Battery BMS not equal to Victron Smart shunt/Cerbo GX

[MerMike]

I'd check the zero current calibration before I looked at changing the shunt sensitivity. i've never had to change that setting either. I have seen users inadvertently hit the zero current calibration button while the system is active.

[jerrypeters]

Thanks again for all the feedback.

I think I have attached the battery settings from the shunt. (Not sure how to do that.)

DVCC is not enabled. I'm not sure how to use it with Kilovault batteries that are not network connected (but I haven't investigated this.)

So ... I measured actual amperage with a clamp meter at the shunt. The results are very confusing. First I disabled all of my solar chargers and tried to stabilize my load. My first reading with whatever load was on at the time was -4.7 on the shunt as reported by Victron Connect and -1.9 on the clamp meter at the exact same time. I then took several other readings with progressively more load added. The results were:

Shunt reading Clamp Meter
-5.2 -2.2
-8.3 -4,2
-19.8 -10.3
-22.4 -11.5

While not exact, it looks like the shunt readings are roughly twice the actual reading from the clamp meter.

I don't believe there is any load present outside of the shunt because the wiring goes directly from two batteries to a Victron Link bus and from there directly to the shunt. There are no other connections.

Also I reset the shunt before I began this process by removing the power feed for a five minute period.

Finally, and somewhat off topic, the Kilovault batteries have an led display which indicates full SOC.

[smac999]

if the shunt readings are twice the real number. you have maybe changed the mv/a shunt settings. you might have set it for a 1000a shunt instead of 500a

if you have a 500a shunt it should be 50mv / 500a

it looks like only the bvm has these settings and not the smart shunt. they may have asseimple the smart shunt with the wrong board. since they make a 500a and 1000a version. you probbaly need a replacment.

[MerMike]

Disconnect the loads from the system side of the shunt and calibrate the zero current value. Before you push the button, check it against your clamp meter and make sure the clamp meter is reading zero. If it isn't something else is going on.

I think a bodged shunt is more likely, but you can try this while you're working on a replacement.

[CarlF]

My Kilauvaults and Victron BMV 712 shunt track very well (Peukert 105). Never more than a few percent off.

[Zanshin]

The "Misc" tab in the app should look something like my settings:


One other thing you might want to do is switch your multimeter to volts and read the voltage difference between the two parts of the shunt. With a known current and a known mV value, you can calculate the shunt's actual value: 50mV / 500A = 0.1mV per Amp. So a 10A load should produce 1mV voltage drop.

If you do have a 1000A shunt rather than the 500A one (which looks likely), you can change the value in the app screen shown above. Both models look alike. The only downside to the 1000A shunt version is that it is only half as sensitive and thus can be somewhat less accurate; but not significantly.

[MerMike]

Originally Posted by CarlF

My Kilauvaults and Victron BMV 712 shunt track very well (Peukert 105). Never more than a few percent off.

What Charge Efficiency setting are you using?

[SailingHarmonie]

Originally Posted by smac999

Measure with a clamp meter. You can zero calibration the shunt if needed. But I have install 100.’s and never had one off that needed calibration. I’d bet it’s wired wrong and you have a wire on the battery side of shunt.

I agree with this, with an additional comment... if there were loads that were connected on the battery side of the shut the current through the shunt would be low not high. Now if the solar panels are connected on the battery side of the shunt, there is your issue. Based on the net size of the error, I'm actually betting that's the answer, or something very similar

I recently redid a Victron monitor that was installed by a "professional" who had no idea what he was doing, and installed it in totally the wrong place. All the solar controllers were on the battery side of the shunt, along with the alternator. Probably cost the boat owner the better part of 3 grand to rewire the boat to make it right after that yahoo go finished with it.

