Smartgauge w/ 3 batterys

[Stu Jackson]

Originally Posted by Dockhead

>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
I'm amazed the amp counting meters don't use this to recalibrate themselves. But likewise, this additional data would fill in ALL the gaps in what Smartguage knows.>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>

Perhaps I'm misunderstanding, Dockhead, but my Link 2000 does exactly that by using the kWhr counter, assuming you've corrected the default charging voltage and %.

As noted in my Gotcha article:

It WILL eventually register full because the way it works is that it does count the AH down by using the kWHr function internally. The "trick" here is ONLY to prevent it showing full prematurely with the "default" 13.2 V and 2% "calculators," and not never ever showing "full." It will show full when the kWHrs counts to zero, and WON'T do it PREMATURELY with the float voltage and percentage of charge "method."

Showing full means it will reset itself.

Also included in that article is a post from Maine Sail about how he resets his coulomb counter manually, weekly, when he knows his bank is fully charged. I do that, too.

[sailinglegend]

Originally Posted by Dockhead

....... you need at least amps - amps and voltage give very accurate SOC during charging, maybe most accurate from any method. I'm amazed the amp counting meters don't use this to recalibrate themselves. .....

Amps and volts will give a very accurate 100% SOC - BUT - this is at the absorption voltage of 14.4v and at a current of 0.5% of the battery Ah capacity.

You're amazed that the amp counting meters don't use this to recalibrate????

But you should know that any charger will have dropped to float long before the current is 0.5% of the battery capacity. Chargers are designed to charge a battery safely, not to overcharge a battery, which is why they all drop to float well before a battery is fully charged. At the lower float voltage no gassing occurs and the battery will finally get to 100% - but it takes a very long time.

The only way to check that a battery is full is to turn the charger off and back on to force it back into absorption mode at 14.4v and keep doing that until the current is 0.5% of battery capacity.

This is why SmartGauge is so smart.

[conachair]

Originally Posted by sailinglegend

Amps and volts will give a very accurate 100% SOC - BUT - this is at the absorption voltage of 14.4v and at a current of 0.5% of the battery Ah capacity.

Sounds very authoritative - where did you get the 0.5% figure from?

I emailed Trojan about this & the reply was "The finishing rate for a fully charged battery is 1-3% of the total capacity of the battery bank (2.25A - 6.75A)."

Though I'd probably put it a bit lower for these batteries on this boat. At a guess it would vary a little with battery types & age.

[ayates]

Originally Posted by noelex 77

I posted a review of the Simarine Pico monitor, which may be of some interest here:
http://www.cruisersforum.com/forums/...147499-78.html

Brilliant! Thanks ever so much! That unit looks sweet. I have been looking, without success, for a unit that would log multiple DC current channels.

The thread page with your Pico review is actually: http://www.cruisersforum.com/forums/...47499-103.html


Allan.

[john61ct]

Originally Posted by conachair

Sounds very authoritative - where did you get the 0.5% figure from?

I emailed Trojan about this & the reply was "The finishing rate for a fully charged battery is 1-3% of the total capacity of the battery bank (2.25A - 6.75A)."

Though I'd probably put it a bit lower for these batteries on this boat. At a guess it would vary a little with battery types & age.

Each vendor has its own spec.

Flooded batts are very robust, don't get harmed by going over (lower amp rate).

More sensitive chemistries, should stop sooner (higher amp rate).

If you made the mistake of buying from a vendor that doesn't make it easy to get such specs, .005C for flooded, .02C for others is generally fine

[john61ct]

Originally Posted by ayates

Brilliant! Thanks ever so much! That unit looks sweet. I have been looking, without success, for a unit that would log multiple DC current channels.

The thread page with your Pico review is actually: http://www.cruisersforum.com/forums/...47499-103.html


Allan.

That link didn't work for me either

[john61ct]

Originally Posted by sailinglegend

any charger will have dropped to float long before the current is 0.5% of the battery capacity.
…
The only way to check that a battery is full is to turn the charger off and back on

Not true for all charge sources.

