Monitoring Current: Shunt vs CT

[waoj]

I'm interested in monitoring the current going in and out of my 12 volt car battery and it's voltage over time.  I want to thoughtfully distinguish between bad battery and phantom load causing a good battery to run empty.

This is a common task so there are already cheap products I can buy to do it.  I have seen most kits use a shunt, and a few use CTs.  Drok calls their CT a Hall sensor?

On one hand I worry that the shunt isn't going to handle the peak amperage demands of a car battery.  On the other my car is a phev, so there's no starter in the traditional sense so blowing up the shunt is less of a worry.

I also care for accuracy. This is DC and I have a suspicion that a moving vehicle that changes orientation related to the Earth's magnetic field might screw with a CT measuring DC. And because I'm wanting to measure over time I can't just take it off and zero it all the time.

I like the idea of a clamp-on CT, not having to cut into my factory battery wire to get it over the terminal head. I would rather not cause permanent alterations to the vehicle so I'll try to source a negative battery terminal and negative battery terminal clamp for whatever meter I choose so that it can clamp inline and be removed without damage.

I've seen CTs used by electrical utilities.  So I suppose they must be fairly accurate.  Or maybe the utility cares more about reliability than accuracy.

So what are the eev blog opinions between shunt and single piece CT, and I suppose clamp on CT, in terms of accuracy?

[coppice]

A CT can't measure DC. Its a pure AC device. If you want to measure DC without breaking into the circuit to fit a shunt, and having galvanic isolation issues with the measurement circuitry, you should be looking at things like hall sensors and flux gates. Hall sensors have issues with magnetic fields, and offset errors, but can be cheap. Flux gates are more accurate and flexible, but more costly.

[DenzilPenberthy]

One of these will do your job well. It comes with a current shunt that's good for 500A and I think they sell a bigger shunt if you need it.

https://www.victronenergy.com/display-and-panels/bmv-712-smart

Has it's own LCD display and links to a free Bluetooth app as well.  Very well engineered and has a good algorithm that incorporates Peukert factor etc.

I have a DC hall effect clamp meter that I use for automotive and off-grid solar work. It is amazing and super useful but you have to zero it before applying it to the wire every time and the zeroing will only stay zeroed for a few 10s of seconds max before drifting. So I wouldn't give it any hope of being super accurate even right after zeroing, and no hope after even a few minutes.  It is also not sensitive enough to low currents to see small parasitic draws.

CTs like electrical utilities use only work with AC

[waoj]

That is a neat product!  And spec sheet actually says 1ma of current to run it, which sounds really good for bluetooth, though maybe that's just idle current.

A lot of reviews complain about bluetooth distance and they mention they sell a bluetooth addon that extends range.  I like that it lists a windows app, which tells me it probably has a bluetooth serial port that I could probe, datalog with my own scripts.  But the manual says i need a custom usb cable to use the windows app

I did find this one also: https://www.amazon.com/gp/product/B01MT4583U but it clearly is not a shunt, just a voltage monitor.

I mean, I do care about voltage.  But I want to know if my battery is dying because it's crap, or because something is puling too much current in sleep mode.  So far I've found it stabilized when I unplugged the EVSE cable.  So I'm thinking something is bad about the EVSE that is keeping the HVB charge controller computer awake, but it's not in "ignition on" mode, so hvb does not top up the 12v bat.  I thought I read that the EVSE was supposed to act as a trickle charger for the 12v once hvb was full, but its clearly not.

[waoj]

So... I didn't buy the 712.  I thought my significant other would freak out of they saw a gauge and wire. And it might draw enough current to cause a problem maybe in an already weak battery.

I bought a SmartShunt, and I absolutely love it!  The Bluetooth range is a little sad but I can receive it while in the air conditioned part of the house at least!  I installed it right before leaving on a road trip yesterday and been driving throughout the night.

Here's a screen shot showing a pitstop.  From what I'm gathering the battery is "consuming" 2.8w into waste energy while driving once things stabilize and that doesn't seem too bad.  Dropping below 11 isn't good.  But being a phev it doesn't need much to boot up the big battery's computer. I bought two jump packs just in case.

I wired it in line with a 100a fuse just to be safe because I didn't buy the 500a shunt but the cheaper 100a one.  If it's any indication, I had to drill out a Ford factory lugs a bit to fit over the SmartShunt's  bolts.  So I am betting that Ford expects to see the less current than the shunt!

I measured a spike to 110a going into the battery in live view once when starting the car, but I suspect it was extremely instantaneous, because when I looked back in historical view it had averaged down to 30a while it was picking the battery back up from it's sad upper 10v potential.

Liking this tool in the American Southwest!

[andrewtaylor]

why don't you something the professinals do for mor than 20 years,
desined to withstand the automotive -40C to 105C,
esd, iso pulse in the car wire harness?

Reineventing the reel is a fruitless project ,-)

[waoj]

There actually is one on the battery already.  I thought I'd
Of trying to tap into that, but I still need something that data logs, and I but that Bosch component is just the sensor

[andrewtaylor]

but I still need something that data logs, and I but that Bosch component is just the sensor

It's a sensor with high performance ADC and with LIN databus --
so logging these data is  not rocket science at all.

[bastl_r]

A few years ago I got such a car battery sensor from a scrap.
A shunt with 80uOhm was installed in it.

[andrewtaylor]

A shunt with 80uOhm was installed in it.

That's right.
It's resistive material welded to copper via electron beam welding.
Patent for this technique had a company named ISABELLENHÜTTE, but the patent validity may have timed out.

Shunt has been tested to withstand 1000A for 1 minute -- which is much more even a 120Ah lead acid battery can deliver over time , even when fully charged & fresh from shelf 

[waoj]

Have you ever gotten milliamp precision out of a Victron?  I seem to only see one decimal place of amps.  But I'd like to look more closely and see how many MA are passing through while the vehicle is fully off to make sure it's super low, 50ma for instance.  If the victron "rounds up" then I guess 0a means less than 500ma and 0.1 means 51 to 149ma.  But if the victron doesn't round up then 99ma will report as 0a.  And I think there's a signicficant difference between 50ma parasitic load and 90ma if I want the battery to go many days without dying on me.

[guenthert]

Make that a separate measurement using a run-of-the-mill hand-held DMM.  Just don't attempt to start the car while the DMM is in circuit.

In a car, such a concern surprises me a bit.  For a number of years, I used my car rarely (some 6kmi in 12 years) and in summer not at all for months.  Never had a problem starting the car and never had the need to exchange the battery (that was in mild California weather, the car was garaged and it had a huge battery).

My 2001 BMW R1150GS however always caused trouble.  No battery lasted longer than two years.  Once I noticed even sparks flying when reconnecting the battery.  I think the old on-board computer (with its real time clock) drew too much power even when the engine was off.