This discussion pertains to putting a meter in the negative lead of a HV
power supply, to keep the DC HV off the meter. To protect the meter and
wiring from the transient overvoltage that could occur when (yes it will
happen too) the B+ arcs to chassis, either through your precious tubes, or
somewhere in the power supply or connections. The idea of floating the
negative and positive sides is great, but eventually there has to be a
connection to the cathode side of your tube, and subseqent grid connection
(through the bias, or chassis for zero bias tubes). It would be tricky to
float the cathode entirely for DC, and insulate it same as the metering,
while holding it at RF ground to the deck. So we just tie the negative to
ground, but first run it through the plate current meter (connnected
backwards in polarity).
Thanks, Ian, G3SEK, for the improvement on the meter protection. It is
certainly easier to come up with a big rectifier diode than it is to come
up with a big zener. The ratings of zeners are no where as husky.
Just be sure and design the shunt for the plate current meter to be below
the diode turn on knee. This is slightly complicated if you are unsure of
the shunt resistor in your meter. Here is how to do it:
An example fictious circuit:
Sometimes the meter is already a 1 Ampere movement, and has all the stuff
inside (resistors) to set the range. In that case, you need to know what
the shunt resistance is, to know what the voltage drop across the meter is
for full scale operation. Then you select the # of series diodes to put
across the meter (one end is at chassis potential, where the power supply
negative lead becomes the common).
If you have a milliammeter, the method to set it up for plate current is as
follows. You can closely tailor the protection to the meter movement range.
To be sure that you never get into partial conduction of the diode during
full current operation,I would guess about 0.5 Volts = full scale on meter.
If the meter is a milliameter it has a fairly high internal coil resistance
(dozens of Ohms). If one has a 1mA FS movement, put a shunt and series
resistor to desensitize it to work at say 600 mA or 1Amp full scale. From
Ohms law, O.5 Volts / 1 Amp = 0.5 Ohms shunt needed. This resistor has an
Ampere for your legal limit amplifier. So it dissipates a maximum of 1/2
watt. Now, with 0.5 volts for the meter, and a 1 mA meter used, we need to
limit the current in the meter to 1 mA for 1 Amp through the shunt (0.5
volt drop). Take internal resistance of meter, or measure what voltage
produces full scale. (careful, use a series resistor on the power supply or
you'll blow the movement). You may also contact the meter manufacturer for
the internal resistance. For a Weston Mustang 2000 meter, it is 85 Ohms for
1 mA movement. That says it will take 0.085 volts to peg the meter at full
scale. You need to drop (0.5 - 0.085) in a resistor, with 1 mA flowing in
it. R = (0.5 - 0.085)/0.001 = 415 Ohms. Power here is minimal.
The ciruit is then a 1 mA meter for plate current, with a series 415 ohm
resistor (select from 430 Ohm or 390 Ohm values, or better yet, choose the
nearest 1% metal film value) to the movement. Then a 0.5 Ohm power resistor
is a shunt for the entire leg, with the 1 Ampere plate current flowing in
it. A big fat diode (bridge or whatever) that can withstand 50 Amperes or
so is connected in the direction across this network so that the plate
current biases the diode towards on. But in doesn't conduct unless there is
an arc and a spike of plate current, which (if the diode has 0.6 volt turn
on) will peg the meter momentarily but hopefully not blow it out the panel.
I haven't built this exactly, so maybe there are pitfalls i missed. I look
forward to any constructive criticism. I do believe that metering the
negative lead is wiser, and protecting the meter is even more so.
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
|