Balmar SG-200 Battery Monitor

[transmitterdan]

The main reason to keep the shunt close to the battery has to do with the voltage drop across the cable from shunt to battery. That extra voltage drop can confuse an instrument trying to learn about battery impedance. Battery impedance can be an indicator of health and measuring impedance change over time can inform about loss of capacity over time.

[john61ct]

The Merlin SmartGauge's methodology did not involve Ah-counting at all, and it was required to be attached directly to the bank posts.

SG200 uses the same or similar techniques, but they are proprietary. Some have said related to electrochemical impedance spectroscopy, but Balmar refers to it as "Active Impedance Compensation".

The results of those calculations are cross-correlated with the other methodology, traditional Ah-counting to give greater accuracy than just counting coulombs alone.

And the combination of the two is also used to derive the SoH estimates, which members here have been discussing doing manually for years.

[Cat36Mahalo]

Originally Posted by nebster

Can you tell us what it does? (I don't mean the marketing concept, I mean, what element is in the circuit that makes it sensitive to a shunt between it and the battery terminal...)

Balmar probably could have used a normal shunt & gauge, but smartly imho separated the function of the display to be the user interface and the near battery device the “business logic / data gathering”. This is programming 101, Personally I think it would have been better to connect the two using N2K rather than proprietary but that probably would have push the price too high.

By separating the two functions allows two or more displays with one sensor or two or mor sensors with one display. Tricked out I could have one smart shunt on house, start and windlass battery, and a display at helm and Nav table.

[nebster]

Originally Posted by transmitterdan

The main reason to keep the shunt close to the battery has to do with the voltage drop across the cable from shunt to battery. That extra voltage drop can confuse an instrument trying to learn about battery impedance. Battery impedance can be an indicator of health and measuring impedance change over time can inform about loss of capacity over time.

Thanks. I have two responses:

First, the discussion in this thread suggests that cable length is not a factor. Instead, it seems that it's specifically the presence of another shunt that somehow interferes. I'm struggling to imagine why that might be the case, and it's also the reason I asked the question.

(I'm also somewhat skeptical, because having that issue would render the device incompatible with a large array of drop-in lithium packs that each have an integrated BMS... there will be active circuitry in line that no one can bypass.)

Separately, putting the above confusion aside, it seems to me the only way that the location of this instrument in the circuit would matter is if they wanted to save folks the trouble of running one extra sense wire (to the negative battery terminal). Does that ring true to you?

[nebster]

Originally Posted by john61ct

The Merlin SmartGauge's methodology did not involve Ah-counting at all, and it was required to be attached directly to the bank posts.

Let's not confuse the thread by discussing another device.

SG200 uses the same or similar techniques, but they are proprietary. Some have said related to electrochemical impedance spectroscopy, but Balmar refers to it as "Active Impedance Compensation".

What does that mean? Are you suggesting they are injecting AC into the circuit to try to characterize the battery another way?

The results of those calculations are cross-correlated with the other methodology, traditional Ah-counting to give greater accuracy than just counting coulombs alone.

I've written at length how I believe that it would be no problem to analyze voltage versus current over a reasonable time series and arrive at a much more accurate state estimate. But measuring the voltage wouldn't require that zero shunts be in series with the battery for the technique to work.

[nebster]

Originally Posted by Cat36Mahalo

Balmar probably could have used a normal shunt & gauge, but smartly imho separated the function of the display to be the user interface and the near battery device the “business logic / data gathering”. This is programming 101, Personally I think it would have been better to connect the two using N2K rather than proprietary but that probably would have push the price too high.

Yes, I understand the separation of concerns amongst the two modules, and I also share your preference for something that those of us with unified monitoring systems could use rather than a physical display.

But my question is really about the circuit-measuring topology, which is a separate issue.

