Destroying or not destroying analog meter movements

[eccerr0r]

Just wondering - I'm sure some here have destroyed analog meters in the past. I'm curious as to how much overcurrent is needed to destroy these things.

I accidentally left my 20KΩ/V analog meter in the 10mA range trying to monitor a solar cell. The solar cell was in subdued lighting at the time (twilight/dusk) and was getting 1mA or so. However I had forgotten about it and later the full sun shone on it...

In full sun, or typically when it's sunny, the cell delivers 100mA (it should be more than this as far as I know, possibly age related decay, dunno, hence me testing this to begin with.) The meter slowly pegged when the sun came up. The meter was resting against the peg for a while, maybe less than an hour but quite some time.

Once I noticed, I turned it back down to the 500mA range. Luckily the meter still seems to work!

How much of an overload is needed to destroy these meters by mistake? Apparently I had subjected it to a 10x overload...

[redwire]

Re: Destroying or not destroying analog meter movements

Analog multimeters have clamp diodes across the meter movement for protection.
The meter will still peg overloaded but the diodes limit what the meter movement gets.
It's usually one pair of diodes, +/-0.7V so you can figure out what 100mA on the shunt gave.

The hair fine wire on the armature is the limit for melting like a fuse, but I find the mechanical tends to go first. The pointer gets bent or it pops out of the bearings. The whack is what really hurts them. Slow rise doesn't do anything.

[eccerr0r]

Re: Destroying or not destroying analog meter movements

I would have thought the hairline meter winding would be the weakest link before mechanical. Strange. Well, the multimeter I have is so old that it does not have the protection clamp diodes - the only diodes it has is for AC rectification. (I wonder if I should add them... and how much it would affect the reading.)

I also wonder if there is a wattage number that could be used to determine damage. The good thing is that it's spread out over a significant region, but yes it could fuse...

[R_J]

Re: Destroying or not destroying analog meter movements

There are low ohm shunt resistors in the meter, only a small current flows through meter movement itself. almost all of the current flows through the shunt resistor.

[eccerr0r]

Re: Destroying or not destroying analog meter movements

While that is true, if 10mA is FS and I pass 100mA, it's still 10x the amount of current through the meter if everything is linear (and it should be!)

(For VTVMs/FETVMs this is different of course since you can design the amplifier to have an ultimate limit in current flow... The worry are for passive multimeters without protection, especially ones that have sensitive meters like those 50µA down to 10µA movements ...)

---

Speaking of diode protection, if the meter is 50µA FS and 2K ohms, it means 0.1V drop at full scale. At 0.1V a germanium diode or schottky diode would leak, but a silicon diode ... not sure how much it would leak at this voltage...

Then the reverse leakage voltage is yet another problem if we're using two anti-parallel... Ouch.

---

A simulator with some random diode model measured around 2µA forward bias at 100mV... that would be 4% error full scale which is pretty bad. The reverse leakage is 150 nanoamps, so less than 1% which is acceptable.

[R_J]

Re: Destroying or not destroying analog meter movements

Its likely the coil can handle the extra current for a short time, but I suspect the coil was getting warm. If it could'nt it would have gone open. It usually only takes a few µA (around 20µA for some movements) for full scale.
Does the meter have its rating on the front? something like FS= xx mv
Here is some meter info http://sound.whsites.net/articles/meters.htm

[petehall347]

Re: Destroying or not destroying analog meter movements

leaving on current then testing volts might be interesting .. i know it blows up dmm.s if not fused properly .

[eccerr0r]

Re: Destroying or not destroying analog meter movements

Right, a 20KΩ/V passive multimeter mathematically requires a 50µA fullscale movement. But this isn't relevant in the current scales at least in this instance because it's relative here.

Suppose we're in the 10mA range => 10mA full scale. This means that (in this case) the 20KΩ/V meter's 50µA meter is flowing through 50µA. Ideally the meter was designed to pass this current and dissipate this heat continuously.

Now if I pass 100mA in the 10mA range => 10x current. This means 500µA is flowing through the meter... So if power (heat) is I²R (and assuming Rmeter remains constant), it's dissipating 100X more heat(!!!)

YOW

And yet it survived... Now the question is if 100X heat is fine, how high can we go?

---

And another question, how high can we go instantaneous or perhaps 1 second (barring mechanical failure as stated above), if 10x current is okay continuous?

[redwire]

Re: Destroying or not destroying analog meter movements

The meter coil is basically acting as a wire-wound resistor at that point, making heat.
Doing some math using a Simpson 260-6 for numbers: Full-scale takes 50µA (250mV) or 12.5uW. The meter coil is trimmed to be 5,000 ohms it's actually lower so my numbers include the trimpots.

A 10mA shunt 25 ohms getting 100mA is now 2.5V, the meter coil 500µA 1,250µW or 1.25mW and the 10mA shunt is dissipating 0.25W

So the meter coil is getting 1/200 of what the shunt gets for power dissipation. The meter movement is really sensitive and not taking much.

I would say a long way to go to heat up the meter coil and bake the insulation or melt the fine wire. #40 AWG fuses at around 1A. They must use finer wire.
If you put 4A through the 10mA shunt, still 20mA for the meter coil until the shunt burns up.

[eccerr0r]

Re: Destroying or not destroying analog meter movements

Sensitivity usually means fragility, or at least it seems to imply it.

Let's ignore the dissipation by the shunt resistors as it's always going to be a multiple of the dissipation by the meter movement as they are in parallel - I'm only looking at the power being dissipated by the meter movement. (I think my meter is 2KΩ resistance but I've tried to ignore that fact and use relative numbers and percentage increases.)

The question is how long is it safe to dissipate 1.25mW in that 5KΩ meter movement winding when full scale dissipation is 12.5µW, ignoring the fact 250mW is being dissipated by the shunt as that 250mW won't heat the movement significantly as it's quite far away from the winding.

The good thing is that it's a fairly long piece of winding wire so the heat distributed along the length of the wire (and surface area helps). The bad is that it's wound close together for size, which reduces the surface area. Then again, the winding is typically on a metal form, which can help dissipate some of the heat.

And yes the fuse point is the ultimate limit that one should never exceed for any amount of time, which is distinct from the 100% duty cycle limit...

---

Ahh... perfect time for a FLIR camera to test this ... just need to keep below the varnish melt/smoke point

[eccerr0r]

Re: Destroying or not destroying analog meter movements

Oh... close call... I put up my experiment up again and the meter was zero... No current was flowing at all.

Or at least that's what it seems like. Of course if the movement coil broke, it should still conduct electricity.

Quickly verified the meter was still intact in voltage mode.

And then found that one of my clip test lead wires broke. Whew.