homemade shunts to measure DC current
Jamie Morken
I would like to make my own shunts to measure currents from 20amps to
500amps or so.
What metal makes a good shunt material? I was thinking iron maybe.
Would a 200amp fuse in a nice case make a decent semiaccurate homemade
shunt?
cheers,
Jamie Morken
Bill Janssen
Jamie Morken wrote:
Check the temperature - resistance curves of your proposed material and
select one
that is not sensitive to temperature. I don't remember the material used
in commercial
units and don't have time right now to look it up.
Bill K7NOM
Jamie Morken
I checked matweb and since I have some 3/8" 316 Stainless rod laying around
I checked it out:
It has a rating of 0.74 ohm*mm^2/meter.
For designing a 100Amp shunt that has a 50mV drop at 100Amps this is what I
got so far:
stainless rod 48.12mm long by 3/8" diameter (0.9525cm diameter)
I am a bit confused on shunt theory though, doesn't the voltage drop across
the shunt
depend on the battery voltage used in the circuit as well as the current
going through the shunt?
From what I've read about shunts if I have 100 amps going through the
circuit a 50mV (fullscale)
shunt will have 50mV across it, but what if I double the battery voltage and
keep the current through
the shunt at 100 amps? Won't the voltage across the shunt increase as well?
cheers,
Jamie Morken
"Bill Janssen" <bi...@ieee.org> wrote in message
news:3D97AE58...@ieee.org...
Joe Leikhim
Simple OHM's law is at work. The shunt is a resistor (albeit very small value)
so look at it as E = IR. so assuming 100 amps * 0.0005 OHM (your shunt) = 50
millivolts.
The voltage of the supply isn't part of the immediate equation, just the current
being delivered.
For lower current shunts (100 amp is pretty big), the resistance of the meter
movement can become a significant portion of the shunt and so should be figured
into the equation as a parallel resistor-shunt.
You can calibrate your shunt by making it a bit longer and attaching the meter
points accross the length you have calculated for the material. Then move the
connections closer or further apart to fine tune it.
Source -------M(+)-----------M(-)-----------Load
The "M's" are the meter attachment points. The shunt is "-------------"
Joe Leikhim
Jamie Morken wrote:
--
Joe Leikhim K4SAT
Jlei...@nettally.com
"Jazz is not dead. It just smells funny." -F.Z.
Eric
shunt,
:-)
--
From Eric
"Jamie Morken" <jmo...@shaw.ca> wrote in message
news:txNl9.422861$v53.19...@news3.calgary.shaw.ca...
Hi,
cheers,
Jamie Morken
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.391 / Virus Database: 222 - Release Date: 9/19/02
Isaac Wingfield
"Jamie Morken" <jmo...@shaw.ca> wrote:
Bob Pease talked about that subject in the April 29 issue of
_Electronic_Design_. See if your library has a copy.
Or, you could rummage around at:
http://www.elecdesign.com/Index.cfm?Action=Pease
and get the PDF ofthe article on shunts.
Isaac
Jamie Morken
> Bob Pease talked about that subject in the April 29 issue of
> _Electronic_Design_. See if your library has a copy.
>
> Or, you could rummage around at:
>
> http://www.elecdesign.com/Index.cfm?Action=Pease
>
> and get the PDF ofthe article on shunts.
>
> Isaac
Thanks for the link. Here is a picture of the shunt my brother and I made:
http://24.77.58.231/new/100amp_50mV_shunt.jpg
It is a 3/8" stainless steel rod hammered into two pieces of aluminum. I
determined
the length of the rod using the number 0.74 ohm*mm^2/meter from
www.matweb.com
The rod length between the two aluminum spacers is approx 48.12mm to give
the shunt a 50mV drop at 100Amps and 5 Watts dissipation.
I hooked it up to a 100X amplifier (max4194) to get 0 to 5V instead of 0 to
50mV
and sent this to a 10 bit ADC of a uC (AVR) to display the current usage on
my computer.
I measured a big 12VDC motor at 50amps when the shaft was stalled, but don't
know how accurate this is.
How can I check the accuracy of this shunt at the high currents? Do I need
to buy a heavy duty
precise low ohm resistor (aka a shunt) to check it or is there an easier
way?
