On Tue, 9 Dec 1997 13:00:45 -0700 John Lyles <jtml@lanl.gov> writes:
>HV CONNECTORS:
>
>Thanks for reminding us about Alden Co for connectors. I have not seen
>their literature in years, but somewhere remember seeing them. Sorry
>for
>the high end recommendations yesterday, I realize that we are hams.
John, I always thoroughly enjoy your posts and detailed information.
Please keep it coming.
There is still nothing basically wrong with the Millen connector when
used with a bit of caution and common sense.
For many years I used them on 2x 4-1000A amps at 5200-6KV Ep. As I
mentioned a few days ago, the plexiglass plate is the only safe way to
use them over ~3KV.
I used to use new...not old and yellow...RG-8/213 for the feed from the
PS to the RF deck and the 37001R at both ends. Altho rated at 5KV, I
believe that MIL rating is very conservative. I leave the shield and
jacket intact and run the shield over to the Black Millen 37001B for the
B- line. I have never had this cable arc, even at 7500VDC . I am also a
believer of HV fuses at any voltage above 4KV.
Since I found Alden about 9 years or so ago I used their connectors for
any serious work. They were kind enough...at a price of course...to mold
a 30KV rated wire right into the chassis mounted female connector for me.
At 18" this gave the builder some location options. I sold these sets for
years and was buying them in 100 qty.
I
>had
>just spent a week searching for a 15 KV connector that can handle 40
>Amps
>pulsed,
>and settled on a custom unit from Caton that is built in a
>Burndy-style MS
>circular connector. It's the Caton series 16, and we're probably
>giving
>them several hundred dollars each for these. But they have to be as
>close
>as 100% reliable for 10 years, at 7000 feet above sea level. And they
>would
>get unplugged and plugged about once a year, for maintenance.
>
>The problem with 10-15 KV on connectors is that now you have to start
>paying attention to gradients and the field enhancement at points
>where the
>dielectric insulation, air and the conductor come together. If the
>dielectric insulation of a cable is, say 1/4 inch thick, and the
>conductor
>is 16 guage equivalent, and the dielectric constant of the insulation
>is
>3.5, then for a given 10 KV across this sandwich, there is a gradient
>(electric field in V/inch) of 10KV/4 or 40 KV/inch, if it was all
>homogenous insulation. If there is an air void, such as a tiny space
>in the
>ground (shield) of the outer conductor of the cable or the connectors,
>then
>the field in this air space is enhanced by the 3.5 dielectric. In the
>void
>the field is over 120KV/inch, even with only 10 KV applied to the
>system.
>This is the triple point effect, not to be confused with ROCKY POINT!
>
>For air at 7000 feet and 40 deg C. (in aircooled equipment) the
>breakdown
>gradient is 75 KV/inch (sea level) x 0.75, or about 57 KV/inch. By
>exceeding this in the voids, there will be partial discharge during
>operation. In a matter of hours, days or weeks, the insulation will be
>degraded by the presence of Ozone, and the polymeric insulation will
>finally erode and arc through. This is very common in cables and
>connector
>designs that haven't considered these shortcomings. We encounter this
>and
>design around it constantly. Especially in 100 KV connectors for
>Klystron
>amplifiers. The cable companies actually use a semiconducting layer in
>the
>center conductor (if it is stranded wire) and around the shield, to
>make it
>appear as a solid conductor without voids. Also, the connector
>companies
>(like Caton, Alden, Reynolds) have to pot their connector insulation,
>often
>using vacuum to ensure that all air bubbles are gone. The old shell
>and
>Teflon insert such as an HN do not have adequate margins for 10 KV
>operation, at our altitude.
>
>Probably the plexiglass sheet and hole with grommet, and wire coming
>through with soldered ends are the most reliable for hams.
That sounds too hammy! Works fine but not something I would want to show
a visitor that was not into ham grunge. For a few bucks more just put the
Millen, HN or whatever in the grommets place.
The banana
>pins
>would be OK too, except would have to be on an insulated sheet. But
>again,
>the safety aspects of it falling loose should be considered. If you
>are
>truly going to put 10 KV through it, think about the gradients around
>where
>the hot conductor and the grounds are close.
Shudder !!!! More hammy save a penny IMO...heck, why not just use GR
plugs in that case??
>
>ROCKY POINT:
>
>G3RZP is correct that there is definitely a process for high power
>tubes in
>which we slowly increase the DC, while watching the current flicker.
>You
>don't need to worry about cutoff if you do this with the filament
>turned
>off. As you increase the hi-pot, arcs will occur within the tube. If
>enough
>energy is dumped into the arcs, it will clean off the sharp points in
>the
>tube (even in vacuum) and raise the threshold of Rocky Point problem.
>This
>is called debarnicle-ing by some manufacturers. I am talking about
>thoriated tungsten tubes here, don't know if it is acceptable for
>oxide
>cathodes. Debarnicaling in an amplifier is not advisable unless the
>proper
>energy limiting resistance is used between the power supply
>capacitance and
>the tube.
Of course one chap at least will require an elaborate parasitic
suppressor network.
The folks at CPI/Eimac will test big tubes in a special
>shielded
>device at higher than normal operating voltage, and record the cold
>emission as well. This is related to the internal fields around sharp
>points, the quality of the vacuum, the cleanliness of the ceramics,
>and the
>production of X rays.
>
>
>TEST FOR PARASITICS:
>
>If the tube is arcing in an amplifier even during cutoff (no idleing
>plate
>current), then one suspects gas, sharp points, etc.
Am I glad to read that statement John !
If the arcs occur
>as
>the bias is adjusted towards turn on, then keep an eye open for the
>parasites.
At ham levels I call them the wigglies since you can often observe them
on P or G meters.
And they are also very visable on a spectrum analyzer and do not
magically appear without notice and then disappear for a few years.
73 Carl KM1H
There is a way I used for broadcast cavities, in which you
>put
>the tube into conduction, in the rf circuit. This means a low value of
>bias
>or maybe zero on a tetrode, with screen voltage normal. Then only
>pulse the
>plate and screen contactors on and off, even just pressing the button
>for a
>short time. The shock of the stepped voltages can make the amplifier
>ring
>in it's VHF or UHF circuit resonance. If you're quick, you can do it
>without damage. If not, sorry about that....
>
>John
>K5PRO
>
>
>
>
>
>
>
>
>
>
>--
>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
>
--
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
|