[Amps] L7 FIXED (was L7 problems)
Jim Thomson
jim.thom at telus.net
Mon Dec 19 02:57:27 EST 2016
## I did not miss that. The point I was trying to make was.... a grid fuse WILL safely work... provided ALL the grid pins ARE directly bonded to chassis.....and the grid fuse is wired between chassis..and negative terminal of grid meter..or grid meter shunt. In that case, the grid is always bonded to the chassis... resulting in zero cathode fault current.
## IMO, floating the grids with caps....which provide for an RF ground, then using a RF choke for the DC return is fubar. The amount of NFB is miniscule at best, resulting in zero decrease in IMD. It actually makes things worse when using chokes....since the chokes used in the heath SB-220 have a dc resistance of 25 ohms. 300 ma of grid current... is 150ma per tube. .150 x 25 = 3.75 vdc of additional bias on top of the 5 v zener used already. So total bias yo-yos between 5 – 8.75 vdc between not talking..and talking on ssb ( PTT locked up). RL drake grid chokes measure just .9 ohm...so the varying bias effect is of course much less. Just .135 vdc.
## so with grid caps and chokes tossed..and all grid pins bonded directly to chassis, drive requirements drop 20-25 watts on the typ 2 x 3-500Z amp. No yo-yoing bias either. Plus now a grid fuse can be correctly wired between chassis and neg of grid meter...or grid shunt. Ok, now you can also stuff a 4-400 etc, into any SB-200.
## Eimac even specs the use of a 3agc grid fuse, wired between chassis and neg terminal of the YC-243 tube..... which is just a socketless 3CX-6000A7 . Takes 3 secs to replace a rear mounted 3agc fuse.
## Even with grid chokes used..... or a resistor or fuse wired directly between grid pin and chassis....still never see a case of cathode fault current occurring. It would require a stupid amount of peak V to arc from anode to cathode. not gonna happen. The choke –fuse- resistor would 1st have to open off, then the B+ would have to arc from anode to cathode. Sure, the choke –fuse-resistor will open off, but the B+ still wont arc to the cathode.
later... Jim VE7RF
From: Jim W7RY
Sent: Sunday, December 18, 2016 10:08 AM
To: Jim Thomson ; amps at contesting.com
Subject: Re: [Amps] L7 FIXED (was L7 problems)
Jim... You must have missed this statement from W8JI's site...
"The grid, being directly between the anode and cathode, shields the filament-cathode from the anode. Very little cathode fault current flows unless the grid is floated on resistors, chokes, or fuses. F1 in the grid is actually a very bad idea.
73
Merry Christmas
Jim W7RY
-----Original Message-----
From: Jim Thomson
Sent: Monday, December 12, 2016 9:09 AM
To: amps at contesting.com
Subject: [Amps] L7 FIXED (was L7 problems)
Date: Sun, 11 Dec 2016 19:21:23 -0600
From: "Jim W7RY" <jimw7ry at gmail.com>
To: "Paul Kraemer" <elespe at lisco.com>, <amps at contesting.com>
Subject: Re: [Amps] L7 FIXED (was L7 problems)
Better to put a diode here to protect the meters. See the link:
D2 in the last diagram.
I put a spot for one on my SB220 replacement diode/metering board.
73
Jim W7RY
## Use several 6A10 diodes, in parallel, for the function of D2. That assy
is then wired directly between B- of HV filter caps... and chassis. At a rated 400A
surge capability....albeit for only 8.3 msecs...is the reason for 2-4 x 6A10 diodes in parallel.
Then the total surge rating is increased..and safe time extended. Fault current is just B+ divided by
value of glitch resistor. IF you want D2 to actually survive intact, after one or
more events, paralleled 6A10s are the ticket.
## To keep the B- from wandering too far from chassis potential, use RVS connected
diodes. IE: install a single 6A10, between chassis and B-, but this time with cathode, banded
end of diode, facing the chassis. To recap, 2-4 parallel diodes, with cathodes facing the B-.
And a single diode, with cathode facing chassis.
## Worse case scenario is a B+ to grid arc....or a B+ to chassis...aka metal arc. In which case the path for
fault current is from grid to chassis... and then from chassis UP through the paralleled diode assy... and back to the
B- of of the HV supply, completing the loop. If you are relying on the 240 vac fuses or circuit breaker to open off,
during a fault event, that typ takes 35-80 msecs... imo is too long. I install a HV rated fuse in series with the 50 ohm
glitch R. With 3 kv of B+..and a 50 ohm glitch R, the fault current is just 60A. 60A of fault current will open off
a 1 Amp rated HV fuse in less than 2 msecs. The glitch R limits the fault current, the HV fuse interrupts the fault current,
and extremely fast..... which takes a huge load off the glitch R. Event over.
## Looking at the W8JI url posted above. .... http://www.w8ji.com/fault_protection.htm the grid fuse is installed
incorrectly in all his examples. This is a throwback to Rich measures advocating replacing the grid chokes with either
1/2 watt CC resistors... or 3agc glass fuses...wired directly between a grid pin and chassis... fubar.
## Correct method is to toss the grid caps and grid RF chokes....and instead bond all 3 x grid pins of each 3-500Z directly
to the chassis with wide cu strap. Then mount a 3agc type fuse holder on the lower rear panel. Wire the fuse between chassis and
negative terminal of grid meter..... or grid meter shunt. Grids are always chassis bonded. Suck too much DC grid current,
and correctly installed grid fuse opens up. With no path for DC grid current, the PO of the amp drops to zero watts.
Input swr rises to infinity. Xcvr shuts itself down. Simple to replace 3agc grid fuse. Just use a FAST type 3agc fuse, like
a 250 ma type...and even less for a single 3-500Z.
Jim VE7RF
_______________________________________________
Amps mailing list
Amps at contesting.com
http://lists.contesting.com/mailman/listinfo/amps
More information about the Amps
mailing list