Smartgauge w/ 3 batterys

[Exile]

Originally Posted by john61ct

Well the other way to say it is the SG gives an accurate **actual** % SoC based on current reality, not what the owner guesstimates AH capacity may be at the moment.

I've heard of owners with both actually using the SG to try and determine AH capacity by cross-referencing the two, but myself I'd just bite the bullet and do the proper 20-hour load test.

No doubt the SG can be a useful device, but it seems there may be some misconceptions as to how useful.

One feature I was always curious about should you or anyone else know. Let's say you have old batteries and, unbeknownst to you, their capacity is down to 80% from new. So after a full charge does the SG reads 100% SoC, even though that is actually only 100% of 80%? If so, then when you discharge to a SG reading of 50%, would it actually be 50% of 80%, namely 40%?? If this is the case you could unwittingly be destroying your batteries. Or does the SG actually "know" what the remaining Ah capacity is on an older set of batteries?

[lateral]

Originally Posted by Exile

No doubt the SG can be a useful device, but it seems there may be some misconceptions as to how useful.

One feature I was always curious about should you or anyone else know. Let's say you have old batteries and, unbeknownst to you, their capacity is down to 80% from new. So after a full charge does the SG reads 100% SoC, even though that is actually only 100% of 80%? If so, then when you discharge to a SG reading of 50%, would it actually be 50% of 80%, namely 40%?? If this is the case you could unwittingly be destroying your batteries. Or does the SG actually "know" what the remaining Ah capacity is on an older set of batteries?

Don't you have to put in the batteries capacity on SG commissioning?
Then, I guess you have to amend that appropriately as it diminishes?

I find the concept of 20hr capacity test somewhat disconcerting.
Will this be damaging my batteries?
I don't have an inverter or 22.5A load to load up my house bank.
Bit scarey.

[Bleemus]

No, you just put the SG in and within a few charge cycles it settles to current SOC. No need to program it.

[john61ct]

That's right, just choose batt type, FLA, AGM or GEL and you're done.

Important to review Maine Sail's howto, he's got corrections to their wiring for paralleled banks.

[john61ct]

> I find the concept of 20hr capacity test somewhat disconcerting.

Well it does require time and care, obtaining accurate meters and a way to tune the load, like a heater and lights on a dimmer.

> Will this be damaging my batteries?

The key is ensuring it only touches 10.5V at that specific rate, goes no lower and does not stay there, but **immediately** starts charging back up.

Similar process to part of that should be used commissioning any new lead bank to bring capacity up to rated and maximize longevity.

But easy to mess things up and cause irreversible harm, so most would leave to a pro.

Very valuable to assess state of health before a long passage with an older bank.

But as stated no need for the SG, only the shunt-based AH-counting BMs.

[john61ct]

> Let's say you have old batteries and, unbeknownst to you, their capacity is down to 80% from new. So after a full charge does the SG reads 100% SoC, even though that is actually only 100% of 80%? If so, then when you discharge to a SG reading of 50%, would it actually be 50% of 80%, namely 40%?? If this is the case you could unwittingly be destroying your batteries. Or does the SG actually "know" what the remaining Ah capacity is on an older set of batteries?

Yes it does, and exactly the SG's biggest advantage over AH counters, but far from the only one.

Say a nominal 400AH bank is now actually 320.

The AH counter lets you draw 200, down to 120, saying 50% when you're really at 37%! Killing many cycles lifetime every time.

SG says 50% when you've drawn 160, automagically adjusts to the capacity walk-down over time.

For the shunt BMs your choice is guess or do the 20-hour test.

Plus they need to be reset to Full quite frequently, 99% manually.

[Bleemus]

Originally Posted by john61ct

That's right, just choose batt type, FLA, AGM or GEL and you're done.

Important to review Maine Sail's howto, he's got corrections to their wiring for paralleled banks.

Correct. I should have stated you need to program battery type. My mistake. So, yes, a bit of programming is necessary.

[Exile]

Originally Posted by john61ct

> Let's say you have old batteries and, unbeknownst to you, their capacity is down to 80% from new. So after a full charge does the SG reads 100% SoC, even though that is actually only 100% of 80%? If so, then when you discharge to a SG reading of 50%, would it actually be 50% of 80%, namely 40%?? If this is the case you could unwittingly be destroying your batteries. Or does the SG actually "know" what the remaining Ah capacity is on an older set of batteries?

Yes it does, and exactly the SG's biggest advantage over AH counters, but far from the only one.

Say a nominal 400AH bank is now actually 320.