And you can always rig turning them on and off from volt sensing or SoC.

[Stu Jackson]

Originally Posted by sailinglegend

1. Amps and volts will give a very accurate 100% SOC - BUT - this is at the absorption voltage of 14.4v and at a current of 0.5% of the battery Ah capacity.

2. You're amazed that the amp counting meters don't use this to recalibrate????

3. But you should know that any charger will have dropped to float long before the current is 0.5% of the battery capacity. Chargers are designed to charge a battery safely, not to overcharge a battery, which is why they all drop to float well before a battery is fully charged. At the lower float voltage no gassing occurs and the battery will finally get to 100% - but it takes a very long time.

4. The only way to check that a battery is full is to turn the charger off and back on to force it back into absorption mode at 14.4v and keep doing that until the current is 0.5% of battery capacity.

5. This is why SmartGauge is so smart.

1. Agreed.

2. They do and will when you change the DEFAULT values. Heck, they hadda pick two numbers, that's what they did. So, just change them. It really is easy. [This is not for you personally, but for readers of this topic who aren't aware of it. For folks who haven't read The Gotcha and Maine Sail's article.]

3. Yes, but... Most chargers drop to float because their algorithms say do it based on an egg timer approach. Another reason to restart the charger to find amps in AT absorption voltage, not float voltage.

4. Yup.

5. Your boat, your choice.

[Dockhead]

Originally Posted by Stu Jackson

Perhaps I'm misunderstanding, Dockhead, but my Link 2000 does exactly that by using the kWhr counter, assuming you've corrected the default charging voltage and %.

As noted in my Gotcha article:

It WILL eventually register full because the way it works is that it does count the AH down by using the kWHr function internally. The "trick" here is ONLY to prevent it showing full prematurely with the "default" 13.2 V and 2% "calculators," and not never ever showing "full." It will show full when the kWHrs counts to zero, and WON'T do it PREMATURELY with the float voltage and percentage of charge "method."

Showing full means it will reset itself.

Also included in that article is a post from Maine Sail about how he resets his coulomb counter manually, weekly, when he knows his bank is fully charged. I do that, too.

If I'm understanding you right, your Link (I had a Link monitor too, many moons ago) will not show 100% as long as there is current still going into the battery of more than 2% * C at 13.2V. Is that correct?

That's surely a good feature -- I don't think my Victron one did that. But what I was talking about is different -- during the whole absorption phase you can tell precisely what is the SOC by the amount of current at a given voltage. So you could compare that against %SOC you have calculated by another method (either amp counting, or voltage in case of SG) and calibrate it. AFAIK, no battery monitor does this, and it seems to me that it would add a tremendous amount of accuracy. For amp-counters, you are not just calibrating the full point -- you are calibrating a range of different SOC values. I would think that you could even calculate the real capacity of the bank this way.

[Dockhead]

Originally Posted by sailinglegend

Amps and volts will give a very accurate 100% SOC - BUT - this is at the absorption voltage of 14.4v and at a current of 0.5% of the battery Ah capacity.

You're amazed that the amp counting meters don't use this to recalibrate????

But you should know that any charger will have dropped to float long before the current is 0.5% of the battery capacity. Chargers are designed to charge a battery safely, not to overcharge a battery, which is why they all drop to float well before a battery is fully charged. At the lower float voltage no gassing occurs and the battery will finally get to 100% - but it takes a very long time.

The only way to check that a battery is full is to turn the charger off and back on to force it back into absorption mode at 14.4v and keep doing that until the current is 0.5% of battery capacity.

This is why SmartGauge is so smart.

As far as I know, not only at 100%, can you get very accurate SOC from volts and current.

I'll try to remember the source for this idea, and post it.

[Dockhead]

Originally Posted by Dockhead

As far as I know, not only at 100%, can you get very accurate SOC from volts and current.

I'll try to remember the source for this idea, and post it.

Edit: I think I have been posting nonsense -- please ignore what I said about this.

