Dear Stefan, LF Group,
So what do you suggest? Using the caps in series to bring down the voltage
per cap or switching smaller values in parallel?
Well, The first priority is to ensure that the AC breakdown voltage is high
enough - the capacitors will not survive long otherwise. Practically, this
will require series capacitors to reach your required operating voltage ...
So when running 500 W RF, as i did during the tests of the last days,
about 1 kV rms was applied. Surely this caused the breakdown. So if i
design the caps for 2 kV rms, all will be fine :-) 6 A rms will be no
problem for the caps i think, rather for the variometer ;-)
If, for example, one looks at the Wima data for 1000pF, 2kV DC FKP1
capacitors, the RMS voltage limit is 700V up to about 55kHz, above which it
decreases proportional to 1/f. This implies an RMS current limit of only
0.24A. For a 10nF value, the maximum current works out to about 1.1A - the
current limit does not increase in proportion to C. So it would be quite
hard to produce a series combination with a reasonable number of capacitors
with the required current rating.
But the Wima data is based on only 10degreesC teperature rise - presumably
to allow operation at high ambient temperatures without exceeding the
100degreesC maximum operating temperature. In a reasonably well-ventilated
amateur TX circuit, the ambient temperature probably will be below 40
degrees, so perhaps 60 degrees of internal I^2R heating would be the maximum
allowable. This would allow the current rating to be increased by a factor
of about 2.4. So now the 1000pF capacitor could be operated at 0.58A
maximum - the RMS voltage at 137k would then be 680V, still within the
"rupture" limit.
For 500W out, the current into 50ohms load would be 3.2A, so 7 x 1000pf
capacitors in parallel would have an adequate current rating ... A 2 x 7
series/parallel combination should be adequate for the applied current and
voltage, and give the 3.5nF you had before.
I am not sure if capacitor manufacturers would approve of this calculation,
but practical experience shows that the capacitor current can be
considerably higher than the values implied by the data sheet graphs.
Incidentally, I believe the reason the HF RMS voltage limit of these
capacitors is always well below VDC / sqrt(2) is due to corona discharges
occuring in microscopic voids in the film, which cause the dielectric to
deteriorate over a period of time. This does not happen when a DC voltage is
applied. So the AC voltage has to be derated to a level below the onset of
corona.
Cheers, Jim Moritz
73 de M0BMU
|
