• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

An ammeter for low voltage applications

Status
Not open for further replies.

JimB

Super Moderator
Most Helpful Member
This is another of my occasional "show and tell" threads.

While recently working on an RF power amplifier, it occurred to me that it would be useful to be able to monitor the supply current.
The amplifier is powered from 12 volts and draws a current of the order of 10 amps.
Using the current range of the ubiquitous DMM is often awkward.
For my 12v supplies, I use Anderson "PowerPole" connectors, so it would be convenient to have PowerPole connectors built in to a dedicated ammeter.

Looking through my junk box, I found a long dead project which had a small moving coil meter mounted in a diecast box.
A few basic measurements on the meter showed that it had a full scale current of 1mA and a coil resistance of 18.75 Ohm. This corresponds to 18.75mV FSD.

What to use for a shunt to give a useful full scale range of say 20 amps?
On the basis that this was to be a zero cost project, and as far as I know I have no low temperature coefficient resistance wire, I decided to use plain old enamelled copper wire. Probably not the best for temperature stability, but adequate* for what I am trying to do.

Looking through my wire box I found some 18SWG enamelled copper wire.
Looking at some copper wire tables and doing a quick calculation I found that I needed about 8.2cm of wire to give 3.6milli-Ohm for a 20amp shunt.
I cut a length of wire a few cm longer than the required 8.2cm to allow for experiment and adjustment and to make a four terminal shunt resistor.
To calibrate my home made shunt, I used a large commercially made and calibrated shunt resistor in conjunction with my bench DVM to measure the voltage across the commercial shunt while passing a current through the two shunts connected in series.
Using this method I was able to solder two wires on to my shunt in the appropriate places to give the correct current reading on the meter.

The meter scale itself was a bit of a problem. The meter was originally used in some long dead piece of equipment and had a non-linear scale calibrated in dB. Not very useful for my application.
I was about to hand draw a paper scale when I remembered that there is an application somewhere in the interwebs which can be used for printing meter scales.
After a bit of searching I found it, it is called GALVO, as in Galvanometer.

EDIT: Ooops! WRONG, it is called GALVA rather than galvo.

It was originally written in French, but there is an English user guide and version which has mostly English commands.

Anyway, after all this waffle, here it is:

12v Ammeter 006.JPG

And inside:

12v Ammeter 009.JPG

My home made shunt can be seen extending out to the righthand side of the red terminals, and bent into a hairpin loop.

So there it is.
The solder connections to the meter look as though they could do with tidying up a bit.
The hole in the box where the connectors protrude is a bit on the big side. A classic case of measuring very carefully and still getting it wrong!
I also need to make a label for the connectors to show which is IN from the supply and which is OUT to the load, otherwise the meter deflects the wrong way (obviously), making another source of annoyance at busy times on the workbench.

JimB

* As per Eric Gibbs signature line "Good enough is perfect"
 
Last edited:

MikeMl

Well-Known Member
Most Helpful Member

MikeMl

Well-Known Member
Most Helpful Member
I have done similar, but I used more length of smaller enameled Cu wire as the shunt. I wound the Cu wire it around the body of a 2W carbon resistor...
 

JimB

Super Moderator
Most Helpful Member
I have done similar, but I used more length of smaller enameled Cu wire as the shunt. I wound the Cu wire it around the body of a 2W carbon resistor...
I have done similar things in the past.
But for this application I wanted as low a volt drop across the shunt as possible.
When you only have 12 or 13 volts to start with in a high current situation, all those millivolts add up to something significant.

Also, the lower the resistance, the lower the self heating of the shunt at high current. So having less effect due to the temperature coefficient of the copper wire.

JimB
 

MikeMl

Well-Known Member
Most Helpful Member
...
But for this application I wanted as low a volt drop across the shunt as possible.
When you only have 12 or 13 volts to start with in a high current situation, all those millivolts add up to something significant.

Also, the lower the resistance, the lower the self heating of the shunt at high current. So having less effect due to the temperature coefficient of the copper wire.

JimB
Isn't the shunt resistance required to calibrate the meter movement the same regardless of how the shunt is built. I guess I should have said "shorter length of smaller enameled wire". All roads lead to Rome...
 

