meter shunts

[W1RKW]

I need to make some meter shunts.  Anyone have any tips on making shunts?  I need to convert a 50mA panel meter to handle 500mA full scale.  I've tried using a lengths of various guage wire I have on hand but I end up with a big gob of wire of a certain thickness which I prefer not to coil up or wire that doesn't give me the full scale swing. I can't seem to get the perfect setup.

[KF1Z]

What does your meter take for full scale?   about 1mv ?


If so, a 2.4 inch piece of #20 should be about right.


Or, a 0.002 ohm resistor should do it.....


I know there's plenty of other folks with better suggestions.....

I'd like to hear them too!

I don't have an ohmmeter that will read in milliohms, but that would be just the ticket for building shunts.

[Pete, WA2CWA]

I once built a power supply where a needed a meter shunt. Took a piece of a printed wiring board that was stripped of components and bridged all the paths on the left and right sides with solder. Set up a little test jig on the bench with this meter shunt in place with the meter I was going to use and then, using a nibbling tool, nibbled away parts of the board until my meter read what it was suppose to read in the test set up. After 25 years, it still works fine. These type of boards, especially if you can find some with large ground planes on them, can be very useful as meter shunts especially if you have to work with resistance values below 1 ohm or .1 ohms.

[Tom WA3KLR]

Bob,

The rule of thumb with using copper wire for instrumentation purposes like meters shunts is at least 1000 - 1500 circular mils per ampere for reasonably stable resistance with the self-heating of the wire.  So you should use wire greater than 450 circular mils.  Looking in the wire tables, this is #23 AWG.  #22 is common, 642 circular mils.  #22 is 16.14 Ohms per 1000 feet.

You will have 450 ma. through the shunt and 50 ma. through the meter at 500 ma. full scale.  Your meter is probably about 50 millivolts drop.  You should check that with 50 ma. applied.

Example :

Rshunt = E/I = 0.050 V/ 0.450 A. =0.111 Ohm

(0.111 Ohm/ 16.14 Ohms) x 1000 ft. = 6.88 ft. # 22 wire.  Wind some more than calculated and then tweak the shunt in with testing.

You can wind it up on a pencil or a piece of dowel and tie it up.

Good luck.

[VE1IDX]

I like what Nautel did with some of their older solidstate transmitters.They used a long stainless steel bolt mounted on insulators.One lead was fixed at the head of the bolt and the other was mounted between two nuts.You adjusted the meter shunt by varying the distance between the leads along the threaded portion of the bolt.Stainless was used as it has a much higher resistance than copper.You could really fine tune those suckers.

[Ian VK3KRI]

Purits don't use copper wire as shunts beacause of its temperature co-efficient.  I've never let it stop me though. I generally just coil the wire. ( I'm talking Amps not MillAmps here)
                                                                                  Ian VK3KRI

[k4kyv]

Purists don't use copper wire as shunts beacause of its temperature co-efficient.  I've never let it stop me though. I generally just coil the wire. ( I'm talking Amps not MillAmps here)

Even if you're measuring microamps, the ambient temperature will affect the calibration.  I found steel wire even worse.  I once made a meter shunt out of a steel strand from field telephone wire, #20 something.  It worked well, with enough resistance that I only need about 3 inches, but the meter would never read the same from day to day, compared to a known calibrated meter.

I always save old wirewould resistors from multimeters, panel meters, etc that otherwise would be junked.  Whatever they use for wire always seems to take solder.  The wire from ordinary wirewound resistors usually won't.

From my experience, forget about all the mathematical formulae for calculating shunt length.  I never could get any closer than about 20% using that method.  I have had much better success with trial-and-error, using a variable voltage source and a DVM or well calibrated panel meter or VOM as the current standard.  I get a rough idea using a variable  rheostat as a temporary shunt, or by taking a wild guess.  The idea is to get a reading on the meter at the lowest output voltage from the power supply.  Then I increase voltage and thus current, towards the desired current range of the meter, and get a rough idea of what resistance I need.  I suppose I could calculate it, if I wanted to.  Then I take a piece of shunt wire, cut it somewhat longer than what I know I'll need, and test the meter - once again running the voltage to  zero and bringing it up slowly (so as not to destroy the meter).  Then I start trimming the length of the wire until I get the meter to read exactly the same as the calibrated meter.

