Shunt wiring

[transmitterdan]

Wotname,

It's not the volt meter that causes the problem. Take your example of a 1 milliohm shunt. It's impossible to get the big bolt connections to be anywhere close to 1 milliohm. So this creates a higher resistance than 1 milliohm and thus higher voltage than designed across the bolt connections. So the current reading will be too high. But the small kelvin screws don't see this higher voltage drop because they carry no significant current. That's why shunts are made the way they are. Read an explanation here: https://en.m.wikipedia.org/wiki/Four-terminal_sensing

[Jammer]

Originally Posted by transmitterdan

This will not work. There are 4 connections to a shunt. 2 are high current connections. 2 are just for measuring the voltage which is related to current. The big connections can never give the right voltage and the current measured will always be wrong and and it will vary too much with temperature.

Either fix the small broken connections or get a new shunt.

Originally Posted by Paul L

Why would the voltage drop between the two large connectors be any different than between the two small ones? They are only a few cm further apart in the solid shunt, if that.

Originally Posted by transmitterdan

Wotname,

It's not the volt meter that causes the problem. Take your example of a 1 milliohm shunt. It's impossible to get the big bolt connections to be anywhere close to 1 milliohm. So this creates a higher resistance than 1 milliohm and thus higher voltage than designed across the bolt connections. So the current reading will be too high. But the small kelvin screws don't see this higher voltage drop because they carry no significant current. That's why shunts are made the way they are. Read an explanation here: https://en.m.wikipedia.org/wiki/Four-terminal_sensing

Heh. You're both right.

The meter is, of course, always wrong, insofar as there is always some error and uncertainty in the measurement. The questions are: How much error is acceptable? Does using the large terminals contribute to the error in a way that makes it unacceptable when it otherwise would not be?

#12 copper wire is roughly one milliohm per foot. The brass posts on a shunt are much larger than #12 and much shorter than a foot in length. How much resistance is added by the extra distance? Not much, far less than a milliohm, perhaps a few microohms. There is some resistance in the pressure connection itself, and it is difficult to get a good, stable, reliable electrical connection when the large battery cables and the small instrument wires share the same terminal.

Will it work? Sure. Will there be a measurement error? Yes. Enough to matter? It depends, probably not if everything is clean and tight and if you can live with an 0.1A difference or so.

Originally Posted by fritzdfk

I have a little bit unrelated question. Is a shunt generic other than amp rating? Can a various amp meters be connected to a shunt? Only one connected at a time of course.

Originally Posted by Wotname

No, there is no generic shunt. There are some industry standards but no requirements to be a set standard.

Some common numbers are 50, 75 and 100 mV.

Thus for a certain amp rating, say 100 amps, the shunt might be designed to drop (develop) 75 mV when 100 amps is flowing though the shunt. Another 100 Amp shunt might drop 100 mV and so on. You can do the maths to find the resistance of the shunt (by using ohms law).

In the immortal words of Andy Tannenbaum, the good thing about standards is that there are so many to choose from.

[daletournier]

Originally Posted by Jammer

Heh. You're both right.

The meter is, of course, always wrong, insofar as there is always some error and uncertainty in the measurement. The questions are: How much error is acceptable? Does using the large terminals contribute to the error in a way that makes it unacceptable when it otherwise would not be?

#12 copper wire is roughly one milliohm per foot. The brass posts on a shunt are much larger than #12 and much shorter than a foot in length. How much resistance is added by the extra distance? Not much, far less than a milliohm, perhaps a few microohms. There is some resistance in the pressure connection itself, and it is difficult to get a good, stable, reliable electrical connection when the large battery cables and the small instrument wires share the same terminal.

Will it work? Sure. Will there be a measurement error? Yes. Enough to matter? It depends, probably not if everything is clean and tight and if you can live with an 0.1A difference or so.





In the immortal words of Andy Tannenbaum, the good thing about standards is that there are so many to choose from.

Yep your right. I only know because it's working for me. The battery charger is saying 60a,the xantrex is showing 67a which is about right due to solar panel output. It's closer enough for me.

[BlackHeron]

If it is off a little bit how big of a deal is it? Just call them "metric amps" and be happy that you have something that tells you relatively how much you are drawing versus charging. If you are counting how many amps are charging and discharging from multiple meters as long as you are using the same equipment to measure in and out then it doesn't actually matter all that much the unit that is being measured. If not, then it can cause a little bit of a mismatch.

Our sailing club uses knotmeters on their boats that are "Calibrated for Fun." They tell you exactly how fast you are going in "Fun Knots" that are great for comparing how much faster or slower you are going compared to the last time you looked at the speed. It even is good to compare today versus yesterday, or last week.

Comparing speeds from boat to boat? ...not so much

[CatNewBee]

Well, sometimes it might be important to know
if you are measuring metric Amps or Imperial.

[artisanmach]

where's my small terminals?

[Wotname]

Originally Posted by artisanmach

Attachment 161193
where's my small terminals?

