From: Peter Chadwick <Peter_Chadwick@mitel.com>
To: "'George K. Watson'" <watson@epiloglaser.com>,
"'amps'"
<amps@contesting.com>
Subject: RE: [AMPS] PIN stuff
Date sent: Fri, 4 Jun 1999 09:55:43 +0100
>
> George says;
>
> >Does anybody know what the downsides would be
> >of placing a pair of PIN diodes in parallel to increase
> >power handling capability of the circuit.
If you place the PIN diodes in parallel, place them in OPPOSITE
polarity for RF but in series for RF. That way the same bias current
will pass through the pair (or quad) of diodes.
Plus while one diode is trying to "turn off" because of reverse RF
current, the other one will be held in conduction. This reduces
harmonic distortion and increase power handling in particular at low
frequencies where carrier lifetime is a problem.
If you do this power handling goes up at least by the SQUARE of
the added diodes. If you double the diodes, the power handling at
least quadruples.
>
> Unless you increase the current, it won't help the power handling. I don't
> have the 2001 data to hand, but my feeling is that you don't have enough
> bias current. Incidentally, with PINs, it's the low frequency end that
> gets readily compromised, because you need enough stored charge in the
> junction that during the reverse half RF cycle, the diode doesn't stop
> conducting. So 2 PINs in parallel will need effectively the same current
> in each one that you have at the moment to handle twice the RF current.
That's why you put them is series for dc, and parallel opposing
directions for RF.
> I'd suggest that you go for the maximum DC through the diode that you can
> handle without exceeding the diode ratings. That will quite probably
> reduce the dissipation. If you need more, then parallel diodes for RF, but
> keep them separate for DC - you could get one diode hogging most of the
> current, otherwise. Or put a few ohms (less than 10 if you're running
> 500mA) in series with each diode, with a good low loss high current
> capacitor across the resistor. For 160, that means a something like a 0.5
> microfarad ceramic. So you're better off with separate DC feeds to each
> diode, and parallel them for RF.
A waste of parts. If you series connect them for dc while parallel
connecting them for RF (best in opposing directions for RF), you
still only need the same dc current from the supply yet power
capability increases by the square of diode numbers multiplication.
> I'm a bit dubious about the reverse volts - it seems a bit low for the
> power level you're talking about, and you might do better increasing that.
In the receive diodes, reverse volts are critical. In that case, use a
LOT of voltage if carrier lifetime is short compared to frequency and
also use them in a common cathode or anode pair or quad. That
way one or more diodes are trying to turn off while the other is
trying to recover and conduct.
Remember the peak voltage rating must include the dc PLUS peak
RF voltage.
If you hand test your own diodes, you can find many 200 volt
diodes that will handle 1 kV or more. You pay more for the
manufacturer to sort them. They all come from the same batch, the
200 volt diodes are just diodes that are 200 volts or higher, so there
are a lot of 1 kV diodes mixed in.
73, Tom W8JI
w8ji@contesting.com
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|