Never attempt to determine the final length of the shunt BEFORE soldering it to the permanent meter terminals.  Inevitably, even if you think you have the exact length, re-soldering it into its permanent position will shift the calibration enough to cause a noticeable error.  It may be a pain, but when I get close to the final length, I solder each trial length into place so that once I hit on the sweet spot, I can just leave it in place.

If you end up cutting the wire too short, cut another piece somewhat longer, discard the short piece, and continue until you get the calibration exact.

When making a meter read more than about 10 times the basic full-scale current rating for the movement, you will get better mechanical damping if you  place a resistor, several times the resistance of the meter coil, in series with the meter movement, and place the shunt across the meter/series resistor combination.  If you have one, use a 1% or better precision resistor for the series resistance, or make one out of risistance wire.  To be effective, the series resistor needs to be several times the resistance of the shunt.  A low-resistance shunt directly across the meter movement will make the meter sluggish.  (This is the equivalent of trying to turn a generator with a dead short across the output terminals.) 

When shipping or transporting a meter, short the terminals directly together with a piece of wire.  That will reduce the likelihood that the movement will be damaged from flopping around when the meter is inevitably subjected to mechanical vibration and shock.

[W1RKW]

Thanks for the tips everyone.  Working with such small resistances begs the question, is there a way to measure such small resistances with your typical multimeter?  I'm guess some sort of bridge would be the way. Yes/no?

[KB2WIG]

how much resolution does your multimeter have?...    Can you measure uA  and mV?  How accurate is your equipment?...  After considering these things, measure and apply Ohms Law. Its the right thing to do. There are a few techniques to increase the accuracy, or make the readings easier.

Its been a long time since I've done any of this type of stuff, but it can be done..   Mr. Williams had us doing this stuff for a week or two in HS Lectricy class. A hatefull exercise..      one oldie -- with 'analog" meters is to 'buck' the voltage so your reading are in center scale of the meter... increases the resolution..... 

k4kyv has the berries....

Anyway, Yes, da bridge will work....         KLC

[W1RKW]

....Mr. Williams had us doing this stuff for a week or two in HS Lectricy class....

Mr. Potter had us do the same thing in 'lecricity class.  Unfortunately, it's been nearly 30 years since then.  The memory is fuzzy. To much Budweiser.

Will have to experiment with bridges and various shunting methods.  The bolt thing sounds interesting.  Will have to experiment with that too.

[W1RKW]

It seems I have found what I need to get the right config.  I'm using two 2 inch pieces of nichrome wire in parallel. It's amazing, if I'm off by a 1/64 of an inch I see a 10mA change.  Now I need to determine how sensitive to temperature is nichrome wire. Anyone have any guesses?

[Tom WA3KLR]

"Working with such small resistances begs the question, is there a way to measure such small resistances with your typical multimeter?  I'm guess some sort of bridge would be the way. Yes/no?"

If you have a lab supply with adjustable current limit you can use the “4 wire” method of measuring low resistances.  It gives high resolution and is accurate.  The contact resistances of the supply leads and meter are not a factor because the current launched through the loop under test is constant.  The DVM is on voltage mode and it is at 10 or 11 MegOhm input, so the meter’s probes contact resistance not a factor either.

You put a large current as possible through the unknown resistance, within current-carrying capability of the object of course. 

Set the lab supply for just a couple volts or a few volts higher than the drop you expect in the test loop.
Put your multimeter on dc amps and set the lab supply for say 1 ampere exactly or a value of current close to maximum output of the supply.  Record the current limit value.  It does not have to be an exact even number though.

Connect you resistor, wire or whatever to the supply.  Extra clip lead length is o.k. as the current will be the value you previously set the supply to.

Now put the multimeter in the low voltage dc range and measure the voltage drop right at the end terminals of the unknown resistive device itself, near the alligator clip junction, but not on the clips.