Since I send the output of the shunt to a uC it would be easier to do a
software
calibration than mechanically calibrate it. Is there a way to get a very
precise resistance
measurement of the shunt? The shunt should be 0.0005 ohms but this is hard
to verify as my
multimeters lowest setting only shows 10th's of an ohm.
cheers,
Jamie Morken
John Kimball
Message id: <eWbm9.431503$v53.19...@news3.calgary.shaw.ca>:
>Is there a way to get a very precise resistance
>measurement of the shunt?The shunt should be 0.0005 ohms but this is hard
>to verify as my multimeters lowest setting only shows 10th's of an ohm.
You need a 4 wire meter, since the resistance you're trying to read is
*very* low.
Something like this:
http://www.fluke.com/products/specifications.asp?SID=158&AGID=4&PID=20144
Price will probably give you a coronary, though.
Bill Jeffrey
> >Is there a way to get a very precise resistance
That is a VERY low resistance. Does it have to be that low? Assuming you
can build the shunt, I can't imagine how you are going to connect it to
the device to be shunted. The resistance of the connections would have
to be no more than, say, a tenth of the shunt resistance, so each
connection has to be less than 25 MICRO-OHMS! Wow!
Bill
Remove NOSPAM from my address before e-mailing a reply.
Stephen Kurzban
You're pulling everyone's leg, right. I just looked at the
pix and your wiring to this megaAmp shunt would drop more
voltage than the shunt at even modest percentage of highest
rating. Sense wiring to read the voltage drop should be
employed as well as substantial feed wiring... . Nice one
<vbg>.
mike
Shunt looks fine to me.
You've got an extra screw hole in each end block. Hook the sense wires to there.
Use the opposite corners. You want the minimum test current flowing thru the
sense wire.
Run those to your amplifier and A/D converter. Put the current to be measured
thru
the original two screw holes.
You calibrate it with an ammeter. Borrow the biggest ammeter you can get.
Many cheapo DMMs have 20A scales, but watch the accuracy spec.
Pass a current thru the reference ammeter and your shunt. Calibrate the shunt
to match the reference meter.
There are a bunch of issues.
The measurement is not quite four-terminal, because you're
stuffing current thru the same end block that you're using for sensing. You can
improve
this by adding two more blocks (inside the first two) so there's no connection
between the current source
block and the sense block. This will also eliminate the temperature coefficient
in
the end blocks. You'll still have temperature variations due to the rod.
Watch the offset and gain specs over temperature on your amplifier.
They can introduce SIGNIFICANT errors in your measurement. Ground loops can also
be a big problem with hum saturating your amplifier. Look at it on a scope.
Every metal to metal contact is a thermocouple. May or may not be a problem, but
check it out.
mike
Bunch of stuff For Sale and Wanted at the link below.
http://www.geocities.com/SiliconValley/Monitor/4710/
Stephen Kurzban wrote:
--
Bunch of stuff For Sale and Wanted at the link below.
http://www.geocities.com/SiliconValley/Monitor/4710/
Jamie Morken
>You're pulling everyone's leg, right. I just looked at the
>pix and your wiring to this megaAmp shunt would drop more
>voltage than the shunt at even modest percentage of highest
Ya they were the thickest wires I had so I thought I might as well use them.
The wire thickness shouldn't affect the measured Voltage drop across the
shunt,
so the measured current should still be accurate (if I calibrate the shunt
that is)
>rating. Sense wiring to read the voltage drop should be
>employed as well as substantial feed wiring... . Nice one
I knew I should have taken a pic of the sense wires connected to the ADC.
:)
> The measurement is not quite four-terminal, because you're
> stuffing current thru the same end block that you're using for sensing.
You can
> improve
> this by adding two more blocks (inside the first two) so there's no
connection
> between the current source
> block and the sense block. This will also eliminate the temperature
coefficient
> in
> the end blocks. You'll still have temperature variations due to the rod.
Ok this makes sense but I'll need to use a longer rod to do this.
Thanks,
cheers,
Jamie Morken
StevJensen
>>measurement of the shunt?The shunt should be 0.0005 ohms but this is hard
>>to verify as my multimeters lowest setting only shows 10th's of an ohm.
>
If your multimeter has a 10 or 20 amp current range then hook the meter and
shunt in series and send the meters max current through the combo.
Read the voltage across your shunt with another meter.
Ohms law will give you the resistance.
How precise this is depends on how accurate your meters are.