when beeing outside the
>> resonance. I haven't had the tuning-meter (U,I, phase) in line, and
>> just tuned the shack-internal variometer for a maximum signal.
>> After the unwanted fault and replacing the FET i had to find other
>> problems like a non conducting shunt resistor (current sensing). This
>> wasn't visible. For some reason the current limiting didn't work then.
>> A second loud BAAAAAAANG caused me to *let it be* for this week! Now
>> two mains fuses here at QRL are OFF and i don't have the key to put
>> them in again.
>> Anyway, i'm in holiday for the next 2 weeks and they will be back in
>> when i am back and no one will remember that, hopefully, hi.
>>
>> Actually all that stuff should be exchanges in a QSO :-) BTW how about
>> YOUR PA? :-)
>>
>> 73, Stefan/DK7FC
>>
>> Now trying to bring the coil back to its previous L...
>>
>>
>> Am 12.08.2011 23:07, schrieb Minto Witteveen:
>>     =20
>>> Hi Stefan.
>>> I see your CQ on my grabber at this moment.
>>> How does your ERP now compare with yesterday (during your QSO with Ma=
l)?
>>> It looks like S/N is now worse here then yesterday. Reason might be
>>> the loop versus the miniwhip (with the miniwhip the winner). But on
>>> the other hand there seems to be a lot of QRN (because of lightning
>>> at this moment.
>>> 73=E2=80=99s Minto pa3bca
>>> ---------------------------------------------------------------------=
----
>>>       =20
> -
>   =20
>>> Ceterum censeo Carthaginem delendam esse
>>> *From:* Stefan Sch=C3=A4fer<mailto:Stefan.Schaefer@iup.uni-heidelberg=
.de>
>>> *Sent:* Friday, August 12, 2011 20:02
>>> *To:* rsgb_lf_group@blacksheep.org<mailto:rsgb_lf_group@blacksheep.or=
g>
>>> *Subject:* Re: LF: PA3BCA
>>> Hi Minto,
>>>
>>> Am 12.08.2011 19:42, schrieb Minto Witteveen:
>>>       =20
>>>> How much power were you generating?
>>>>
>>>> 73=E2=80=99s, Minto pa3bca
>>>>         =20
>>> It changes all the time. Not even sure since i just have an E field
>>> probe to decouple some power from my TX antenna and feed it to a 50
>>> uA amperemeter, shunted by a potentiometer. This poti was adjusted to
>>> show 100% full scale when using 120 W RF before some weeks. Now
>>> reduced the parallel R on the poti so it shows just 25 uA then. Got
>>> 50 uA anyway by QRO and more (not sure). So it should be arround
>>> 500W. This is -6 dBfs :-)
>>>
>>> But the antenna, i.e. the coil and its cover is very lossy it seems.
>>> When there was rain, i have up to twice the losses :-( Need to search
>>> for an alternative cover (mustn't burn, so a water barrel is no
>>> alternative...).
>>>
>>>
>>> 73, Stefan/DK7FC
>>>       =20
>   =20