From: Winfield Hill
Subject: Re: Meter movements
Date: 16 Nov 2002 07:00:14 -0800
Organization: Rowland Institute
References: <3DD47F82.14E8608@videotron.ca> <firstname.lastname@example.org>
X-Newsreader: Direct Read News 4.10
> A E wrote:
>> Good old meter movements are a bit before my time but I understand that
>> as long as you know the coil's resistance and full deflection current,
>> you're golden. I think that with a simple resistor in series you can
>> make a positive DC voltmeter out of any movement, and a shunt for a
>> current meter.
>> So how come my old HP 6284A (for which I don't have schematics, anyone?)
>> has this complex long tailed pair driving the meter in differential
>> mode? My engineer buddy jokingly said 'they're compensating for the
>> friction of the jewel'. I'm still laughing at that one.
> The copper in the coil of a moving coil meter has a very high
> (positive) temperature sensitivity. This affects the simple
> I= V/R calculation, especially with sensitive meters. It can
> be compensated for mechanically, but that can be very expensive.
> The cheaper (and better) route is to drive the meter with a
> constant current, such as that obtained from the collectors
> of a long tail pair (with emitter degeneration) or an opamp.
> This also allows a less sensitive meter to be used, cheaper
> and more robust.
> Gain switching is also easier.
Here's my favorite meter-driving circuit. Range is set by R1,
for example 4.99k for a 1mA movement with 5V full-scale input.
The maximum meter current is set by R2, talking into account the
supply voltages and the opamp's maximum output capability. If
the input voltage range is too high, e.g. 10V with a 15V supply,
a pair of attenuating resistors can be used on the input.
. IN __ ____
. ----|+ \ R2 | | R1
. | >--/\/\--| / |--+--/\/\--,
. ,--|-_/ ==== | |
. | | gnd
>> I'd expect maybe to have current limiting so the needle doesn't
>> go beyond the range, doesn't slam, etc, but when I switch ranges
>> the needle slams. When I go to a low range, the needle slams...
>> So I don't get it. Why make it complex? Esp if it doesn't seem
>> to do anything? Am I missing something?
> Meters can withstand 1.5x to 2x current overload without damage
> and (knowing HP) there probably is current limiting in there,
> but to that level.
Right, one dares not limit before full scale, and I've observed
that steps to say 120% of full scale appear to be serious slams,
but really aren't and the meter is happily totally unaffected.