[lateral]

Far be it from me to question MS's rationale re being closest to battery neg.
So I have rejigged my cabling/ other shunt to accommodate this against my own intuition.
What was not answered by either Chris or MS was the durability of the smart shunt
in the engine bay with regard to temperature & vibration.
Guess I will find out possibly at my expense.
I guess in the rush (now universally typical) to get it to market the finer points of its install are yet to be established.
Normally I wait a couple of years for this to surface and avoid installation not-so-good practices.
MS has established with some certainty the accuracy of the device, not so much how it should be employed in the different setups.
I guess this will be lesson on listening to the experts and not splitting your house bank in two (port & starboard) with heavy cabling between and neutral bus bar in the middle.
(In the engine bay)

[john61ct]

Originally Posted by nebster

Let's not confuse the thread by discussing another device.

No, let's. It is very relevant, the same or similar techniques / algorithms.

The SmartGauge will not work with a shunt in between either, and for the similar reasons.

"Just measuring voltage" is like saying the NSA just uses physical wiretaps on copper wires.

[john61ct]

Originally Posted by lateral

MS has established with some certainty the accuracy of the device

Have you seen any specific numbers or quantified comparison yet?

My impression is so far one of general approval, which of course should be enough for anyone.

But I am keen to see actual data, similar to what Bruce intended in the BM comparison test bench he set up at OceanPlanet.

Obviously would take a lot of time and effort, especially benchmarking against a representative selection of different bank types.

[lateral]

Originally Posted by john61ct

Have you seen any specific numbers or quantified comparison yet?

.

Flooded Lead Acid:

"did another 10.5V capacity test last night and the 130Ah rated DC FLA is delivering 103.53Ah. This puts it at about 79% SOH. The SG200 is displaying 81% SOH."

Another snip for LiFePO4:

"CALB SE-100Ah 4S - Tested Capacity 87.7Ah
All Charging = 0.3C Based on Rated Capacity
All Discharging = 0 .1C Based on Marine Use Simulation
SG200 SOH Currently Reading 88% (actual 87.7% SOH)"

I ripped out my own LINK-PRO and an SG200 in now going to replace it..

https://www.youtube.com/watch?v=iHJ3Qg2hDWY

10mins on---->

[nebster]

Originally Posted by john61ct

No, let's. It is very relevant, the same or similar techniques / algorithms.

It's not relevant if it doesn't require the same (very odd) wiring topology.

The SmartGauge will not work with a shunt in between either, and for the similar reasons.

And what is that reason? I'm still asking, but you're not answering. Instead, you're back to spouting off marketing drivel. Speak to me like an engineer, or, please, just leave the rest of us be to have our discussion and figure it out.

"Just measuring voltage" is like saying the NSA just uses physical wiretaps on copper wires.

You appear to be incapable of either (a) understanding what I am asking, (b) answering the question I have posed, or (c) both.

Instead, we get to enjoy a mixed metaphor from our resident internet "basement know-it-all." Spare us this nonsense and engage on first principles, or just use your mouse to click over to some other thread.

[john61ct]

Really uncalled for.

If I ever get detailed "engineer level" disclosure about the proprietary inner workings of either device, I'll be sure to pass it on.

For what, just to satisfy your curiosity?

That information is just not available, nor is it required in order to properly implement and benefit from the technology.

[lateral]

Originally Posted by nebster

(I'm also somewhat skeptical, because having that issue would render the device incompatible with a large array of drop-in lithium packs that each have an integrated BMS... there will be active circuitry in line that no one can bypass.)
?

...........for future reference!

Above "flooded LA" was from MS. (Just to clarify)

[john61ct]

Originally Posted by lateral

10mins on---->

Excellent stuff there, thanks for posting

[AndyEss]

Can the OP post a photo of the "smart" shunt?
Is there any circuitry/microprocessors/ remote sensing leads on it?
If it's all buried underneath protective covers, obviously don't bother to remove to destroy your warranty.
I don't think any of us want to reverse engineer and market our own devices, but it would be nice to know just why it needs to be the last one closest to the negative post.
Cable length doesn't seem to make sense as an issue, because the gage of the wire matters to a conductors resistance. Also, the mounting instructions would need to specify some nearly exact length to be between the shunt and the negative post.
I appreciate that Balmar's rep is onsite corresponding with us, and that he's a real person, but any mass product supplier these days has to have a digital presence available in near real time to be successful.