[Amps] Meter Shunts

[Ian White, G3SEK]

Ed wrote:
>I want to get away from the low value hand-made meter shunts. I plan on 
>using the meter (a 0-1 ma) as a voltmeter to measure the drop across 
>more standard values of resistance. Knowing that my Ohms Law is a 
>little   rusty, I am looking for some re-assurance that I'm right. The 
>metering circuit is in the B- CT return and the B- is floating 10 ohms 
>from chassis ground. If I change the meter to a 1 volt voltmeter, do I 
>then need a 1 ohm resistor to measure 1 amp FS of Plate current and a 
>3.3 resistor to measure 300 ma FS of Grid Current ?. Of course there 
>will be a 1k resistor in series with the meter. Is this corrext ??.
>
Pretty close. Right values for the low-value resistors (which need to be 
power types), but the resistor in series with the meter is too high - 
your 1k has to include the meter's own internal resistance. Also the 
accurate calculation is more complicated, as Rich points out, and then 
there's the additional problem of resistor tolerances and the accuracy 
of your meter.

The best way to handle all of this is to make that 1k resistor a small 
trimpot, and calibrate the whole thing by forcing the FS current through 
it from an external LV power supply (with the amplifier cold and dead). 
Use an accurate external meter to measure the true current. Adjust the 
trimpot to show exactly the same current on the meter, and that takes 
care of all calculation and tolerance problems in one hit.


>Next question is exactly where and what is needed for diodes to protect 
>the meter ? I have seen so many different commbinations and placements 
>that I'm totally confused. What is the current consensus ??.

The best option is to configure your shunt resistors so that your 
voltmeters always read 0.5V (or a little less) at FS instead of 1V. Then 
you can connect a single big rectifier diode across the main meter shunt 
- *not* the meter itself. In normal operation that diode will never 
conduct, but even with a small overload it will start to conduct, and 
will give maximum protection for both the meter and the shunt.

As Bill pointed out, the protection diodes should be regular silicon 
rectifiers with a 0.6-0.7V threshold voltage, not Schottky rectifiers 
with a lower threshold. Also the diodes must be able to take the maximum 
possible overload current - use a glitch resistor in the HV+ to limit 
that current to 40-50A.

"3-amp" rectifier diodes have a high enough surge rating to do the job 
without burning out, but 6A diodes will protect against even outrageous 
surges - like an arc from 6kV which jumped straight across the glitch 
resistor... something else blew instead, but the 6A diodes and the 
meters came up smiling.

However, the resistor from B-minus to ground now needs to be much higher 
than 10 ohms, because any significant current flowing through that 
resistor will make your grid current meter read incorrectly. Arguably 
you don't need that resistor at all, because the big diodes are your 
protection against the B-minus ever floating far away from ground.

But if you use an external HV supply, with the meters and diodes on the 
RF deck, there is an increased possibility that a connector or cable 
failure could leave the whole HV supply to float. That could be nasty, 
because your B-minus rail is usually not insulated against high 
voltages. It then does make sense to use a safety resistor as well, in 
the power supply and connected directly from the bottom of the capacitor 
stack to chassis. About 1K would hold it down safely without affecting 
the meter readings.

There are more detailed explanations, with drawings, at:
www.ifwtech.co.uk/g3sek/boards/triode/triode-3.htm
Credit goes to Rich for the original configuration of "stacked" diodes 
and meter shunts.

-- 
73 from Ian G3SEK         'In Practice' columnist for RadCom (RSGB)
                            Editor, 'The VHF/UHF DX Book'
http://www.ifwtech.co.uk/g3sek