[Amps] Response the "QSK switching time"

[Alex Eban]

There's one more reason why there's need of such a high voltage, on the
order of 1000VDC on the diodes and not only on the transmit diode!
When back biased a diode exhibits a varactor effect: it acts as a
capacitance, whose magnitude is inversely proportional with the applied
voltage. Even a few Pico farads in series with the receive path can have a
disastrous effect on the receiver. Sooo the receive path has to be back
biased by high voltage too!
This characteristic is exhibited even by the very highest quality PIN
diodes, like those manufactured by Chelton especially for the purpose. I
know: I used them when I was working, designing multikilowatt amplifiers for
the army.
BTW the bias supply should be quite beefy: the current transients of
charging and discharging this inherent capacitance are quite respectable so
that when working QSK, the dynamic regulation of the bias supply had better
be good, and the switching transistors also should be able to handle pulses
in the ampere range, albeit brief in duration.
Alex	4Z5KS

-----Original Message-----
From: amps-bounces at contesting.com [mailto:amps-bounces at contesting.com] On
Behalf Of Joe Subich, W4TV
Sent: Saturday, December 19, 2009 7:43 AM
To: 'Paul Christensen'; 'Amp Reflector'
Subject: Re: [Amps] Response the "QSK switching time"



> The tick is caused by the moderately-high switching voltage being 
> applied to the T/R diodes.  I haven't taken a good look at that part 
> of the schematic, but I'm curious as to why such high voltage (400V) 
> is needed for switching?

The 400 Volts is used to baas off (reverse bias) the receive path (antenna
to transceiver) during transmit.  In a 50 Ohm system at 1500W the RF voltage
is approximately 275 Volts
(RMS) or about 385V peak.  A back bias of less than 400V will allow the
diodes to conduct during part of the RF cycle and will result in feedback
which can turn the amplifier into an oscillator, destroy finals in the
transceiver or many other nasty things. 

As pointed out by K5JV just a couple days ago, 400V is really too low to
survive system voltages at 1500 W with at SWR levels above 1.5:1 in SSB or
any SWR in constant carrier modes like RTTY.  This is one reason that the
QSK-5 is rated as it is. 
This lack of understanding is one reason the QSK-5 and old DEO QSK-1500 have
a bad reputation, particularly when operated with 8877 based amplifiers with
"all knobs to the right." 

Lon is correct, to be reliable  the diode QSK units really need to use 1000V
reverse bias for the "transceiver to antenna" leg of the switch in order to
prevent breakdown under the combination of high power and high SWR.  By the
way, the 400V reverse bias is reasonable for a 1000W amplifier like the SPE
1KFA or a single 3-500Z and the diode T/R switch is likely to survive
anything the amplifier can deliver into a 2:1 SWR. 

73, 

   ... Joe, W4TV 
   
   


> -----Original Message-----
> From: amps-bounces at contesting.com
> [mailto:amps-bounces at contesting.com] On Behalf Of Paul Christensen
> Sent: Friday, December 18, 2009 7:16 PM
> To: Amp Reflector
> Subject: Re: [Amps] Response the "QSK switching time"
> 
> 
> > I have to comment on Carl's thread.  I have not been able to get 
> > through to them lately, either.  My bone with them is that their 
> > Quality Control is awful, almost non-existent.  Since I
> have started
> > messing with QSK, I have bought two QSK-5, stand alone units, from 
> > them.
> 
> I recently purchased a used QSK-5 to use with an SPE 
> solid-state amp.   The 
> electrical concept of the QSK-5 is quite good although mechanically, 
> it can use some improvements.
> 
> The low-level SO-239 panel connectors use the aluminum cabinet 
> exclusively for the RF return.  I detest construction practices where 
> only the case and hardware are used to complete return paths.  Same 
> issue with their RCA keying connectors.  The panel is used for the DC 
> returns.  If the KEY 2 jack becomes slightly loose -- and looses 
> contact against the cabinet, the QSK-5 will hot-switch - guaranteed.
> 
> To deal with these issues, I installed toothed solder lugs on the 
> SO-239 and RCA jacks.  On the RF connectors, I then brought individual 
> leads from each jack directly back to circuit ground.  On the RCA 
> connectors, I grouped the returns together with #20 buss bar back to 
> circuit ground.
> At least now,
> the QSK-5connections are solid.
> 
> Finally, I replaced the line cord with a detachable IEC type with 
> combination RFI/EMI filter.  I really hate having to find the end of 
> an AC cable when it's time to perform some maintenance on a piece of 
> equipment.
> The IEC cord makes it easy to pull the unit from the desktop.
> 
> I did find it odd that the QSK-5 makes a "ticking" sound when keying - 
> with or without RF applied.  Although a frame signal relay is used as 
> part of the unit's bypass system, the relays are not the root cause.  
> The tick is caused by the moderately-high switching voltage being 
> applied to the T/R diodes.  I haven't taken a good look at that part 
> of the schematic, but I'm curious as to why such high voltage (400V) 
> is needed for switching?
> Fifteen years ago,
> I designed a PIN diode T/R switch using Microsemi UM2110 types and the 
> diodes are fully biased with low voltage and moderate levels of 
> forward current.  The UM2110 is optimized for HF switching and easily 
> handles 1.5KW with a reverse voltage rating of 1KV.
> 
> Paul, W9AC
> 
>  
> 
> _______________________________________________
> Amps mailing list
> Amps at contesting.com http://lists.contesting.com/mailman/listinfo/amps

_______________________________________________
Amps mailing list
Amps at contesting.com
http://lists.contesting.com/mailman/listinfo/amps