Jim gives good advice here. A combination of a glitch resistor to limit
flashover surge from the filter capacitor, plus an HV fuse close to the tube
HV feed, is cheap and important protection.
One additional precaution pertains to tetrode amplifiers. If there is an
internal arc in a tetrode, it is essential to kill immediately the screen
voltage to avoid destroying the tube. In my homebrew GU74B (x3) amp, I use
three HV fuses in parallel just as the HV enters the bottom of the plate
choke. On the downstream side of the fuses, a voltage divider samples the
HV, with a 5V full scale output. If the 5V goes down because of a tube arc,
the screen supply shuts down immediately. Note that these fuses are intended
only as a safeguard against internal tube arcs, and will not blow because of
a flashover elsewhere in the amplifier. With triodes, it's probably best to
place the HV fuses closer to the power supply. In my tetrode amp, I wanted
the fuses also to serve as a diagnostic aid, since the fuses will blow only
if a tube flashes over internally. I thought that was useful information,
especially with NOS GU74B, which are notorious for arcing internally when
first powered up. And, in fact, that is exactly what happened!
The HV fuses and voltage divider circuitry are on a little PCB, which you
can see from a photo at
http://www.w8zr.net/160amplifier/RF%20Deck%20Construction.htm. Like Jim, I
use a second voltage divider across the filter capacitor which meters the HV
between B+ and B-, and also a magnetic circuit breaker across the AC line.
73,
Jim w8zr
> -----Original Message-----
> From: Amps [mailto:amps-bounces@contesting.com] On Behalf Of Jim Thomson
> Sent: Friday, December 22, 2017 11:16 AM
> To: amps@contesting.com
> Subject: [Amps] Henry 2k-4 HV inductor
>
> Date: Fri, 22 Dec 2017 08:37:01 -0500
> From: Ron Youvan <ka4inm@gmail.com>
> To: amps <amps@contesting.com>
> Subject: Re: [Amps] Henry 2k-4 HV inductor
>
>
> < The primary advantage of using a L or Pi filter (rather than a brute
> <force {just capacitance} filter) is the reduced amount of energy stored
<in capacitance
> which is undesirable when an arc is experienced.
> <The extra storage energy and the lack of inductance (which reduces the
<amount of extra
> current from the rectifier/s) add to the amount of <energy that shall be
dissipated in the
> tube/s or other fault.
>
> < It is a safety thing. (for the tube/s)
>
> <Pi filters are seldom used for low Voltage supplies (100. Volts and
> <under) because little "arcs" at these Voltages.
> --
> < Ron KA4INM -
>
> ## IF the GG triode arcs from the anode to the grid.... which is bonded
to the chassis....
> then the tube is probably defective anyway. If you hi pot test the tube,
just prior to the arc
> event, you will find that it hi pot tests on the low side...like way on
the low side.
>
> ## You dont require the huge dc resistance of the choke to limit the
peak fault current
> either... which wont work anyway, unless the choke is
> dead last. The dead simple method is to use a 50 ohm
> glitch resistor in series with the B+. And precede the 50 ohm glitch
resistor
> with a fast HV fuse. Max fault current is just B+ dcv divided by 50
ohms.
>
> ## EG, 3000 vdc / 50 ohms = 60 A fault current. For a table top amp,
> use a 750 ma to 1 A rated HV fuse. The 60A of fault current will open
off a 1A
> rated fuse in less than 2 msecs every time.... event over. And no follow
on current
> from the rectifiers either.
>
> ## The 60A fault current will be there, regardless of the HV capacitor
value.
> The HV fuse will open off in less than 2 msecs, whether you use 25
uf..or 312 uf.
> I have tested the concept 8 ways to sunday. Tested 8 x 200 uf caps in
series..and
> also 8 x 2500 uf caps in series. Filter cap value has no bearing at
all.
>
> ## Also tested using 20 x 2000 uf caps in series on a 6700 vdc supply,
same results.
> Also tested on a 8 kv supply using 24 x 3900 uf caps in series.... also
using a 50 ohm
> glitch. Both the 6700 + 8kv supplies use a Buss 3A rated HV fuse.
Same deal, 3A
> fuse open off in less than 2 msecs every time. 6700 / 50 = 134 A fault
current.
> 8 kv / 50 ohms = 160A fault current. 134A / 160A fault current opens
off the 3A rated
> HV fuse stupid fast. On the 8 kv supply, the concept was tested at
least 11 times.
> Nothing happens...except for an open HV fuse.
>
> ### Its 2017, and chokes, resonant chokes have gone the way of the do-do
bird. Put the
> money + weight into the HD quality xfmr, with either a doubler or a
FWB..and a high C
> filter cap, HV fuse, and 50 ohm glitch Resistor. If you want bomb
proof, then install a
> 2nd
> HV fuse.... and wire 2nd hv fuse between plate xfmr secondary + diode
board. 2nd HV
> fuse
> goes in just ONE leg of the sec. Ok, now you can short the B+ to the
B-.... or B+ to
> chassis,
> right at the HV filter, and nothing happens, except for open HV fuse #2.
Diode board
> remains intact.
> We tested that concept several times as well. The HV fuses are fast and
cheap, and faster
> than a HV mech relay or even a sped up vac relay.
>
> ## I use a HV meter in the HV supply, wired directly between B+ and B-
of the filter cap.
> 2nd
> HV meter is in the RF deck, between B+ and B-. HV meter in RF deck is
AFTER the 50
> ohm glitch.
> HV meter in B+ supply is BEFORE the 50 ohm glitch and mating fuse.
>
> ## IF the HV fuse associated with the 50 ohm glitch opens up, the HV
meter in the RF
> deck will read
> ZERO ! However, the HV meter inside the B+ supply will still read
normal B+ voltage
> level !
> IF BOTH hv meters read ZERO, then you know that the 2nd HV fuse, between
xfmr sec
> and diode board,
> has gone open. I also use a magnetic hydraulic circuit breaker in the
incoming 240 vac line
> to B+ supply.
> These are the super fast trip type, with the oil evacuated from the
internal oil reservoir.
> They are known as
> instant trip or NO delay types. They do not rely on heat activation
like a normal breaker.
> Ok, no you can
> lay a dead short across the sec of the plate xfmr..and breaker open off
asap, xfmr remains
> intact.. also tested that concept out several times.
>
> ## With all of the above installed, short anything, anywhere you want,
and everything
> remains intact.
> Breakers are cheap, HV fuses are cheap, ditto with the 50 ohm glitch
resistor.
>
> Jim VE7RF
> _______________________________________________
> Amps mailing list
> Amps@contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|