[ARC5] RCA AVR / AVT Set - Vibrator Spike Suppression

[Ronnie Hull]

You been letting Moquisha work on your radio's again?

Sent from Ronnie's IPhone

On Dec 12, 2012, at 9:45 AM, "David Stinson" <arc5 at ix.netcom.com> wrote:

> Been working on the "spikey" AVA-126A vibrator power supply for the RCA AVT / AVR series of light aircraft radios.  You likely remember the issue
> with the small snubber resistor on the high-voltage side
> overheating, which I addressed by replacing it with
> a 10-watt wire wound.  The set seems to function normally.
> As a reminder, here's the transformer input functional diagram:
> 
> http://home.netcom.com/~arc5/AVRAVT/AVA126Vib.JPG
> 
> I've begun looking at the spikes on the primary side of the vibrator transformer and have some scope photos
> of the waveforms at the vibrator contacts to the primary.
> One of my probes has "joined the choir invisible" so I can't give you a dual-trace look, but the waveforms
> sync OK.  The scale is 10 volts per division and the waveform base line is 0 volts. Here is contact #1:
> 
> http://home.netcom.com/~arc5/AVRAVT/C1NS.jpg
> 
> Pretty nasty on "make" and a 30 Volt spike on "break."
> Note the negative spike at the start as well.
> 
> Here is contact #2:
> 
> http://home.netcom.com/~arc5/AVRAVT/C2NS.jpg
> 
> This one is also nasty on the front end and has a 35 Volt
> spike on "break."
> 
> Here's my "test bed" with a pair of suppressors tacked in.
> Primary Voltage was held to 12.5 VDC in all cases.
> Now don't laugh or snear at the "new" caps tacked-in here and there;  the idea is to keep the original parts in place, disconnect them and tuck the replacements in where they will fit:
> 
> http://home.netcom.com/~arc5/AVRAVT/AVA126Asup.JPG
> 
> and here's where the suppressors went:
> 
> http://home.netcom.com/~arc5/AVRAVT/AVA126VibS.jpg
> 
> The last one I worked-on had all the old caps removed and the new ones didn't look any better ;-).
> The gray wire-wound is the "new" snubber resistor with
> a temperature probe inserted.  The large, orange disc ceramic
> under it is the new snubber cap (.02 @ 1.4 KV) Without the suppressors, the temperature of the resistor stabilized at 155 degs F.
> 
> I had some 15-Volt, 1500 Watt transient suppressors and some 500 Watt, 22 Volt ones.  I first tried the 15 Volt set with load resistors behind them
> to "sink" the nastyness, starting at 47 Ohms.  No joy-
> 15 volts is too low.  They sink too much of the waveform
> and overheat unless one uses 33-47 Ohms, and that reduces
> their effect to insignificance.  Lowering to 7-10 Ohms
> to make them effective and the suppressors will fry.  A 15-Volter without the resistor makes the waveform look sweet... for about 3 seconds.  *POOF!*  ;-)
> 
> I switched to the 22-Volt set.  These are taking far less of the waveform and run very slightly warm without "sink" resistors.
> They clip the worst of the spikes at about 22 volts.
> Here is contact #1 with the suppressor:
> 
> http://home.netcom.com/~arc5/AVRAVT/C1S.jpg
> 
> The negative spike is clipped at 0 Volts, somehow.
> 
> Here is contact #2 with the suppressor:
> 
> http://home.netcom.com/~arc5/AVRAVT/C2S.jpg
> 
> I've run the supply for an hour without significant heating
> in the 22 volt suppressors.  I get a slightly better waveform
> on the second leading spike on Contact #2 with a .047 uFd
> cap across that suppressor.  Dunno why just that one spike.
> Other values have no effect.  Curious.
> 
> With the 30-Volt-plus spikes clipped, the temperature
> in the high-Voltage end snubber only dropped a small amount,
> leveling at 145 degs F.  So there's still a lot of  "nasty" over on that side.
> 
> I'm planning to get some 20 volt, 1500 watt or higher units and give that a go.
> 
> OK all you folks smarter than me-  Comments?
> 
> 73 DE Dave AB5S
> 
> 
> ______________________________________________________________
> ARC5 mailing list
> Home: http://mailman.qth.net/mailman/listinfo/arc5
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:ARC5 at mailman.qth.net
> 
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html