On top of the big terminals.
.
.
.
.
.
.
.
But clearly this won't work and will give wildly erratic and unstable results

[senormechanico]

For that shunt (post # 36) I would install the big current carrying wires immediately above the lowest washer, then a big nut, then an insulating washer, another nut, the twisted wire pair and finally, another nut.
The nut just above the insulator would be tightened against the very top nut so the twisted pair sees only the shaft voltages across the shunt itself.
This would keep the twisted pair connected ONLY to the threads which would accurately represent the voltage drop across and proportionally, the current through the shunt.

So there !

[Wotname]

Originally Posted by senormechanico

For that shunt (post # 36) I would install the big current carrying wires immediately above the lowest washer, then a big nut, then an insulating washer, another nut, the twisted wire pair and finally, another nut.
The nut just above the insulator would be tightened against the very top nut so the twisted pair sees only the shaft voltages across the shunt itself.
This would keep the twisted pair connected ONLY to the threads which would accurately represent the voltage drop across and proportionally, the current through the shunt.

So there !

No disagreement from me....but.... I would connect the small twisted pair of wires directly under the small bladed screws that are at the top of each post just as the manufacturer intended.

[belizesailor]

Timely thread, Im getting some inaccurate readings at my volt/amp meter...another maintenance task added to the long list. [emoji45]

[transmitterdan]

Originally Posted by artisanmach

Attachment 161193
where's my small terminals?

The two small straight blade screws on top of the threaded posts.

[senormechanico]

Originally Posted by Wotname

No disagreement from me....but.... I would connect the small twisted pair of wires directly under the small bladed screws that are at the top of each post just as the manufacturer intended.

Even better! I didn't inflate the pic, and mistakenly thought we were talking about the OP's shunt.
Your post amounts to the exact same connection as my detailed diatribe.

I know, I know. PAY ATTENTION !!

[Wotname]

Originally Posted by Paul L

......

Originally Posted by transmitterdan

Wotname,

It's not the volt meter that causes the problem. Take your example of a 1 milliohm shunt. It's impossible to get the big bolt connections to be anywhere close to 1 milliohm. So this creates a higher resistance than 1 milliohm and thus higher voltage than designed across the bolt connections. So the current reading will be too high. But the small kelvin screws don't see this higher voltage drop because they carry no significant current. That's why shunts are made the way they are. Read an explanation here: https://en.m.wikipedia.org/wiki/Four-terminal_sensing

Hmm... perhaps I'm not explaining myself very well or heaven forbid, perhaps I'm wrong.
First up, I understand all about four terminal measurements, especially low ohm meters etc. I've being doing this stuff for a long time now and have used all sorts of field and lab standard instrumentation as I am sure you have.

Second, the best way to use a shunt is of course the proper way, using the terminals exactly as the manufacturer intended.

However the OP had a problem and was looking for a "field" fix and after that we have got into a discussion about connecting the meter wiring directly to the "big" studs and the possible problems.

I maintain that it would be OK to connect the meter wires to the top of the big studs if that is the only fix available in the interim. There would not be large errors or inconsistent readings. If only one of the "kelvin" screws had failed, I would then use the other one as normal and field fix the failed side. If it was me, I would probably solder the meter wire directly to the brass block however fitting a large ring terminal to the meter wire and placing it on the very top of the big stud would get the amp meter functional again albeit with some error.

My reasoning is that surface of the brass block will have the near zero PD between any two points on it (maybe a some tens of micro volts). By extension, the PD between any side of the brass block and the top of the stud should only be in the order of hundreds microvolts to perhaps a few millivolts. Even if say it was 10 millivolts, there will only be a 10% error. It really boils down to the value of the current flowing though the stud and the resistance of the stud. I suggest this will not be a high number.

I understand the resistance between the large cable lug and the brass block can be high (i.e. > a few milliohms) but the degree of error will be determined by the value of the current flowing though the stud times the resistance of the stud. This will (in effect) give the PD between the existing small terminal and the top of the stud.

If I still had access to some of our old lab equipment and a micro voltmeter, I would run some tests but alas, this is no longer possible .

[Wotname]

Originally Posted by senormechanico

Even better! I didn't inflate the pic, and mistakenly thought we were talking about the OP's shunt.
Your post amounts to the exact same connection as my detailed diatribe.

I know, I know. PAY ATTENTION !!

[transmitterdan]

Guys,

Can you get a shunt like the OP had hooked up "wrong" to give some reading? Yes, and if that's a definition of "it works" then ok, it will work. But there is significant resistance in the studs and the brass block. So there is no good way to get an accurate reading using the big studs. For a small current shunt of 25-50A the error will be negligible. But for a big 250-500A shunt the error will be significant. That's my definition of "it won't work".

If the small screw holes are damaged the "best" field fix would be to drill out and tap to next larger screw size. Or drill and tap same size holes on opposite side of shunt. All cruisers ought to have a small set of drills and taps IMO.