The AM Forum

I need a couple of meter shunts for monitoring DC current in the 30 & 100 ma area. A quick calculation shows that they need to be in the 0.7 ohm area. IIRC you could buy wire specifically made to wind meter shunts. Winding a shunt would allow me to tailor the shut and calibrate the meter.

 The alternative would be to dig up some carbon comps and bring “it in” with a triangular file. Seems like mechanical integrity might be compromised here and kind of a pita to get it right.

I could possibly use wire out of a power resistor. But I’ll bet the wire scavenged from a power resistor is brittle and not very friendly to being reworked.
 
Suggestions?

Thanks again
Ted / KC9LKE

Ted,

Check the Mouser catalog (or online at www.mouser.com) and they probably have them, you may have to use a couple in series or parallel to hit your exact value.   Keep in mind that the meter movement itself has some tolerance also. 

Otherwise find some resistance wire (possibly from a defunct WW resistor or pot).   I would avoid the filing a carbon resistor routine.  I know this was a cherished ham method in the past but the result isn't necessarily stable and an unstable precision resistor doesn't seem to make much sense :)

YUP!

Thanks
Ted

Usually the wire on ordinary wire-wound power resistors won't take solder.  I use the enamelled or cloth-covered wire taken from old wire-wound precision resistors in the junk box or taken from defunct multi-meters. I never throw those resistors away, but keep them just for the purpose. Usually the wire in those resistors will take solder.  Small gauge enamelled copper wire will work too, but it may not have as stable a temperature coefficient as wire specifically produced for precision resistors. I suspect the wire in the resistors is some kind of special alloy designed to take solder and have stable resistance with varying temperature, not pure copper nor pure nichrome.

I sometimes calculate the shunt resistance needed to get in the ballpark, but always make the final shunt by trial-and-error, intentionally making the shunt wire a bit too long and clipping off the length a little at a time to shorten it until the meter reads precisely on scale, using another meter in series to monitor the current.  The closer I get the shunt to the correct length, the more I turn up the current towards full scale.  I usually employ a small DC power supply with a rheostat in series to get just the current I need.

Be careful to always make good connections and double-check everything while testing the shunt; a bad shunt connection could damage or destroy the meter movement. I usually solder the shunt in place each time, so that once the magic length is achieved, I can just leave it in place. Inevitably, soldering the shunt in place after the proper length is found, will result in error, especially with a short shunt wire. Never use alligator clips to temporarily connect the shunt wire; they have a bad  habit of popping loose just about the time you turn on the juice to fry the meter.

Be careful measuring the resistance of a meter movement with an ohmmeter.  Some meters send enough current through the circuit when measuring low resistances to damage a meter movement (learnt from experience!). A good trick to calculate the internal resistance of the movement is to get the meter to read full scale, and then shunt just the right amount of resistance across it to make it drop back to half scale.  That resistance is the same as that of the meter.  A decade box would be ideal.  Otherwise, use a wirewound pot or slide-tap resistor and then measure the resistance you come up with, with an ohmmeter. This will get you started towards figuring out how much shunt resistance is needed.

Here's an alternate way of making a standard 1 ma @ 100 ohms meter read 0 - 30 ma without using low resistance shunts: Take a 47 ohm resistor (5% would be nice, 1/2 watt or larger) and assume you are running the 30 ma current through it. This would give a voltage drop across the resistor of 1.41 volts, which probably can be tolerated in almost any circuit without changing it's parameters. Connect the 0-1 ma 100 ohm meter across this resistor with another resistor in series with the meter. You will need one that allows the meter to show full scale (actually 1 ma) when 1.41 volts is across the resistor in the line to be measured. This would be a total of 1410 ohms,  the meter resistance is already 100 ohms so you  need an additional 1310 ohms to make it read right. This can be made by using a   2200 ohm and a 3300 ohm resistor (5 % again, 1/2 watt is fine) in  parallel, this combo would give Rtotal of 1320 ohms, close enough for government work. The same logic will make any low range meter read whatever.

Mouser has low/fractional ohm value resistors, not expensive. You could by a few or various close values and use the one that gives you the best meter indication.

As Ed points out, you can get close with low value resistors and then pick up the difference using conventional hookup wire. A typical American Wire Gauge Copper Wire Chart is here: http://diyaudioprojects.com/Technical/American-Wire-Gauge/ The beauty of this is that no special wire is needed, you can solder it, and the wire lends itself easily to being trimmed to get you exactly the value you need. I've never had to go buy or cobble specialty "resistance wire" to make a meter shunt.  

why do you need special wire?  I used normal 26 gauge copper wire on the plate current meter of my HB linear.  Took about 2 ft to get the right resistance.  I covered it in small heatshrink tubing and coiled it up on the back of the meter.  Works FB.

Rich

Typically, "resistance wire", depending on the type of wire and gauge, has more resistance per ft then a typical copper wire of the same length and gauge. Some resistance wire also have some good temperature attributes if that's needed.

A fine publication here details all sorts of resistance wire.

"Alloy Handbook of Electrical Resistance " by the Wilbur B. Driver Company

http://www.bunkerofdoom.com/lit/wbd/wbd_nopic_pub.pdf  - 55MB version

http://www.bunkerofdoom.com/lit/wbd/wbd_full_pub.pdf - 255MB with lavish illustrations

WBD was bought by GTE in 1967. I don't know where to buy the products any more.

a few months ago ER had a good brief article on how to make meter shunts and multipliers.  it was in one of the issues of the past 8 months.  I forget the author but it was a really good two or three page article. 