The AH counter lets you draw 200, down to 120, saying 50% when you're really at 37%! Killing many cycles lifetime every time.

SG says 50% when you've drawn 160, automagically adjusts to the capacity walk-down over time.

Ok good. That answers my question & speaks to the SG's usefulness. Thank you.

For the shunt BMs your choice is guess or do the 20-hour test.

No, without a SG your primary telltale would be your voltage meter, right? You can also use your amp counter as a cross-check which will also help assess your capacity.

[Exile]

Originally Posted by john61ct

> I find the concept of 20hr capacity test somewhat disconcerting.

Well it does require time and care, obtaining accurate meters and a way to tune the load, like a heater and lights on a dimmer.

> Will this be damaging my batteries?

The key is ensuring it only touches 10.5V at that specific rate, goes no lower and does not stay there, but **immediately** starts charging back up.

Similar process to part of that should be used commissioning any new lead bank to bring capacity up to rated and maximize longevity.

But easy to mess things up and cause irreversible harm, so most would leave to a pro.

Very valuable to assess state of health before a long passage with an older bank.

But as stated no need for the SG, only the shunt-based AH-counting BMs.

There are definitely some mixed opinions about whether doing a full 20-hour capacity test will harm your batteries, and I personally wouldn't do it. Some portion of a full test -- say 10 hours @10 amps or whatever is appropriate for your bank size and your schedule -- would suffice for me.

And needing to "commission" new batts. by also discharging down to 10.5v and then fully recharging is also a matter of debate. I would check with your battery mfg. first. I think all that Lifeline suggests is an initial charge up to 100% before applying any loads, and even that may not matter in the long run. There was just a recent thread on this topic as I recall.

[john61ct]

yes here tis http://www.cruisersforum.com/forums/...es-190394.html

[ramblinrod]

Originally Posted by john61ct

Even with the SG lower accuracy, it is overall **more** accurate than AH counters.

And much more so, once charging is done.

If this is critical for you, you could briefly interrupt charging.

Or the ideal would be to use both, which many with expensive banks indeed do.

But it takes knowledge, the right tools and regular attention to keep the AH counters accurate, not least regular 20-hour load tests to determine true AH capacity as that walks down over time.

The most useful parameter is ensuring saying well above 50%.

Ideally, stopping charging is done by looking at tail current amps declining.

No, you cannot briefly stop charging and get accurate SOC results based on battery voltage.

In my case, it takes about 15 minutes of 5 A load, to represent "resting" after charge voltage is removed.

Who wants to be draining batteries, when they are supposed to be charging, just to make the monitor read accurately?

[john61ct]

No one is discussing getting SoC from simply reading voltage as a human would, I don't see how that is relevant

SG is running its proprietary algorithms around 20,000 times per second against its historical data store. I certainly doesn't need resting voltage, which btw takes at least 48-72 hours.

The voltmeter display is of course enormously dampened.

[john61ct]

Originally Posted by Exile

No, without a SG your primary telltale would be your voltage meter, right?

Reading voltage in this context does not give accurate enough info to be useful for SoC.

Originally Posted by Exile

You can also use your amp counter as a cross-check which will also help assess your capacity.

There is no accurate measure of capacity other than the 20-hour load test.

If this cycle you benchmarked 100% full with e.g. charging at Absorb until tailing current drops to .005C, then you can know how many AH down from "Full" you are. But you have no way of knowing what SoC% that is, without having recently performed a load test.

[lateral]

Well, sat on the fence long enough.
Ordered my SG. It was either that or a BM which I knew I would never calibrate properly.
My boats been splashed 2yrs & I still haven't calibrated the paddle wheels
or commissioned the AP. No hope for a BM.
Probably result in the same conclusion as Keno and sell it like SoCal.

[ramblinrod]

Originally Posted by john61ct

No one is discussing getting SoC from simply reading voltage as a human would, I don't see how that is relevant

SG is running its proprietary algorithms around 20,000 times per second against its historical data store. I certainly doesn't need resting voltage, which btw takes at least 48-72 hours.

The voltmeter display is of course enormously dampened.

Don't be baffled by BS. So a microP with a slow clock can process math quick. So what? The SG is capable of measuring voltage and time. No more, no less. It cannot perform magic not miracles. As I stated, I can monitor my battery voltage at sunrise and sunset, and know well enough what is going on. Sampling voltage 20 times per second and running 160 algorithms mpay help my batteries last 5% longer.
That would save me $3.14 per year.

For $400 plus installation? No thanks.