Calculate the resistance by Ohm’s Law :

R = E/I, where E is the voltage drop across the unknown resistance and I is the current limit value you set the lab supply to.

I once measured bonding resistances on a rack cabinet to a resolution of about 10 micro-Ohms, I think it was, by using this method.  I had a high resolution DVM and I think I was using 4 amperes – the limit of the lab supply available at the time.  I could easily see the resistance of the rack rails too.

- - - - - -

Nichrome wire has a temperature coefficient about 1/20th that of copper but it all comes down to what the actual temperature rise of the shunt is.  If your 50 ma. meter is about 50 millvolt drop as I guessed before, then the power dissipated in the shunt is approximately 0.050 V x 450 ma. = 22 milliwatts.  Not much, but it all boils down to what size the nichrome wires are?  Each will be carrying 225 ma. 

[W1RKW]

Hi Tom,
I don't have a mircometer so I'm not sure what the nichrome wire size is. If I were to guess I'd say about 24 or 26ga. I measured a foot length with my Fluke and it does give me about 2.2 to 2.4 ohms per foot.  I'm looking at about 1.5 inches across the terminals of the meter.  If I run the numbers and compare what you calculated, it's pretty darn close. In any event, by trial and error I managed to get the right size without having to coil up a long length of wire. Plus I found out a couple of times how sharp this stuff is. I don't know how many times I impaled my fingers. I'm glad I saved that wire. I was going to get rid of the various pieces I had after completing my dew heaters for my telescope equipment. It came in handy. Thanks for your help.

BW

[k4kyv]

Isn't nichrome one of those alloys that won't take solder?

I have noticed that the wire in most wirewound power resistors is riveted to the terminals, not soldered.

But the stock shunt wire used in all the meters I have ever encountered does take solder.  Also, I have salvaged the wire from wirewound 1% precision resistors to make oddball low-value resistances such as homebrew meter shunts, and never had a problem soldering it.

[W1RKW]

The nichrome I have here is won't solder with standard tin/lead.

[k4kyv]

The nichrome I have here is won't solder with standard tin/lead.

Then, what kind of solder do you use?

[W1RKW]

Didn't use any solder for the shunts.  I simply bolted them to the terminals on the panel meter. For my dew heaters I'm not sure how I connected the nichrome to regular wire.  I built them a while ago so I don't remember.  I may have simply crimped the wire.
Silver solder doesn't work either.

[Tom WA3KLR]

Nichrome wire is primarily used for heater wires. It is not solderable.  If you look in to toasters and those 1000 - 1500 space heaters for your home, they have to crimp the nichrome wire connections.

Nichrome is used occasionally for shunts however.  I have a 15 Amp 120 V true power meter that uses a nichrome shunt.  The shunt is about 3” long, thin and hollow tube, probably for maximum surface area/cooling.

I looked up a nichrome wire table and #26 is about 2.6 Ohms per foot, so that is probably what you have.   My guess is that your nichrome shunt will actually have worse temperature drift (relatively speaking) than if you use a copper wire.  Since the resistance per length is much higher, there is more heat per length than copper wire.  But the nichrome is 1/20 the temperature coefficient of copper.  Opposing factors here, everything is a tradeoff in engineering, that is the way things always are.  To have a good nichrome shunt with low temperature rise, I would think that it must be a few wire gauges larger at least than #22 which I quoted for the copper shunt earlier.

But if you can test the shunted meter and don’t see any significant change in reading, you are o.k.

The previous remark that purists don’t use copper is the kind of remarks in these hobbyist forums that can throw you off.

[Jim, W5JO]

All remember that Johnson used Nicrome wire for their meter shunts in the Vikings and Valiants.  As I recall one of the books state that if the shunt needed to be revmoved for any reason to crimp the wire tightly to the solder terminal and flow solder over the connection to keep it immovable.

[W3SLK]

I thought that you could solder to Nichrome. Its just you had to use LOTS of heat in order to overcome the 'sinking' effect of the wire itself. Naturally it is designed to radiate heat so it sapps all the heat when soldering.