Most 1ma meter movements I've worked with usually measure about 50 ohms.  I use a similar method in my HB xmtr.  Used a small resistor in series with the current to be measured to produce a 1 volt drop.  Then put two 470ohm resistors in series with the 1ma meter movement which measures the voltage 0-1volt.  Example, I use a 5ohm resistor in the cathode of a 6146.  With 200ma through the 5ohm resistor the drop is 1 volt. The 1 volt drop is measured by the 1 ma meter that has the two 470ohm resistors in series with it.  The scale on the 1ma meter movement is 0-200ma.  To be more accurate, I use 5% and 1% resistors and then I measure each to select ones that are exactly what I need. (I have thousands of resistors on hand).  You should take into account the fact that the 1000 ohms of the volt meter circuit will be in parallel with the 5ohm resistor.  So, the 5ohm resistor will be selected to be a little above 5 ohms so with the 1k ohms in parallel you end up with 5.0 ohms (probably you don't need to be that exact).

As you may see, there are many ways to measure currents and voltages in a xmtr, using the right meters and meter circuits.  A single 1ma meter (with the right scale or scales on it) and a rotory switch can be used to measure different voltages and currents in a xmtr.

  That wire is nichrome wire. Yes soldering is impossible with ordinary solders. Might be possible to spot weld to something, but the technique would need some experimentation. Another idea is to just crimp each end of the nichrome wire to a terminal lug sized for the wire, and the meter studs. Since the wire might be too small, then just use the bench vise method to smash flat the crimp area of the lug to capture the wire. To increase the contact area make the wire loop back within the lug in a 'U' shape.

   Getting the length right is always tricky. One idea comes to mind. Take 1 foot or more of shunt wire, and hook up to a lab DC supply set for constant current mode. Crank up the current until you have the amount you need for meter full scale current. Then probe the wire to see what length drops the full scale meter voltage (often 50mv).

  Once you know the length for say 1 ampere gives 50mv, crimp lugs as described above on each end where the desired wire length is placed between the crimped ends. Done!
 

  This technique will work well so long as the completed full scale current is much greater than the meter sensitivity full scale current. When this is not the case, like taking a 0-1 ma meter and making a 0-2 ma meter with a shunt, then you need to do some further calculations with ohms law. In this case ordinary resistors could be used since the shunt resistance would likely be in the range of off the shelf 5% carbon film resistors.

  The nichrome wire will be more stable than ordinary copper wire so long as the crimps are done well.

Jim
WD5JKO

I have found this useful:

http://www.jaycar.com.au/images_uploaded/shuntmul.pdf

Rich

Thanks rich. I printed off a copy.

Its like Don stated, he can make exactly what he needs when he needs it.
I was hoping to find a source for the wire but like others have stated, there is more than
one way to skin a cat and I appreciate everyone’s suggestions.   

Thanks to all for the Ideas
Ted

Ted,

With any type of HB shunt make sure that your connections are solid because if the shunt fails you are likely to damage the meter movement from overload.  In some instances a pair of diodes across the movement may protect it from burnout in case of circuit failure.

I planed to fuse them, but adding a crowbar is a good idea. Thanks.

Ted

In my exciter I have a meter (I believe it's from a heath dx100, 46.1 ohms, 1mA
FS) which is switched to measure 0-10 mA (-50 volts through shunt), 0-50 mA
(300 volts through shunt), and 0-150 mA (600 volts through shunt).

What diodes should I use ?

73 de Peter

"600 volts across shunt"
600 / .6 = Alot of um  ;D

OK I re-read Your post, at 46 mV = 1 fuse it also.

I've seen zener’s used as crobars on power supplies.
They were a PITA.
Their failure pattern could have been used as a random number generator.

Ted

Thanks Ted, but where can one find 1mA fuses ?

Also, in the (switched) situation I'm in, what kind
of diodes can one use ?

It just goes across the meter. No mater what range your in you still have 42mV 1mA FS.
That never changes.

I haven’t looked for fuses yet. Probably won’t find a 1mA fuse. ;D

I'll probaly use a 1N4007 got a bunch of them.

The idea is some added protection for the meter, it’s not perfect. Best case, when the diode does conduct due to a failure, your still going to have 0.6 across the meter and the fuse will open do to the conducting diode. Hopefully the meter will live through having 0.6 volts across it for a short period of time.

Ted

Since we are on this topic....

I am adding a string (40) resistors to use a 10 mA meter as a 0-3000VDC voltmeter.  The meter actually draws .00842mA full scale and so I getting a power dissipation of 25W for the resistor string!

Perhaps meters, used to read high voltages, have to draw small currents to avoid this?

ERich
 

Not sure what kind of filter cap setup you have but if you have multiple series capacitors with equalizer/bleeder resistors you can meter close to the ground end of the string.   Otherwise the option is to use a higher sensitivity meter; a 1 mil movement would cut your power dissipation down to under 3 watts.

AHH, that is a good idea! 

Rich

Along the lines of what Don W2DL mentioned...

For the values you've mentioned I've always had good luck getting accurate metering by going a little higher on the shunt (in your case probably 1 ohm) and then adding a pot in series with the meter to bring the reading back down to exactly where it needs to be.   

With some protection diodes across the meter the pot usually also serves as a sacrificial device if the shunt fails. :)

Mark

Typically 1000 ohms per volt. Multiple resistors are still used so as not to exceed voltage across each resistor. Back to back diodes across the meter for safety in case the meter multiplier were to short through a carbon trace.