JimB

Super Moderator
Most Helpful Member
I guess I should have said "shorter length of smaller enameled wire".
OK, I see what you are getting at.
Another part of my thinking when sizing the wire for the shunt was that I would need to adjust the value of resistance to get the calibration right.
There are a couple of obvious ways of doing this:
Adjust the position of the voltage sensing taps by heating the solder and sliding them. With this method the taps are easily slid in both directions, and a longer length of wire gives more easy resolution to the adjustment.
Using a file to reduce the cross section of the wire to decrease the cross sectional area. If you go too far with this it is difficult (impossible) to go back.

JimB
 

spec

Well-Known Member
Most Helpful Member
Very useful info Jim.

I suppose you know that the coil of your meter is probably wound with copper wire and, as you have used copper wire for the shunt resistor, I think you have embodied a degree of temperature compensation.:cool:

spec
 

spec

Well-Known Member
Most Helpful Member
OK, I see what you are getting at.
Another part of my thinking when sizing the wire for the shunt was that I would need to adjust the value of resistance to get the calibration right.
There are a couple of obvious ways of doing this:
Adjust the position of the voltage sensing taps by heating the solder and sliding them. With this method the taps are easily slid in both directions, and a longer length of wire gives more easy resolution to the adjustment.
Using a file to reduce the cross section of the wire to decrease the cross sectional area. If you go too far with this it is difficult (impossible) to go back.

JimB
The bigger the shunt resistor the better the cooling.:)

spec
 

MikeMl

Well-Known Member
Most Helpful Member
...I suppose you know that the coil of your meter is probably wound with copper wire and, as you have used copper wire for the shunt resistor, I think you have embodied a degree of temperature compensation.:cool:...
Not likely. Say the meter was intrinsically internally compensated (an expensive meter would be). The resistance of an external shunt made from Cu wire would have a tempco as described here and in the linked calculator. Note the resistivity tempco of the various materials shown in the calculator...

It is obvious why NiChrome is the material of choice for shunts...
 

spec

Well-Known Member
Most Helpful Member
Not likely. Say the meter was intrinsically internally compensated (an expensive meter would be). The resistance of an external shunt made from Cu wire would have a tempco as described here and in the linked calculator. Note the resistivity tempco of the various materials shown in the calculator... It is obvious why NiChrome is the material of choice for shunts...
Yes agree, but I was thinking about-run-of-the-mill movements which, as far as I know, are just copper wire coils.

spec
 

JimB

Super Moderator
Most Helpful Member
The resistance of ....... made from Cu wire would have a tempco as described here and in the linked calculator.
Which reminds me of an interesting thing.
If you need to measure the temperature of a motor or transformer, it is easily done be measuring the winding resistance when the motor is cold and again when it is hot. Then a simple calculation will give the temperature rise.

JimB
 

dougy83

Well-Known Member
Yes agree, but I was thinking about-run-of-the-mill movements which, as far as I know, are just copper wire coils.
That sounds good for ambient temperature compensation, but nichrome will better handle the error caused by the shunt self-heating when the 20A are flowing through it.
 

spec

Well-Known Member
Most Helpful Member
That sounds good for ambient temperature compensation, but nichrome will better handle the error caused by the shunt self-heating when the 20A are flowing through it.
Quite correct.:)

Incidentally the temperature coefficient of copper wire is 0.004041, so if, for example, you had a temperature increase of 10 Deg C, the error would be +4% (if I have the sums right).

spec
 
Last edited:

schmitt trigger

Well-Known Member
Nice "zero cost" project!! I enjoy your posts very much.
As long as the wire doesn't heat too much because of the current flow, the accuracy should be adequate for an analog movement.

One question, though:
I googled, yahooed, binged and duckduckgoed "GALVO" but did not come out with any results pertaining to analog meter drafting software.
 

JimB

Super Moderator
Most Helpful Member

dr pepper

Well-Known Member
Most Helpful Member
Andersen connectors are very good, I use them a lot on mobile plant, some carry a lot of current.
 
Status
Not open for further replies.

EE World Online Articles

Loading
Top