[W1RKW]

.....I looked up a nichrome wire table and #26 is about 2.6 Ohms per foot, so that is probably what you have.   My guess is that your nichrome shunt will actually have worse temperature drift (relatively speaking) than if you use a copper wire.  Since the resistance per length is much higher, there is more heat per length than copper wire.  But the nichrome is 1/20 the temperature coefficient of copper.  Opposing factors here, everything is a tradeoff in engineering, that is the way things always are.  To have a good nichrome shunt with low temperature rise, I would think that it must be a few wire gauges larger at least than #22 which I quoted for the copper shunt earlier.

But if you can test the shunted meter and don’t see any significant change in reading, you are o.k.....

Tom,
OK on #26 nichrome

I did kind of a crude test to see what sort of drift I might experience when I heat up the nichrome shunts.  With 250mA running through the shunted meter so it would read mid scale I saw very little drift when the tip of the soldering iron was put next to the shunts for approx 90secs.  I had  a little less than 20mA of drift with the tip about a 1/4" away.  At about 450mA through the meter I got a similar result. My conclusion: I don't know what the actual temperature was at the nichrome but since the chassis will be open, there's no direct heat source near the shunts and there will be plenty of air circulation while in the rack the temp should be pretty constant on the shunts and have litte drift. At least in theory that might work.  I guess I'll eventually find out.

[W3SLK]

This is the reply I got from Rich Measures, AG6K, from the ham amps group,http://groups.yahoo.com/group/ham_amplifiers/ when I posed the question to him about soldering nichrome: Ni-Cr alloys can be soldered with 95%-Tin, 5%-silver (plumbing) solder. The melting point is c. 221ºC. J. W. Harris Sta-Brite flux - which is made for soldering stainless steel alloys -works well for soldering Ni-Cr. Before soldering, clean the oxides off of the Ni-Cr with coarse steel wool or carbide paper. Apply drop of silver solder to tip of iron, touch drop to the joint to be soldered, remove iron, apply drop of flux. The flux will sizzle -which activates the flux. Re-apply soldering iron and the joint should suck in the drop of silver-solder. The flux residue should be
removed with running warm water and a toothbrush. If the iron is temperature controlled, 600ºF seems to work well since higher temperatures break down the flux into useless brown goo.

[W1RKW]

Interesting note Mike.  When Don mentioned about soldering nichrome I tried it with the silver solder I have and I had no luck in getting the solder to flow. I wonder if the composition of the type of solder I have is different from what AG6K uses or maybe I didn't get the right temperature, maybe to high?  I'll have to mess around with it. Thanks for sharing that info.  Looks like it can be done with the right technique.

[WA1GFZ]

acid flux will allow you to tin the wire so you can solder to a terminal.

[Rob K2CU]

I have also found that acid core solder works with tinning those tabs that connect battery cells together in battery packs when you want to break one up for reuse or replacement of a single bad cell. These are the welded tabs, not the ones designed for soldering.

As to meter shunts, there is probably a shunt within the meter already to make it do 50 mA. It may well actually be a 1 mA movement in there. youi could consider removing the internal shunt, then use any stable shunt that would be less than one Ohm. the 500 mA would develope at most .5 volt. And shunts of less than one Ohm, but more than a tenth Ohm are readily avaiable at places like Digikey.  You can also find them in discarded PC swithing power supplies, etc.  So, say you find a .39 Ohm resistor. and you know you have a 1 mA movement. you can make the movement into a little voltmeter with a series resistor similar to idea suggested above, except that the current trogh same will be much less as it is no longer a 50 mA meter. So, for this example, 500 mA would develope .195 V across the .39 Ohm resistor. To get the 1 mA, you would need to have the "new" mini voltmeter have a total series resistance of 195 Ohms. if you use a 250 Ohm miniature trim pot, you can set it exactly.  As to power dissipation in the resistors, the .39 would come to  97.5 mW and the pot would see roughly .195 V x 1 mA or 195 uW....nothing.  SInce the pot is asjustable, the effect of 195 ohms in parallel with .39 would be corrected when you calibrate the thing.