Totally agree that these shuts are really very accurate. There can be a very small zero offset, 10's of milliAmps) but nothing like 2 Amps. If you are seeing a consistent bias on the high side for the shunt, and you are 100% sure the wiring was done right, the next place I would go is the connections. High resistance connections will translate into Amp readings that are higher than they should be. Anything that increases the voltage drop across the shunt will add to the error. Even things like putting the stainless steel washers UNDER the cable terminals will cause problems. (Yes, people do this, for some odd reason...) and even a little bit of corrosion can make a mess of things.

This should be easy to measure... which one is right? Then why is one wrong? Assuming a design fault in the shunt is very, very far down the list of possible causes.

[jerrypeters]

Charge efficiency rating set at 99%

For some reason my "Misc" screen on the shunt app does not provide shunt specs and therefore I have no way to change the shunt from 500 to 1000. Not sure why my app is different or how to spec the shunt model on the Cerbo. More research ...

[smac999]

Originally Posted by jerrypeters

Charge efficiency rating set at 99%

For some reason my "Misc" screen on the shunt app does not provide shunt specs and therefore I have no way to change the shunt from 500 to 1000. Not sure why my app is different or how to spec the shunt model on the Cerbo. More research ...

As I posted above. I think only the bvm has this. Because you can wire different shunts to the screen. The smart shunt you buy one model or the other.

I think you need to exchange it for a good one.

[Capt.Don]

JP -- is the Kilovault app showing 1/2 amps of the Victron SmartShunt?

Do you have a known load, like a string of 12v incandescent light bulbs in series (interior house light?)? Assuming each bulb is 1.5a (which you can check), 4 bulbs would give you a known 6 amp draw? Disconnect everything else and compare.

Don

[Zanshin]

@Jerrypeters - Is there any update on your problem?

[JBP]

Originally Posted by MerMike

I'd check the zero current calibration before I looked at changing the shunt sensitivity. i've never had to change that setting either. I have seen users inadvertently hit the zero current calibration button while the system is active.

It certainly sounds like you need to check and reset the zero current calibration as MerMike said. Also turn on automatic synchronization so the Smart Shunt resets to 100% when the battery is fully charged. I've found that my smart shunt does creep if not synchronized occasionally (though my SOC creeps up, not down).

[jerrypeters]

Once again, thanks to all for the help.

Here's an update ... The problem with the shunt vs bms has reversed itself! Now the shunt is under reporting load and overreporting SOC. I did nothing to change anything. It just changed by itself in the last two days.

For example Tuesday morning at 6AM the SOC reported by the Cerbo was 97% and the SOC on the battery bms was 91 after a whole night of no charging. By 9AM that day with minimal solar, the Cerbo SOC was 100% and the bms was 94%. At the same time (9AM) the current on the Cerbo was +3.7 and the bms was +1.6. Wednesday at 6AM a similar result – Cerbo =96% and -4,8 amps, bms=92% and -6.6amps. This Is exactly the opposite of the problem of the last several weeks/months!!??!.

I did not have time to do a clamp meter test on the shunt but I suspect the clamp meter would be higher than the shunt. This makes no sense but I think still points to a shunt problem. It’s probably a good idea to replace it and continue to monitor.

Unfortunately, I had to leave the boat and won't be able to follow-up for a few months. I'm going to replace the shunt and I'll see what happens and report back to the forum.

Thanks much again.

[CarinaPDX]

The problem is that everyone here is still thinking like FLA batteries. With an FLA battery the current going into the battery during charging is the charge. With these integrated BMS+cells batteries the current going into the batteries is not all reaching the cells. There are at least two or three circuits used by BMS devices so it is hard to say exactly where the power is being lost, but consider this: there are 4 cells in each string, and when one cell in the string is fully charged it must be disconnected from the charge current, while the other cells continue to charge. One solution is to switch a parallel shunt across the charged cell, which of course would convert current to heat. I have to admit that I am no expert, and trying to work out the various BMS circuits is mind-numbing for my old (EE) mind, but suffice it to say that the charging circuits inside the case are taking some of that missing power. Whether there are other explanations I don't know, but I do know that just because it looks like a battery doesn't mean it acts like one when it contains a BMS.
\
Greg