Here is something interesting: https://ogst.ifpenergiesnouvelles.fr...ogst110109.pdf

Indicates that charge acceptance is not constant, but changes a lot according to recent charge history.

[noelex 77]

I am not sure why the link is not working for some. I have pasted the information below to make it easier to find. Sorry for the duplication just ignore the post if you have already seen this:

Battery Monitor: Simarine Pico.

If you are an electronics geek (like me) you are going to love this device.

The battery monitor functions in an unusual way, but there will be more about that in another post. First I wanted to outline its capabilities. It will do things no other battery monitor will do. Here is a brief list (bear in mind some of these options require additional shunts etc):

Monitor 6 battery banks
Display current information (such as solar panel input, fridge output etc etc) from 24 different shunts.
Display temperature information from 10 different locations
Display information from 6 tank level sensors
Display atmospheric pressure (including a nice graph)

These inputs can all be individually named and are displayed in digital format, as well as an analogue bar graph that can be tailored for the individual input. So the analogue display will show full deflection at the maximum expected value. There is also adjustable minimum and maximum alarms that can set for most parameters as well as data logging and calibration for many inputs. The data can also be sent via wifi and displayed on a tablet or smartphone.

I cannot imagine anyone will be using all these options, but my planned installation might give some impression of how this might work practice, although initially not all the capabilities listed below will be installed.

The main home page will display:
House battery voltage
House battery SOC
Net current into or out of the house battery
Solar panel input for each of the three solar panels
Fridge current output

Further pages will display:
Start battery voltage
12v system voltage
Barograph (in graph form)
Temperature:
It would be easy to go overboard here but I can see the following as potentially useful:
Air
Water
Battery
Fridge
Engine
Engine exhaust
Engine alternator
Some of the above would befit from alarms, which are easy to set.

Other options include monitoring the alternator output although this will require the purchase of an additional shunt.

So how does it perform?
I have only set it up on test bench so far so a detailed performance review will have to wait, but the initial impression is very favourable. The display, while small, is gorgeous and the software is excellent, enabling extensive customisation while keeping things relatively simple.

The battery voltage is shown to 2 decimal places (much better than one) and the reading is spot on. Many battery monitors get even this simple but important measurement wrong. Errors of around 0.1v or more are not uncommon.

Shunts of 500A (and more) are available which can be important for larger boats.

Drawbacks:
It is expensive. There are also so many capabilities that it is tempting to make the installation complex. There needs to be some thought whether the information will be useful rather than just cool to have.

The main question is how it performs as battery monitor. I talked to the developer at the Dusseldorf boat show several months ago and he is a very smart guy who understands battery technology and what is needed to design a good battery monitor. The monitor is sophisticated and works in a slightly different way to most other battery monitors (more about this in a future post). It will be interesting to see how this translates in practice.

A bit of background on battery monitors may help put the Pico into perspective.

There are two quite different approaches to producing a battery monitor.

The first monitors battery voltage and importantly the current going into (or out of the battery). In the simplest form this type of monitor counts the amp hours going into and out of the battery and uses this to display the net amp hours that have been drawn from the battery. If you enter the battery size it can also estimate the net state of charge in percentage terms. The better models are more sophisticated and take into account Peukert's law which simply means the battery capacity diminishes more rapidly than might be expected when there is a high current draw relative to the battery size. They also take into account battery efficiency (not all of the current entering the battery is converted into capacity. Unfortunately, this changes with voltage). A small number of the better models also use battery temperature as this effects battery capacity. As even a sophisticated counter will gradually drift out of synchronisation over many cycles, there also needs to some means of automatically zeroing the meter, usually when the monitor recognises that the battery is fully charged. How this is done varies.

The above brief explanation shows this type of meter is complex, and not all models are equally good. In my view, the better models are accurate, but they need to be set up well both physically with the correct wiring, and you need to enter battery details such as the nominal capacity and the particular Peukert's exponent of your battery type. Unfortunately, it is rare to see one set up correctly.

The second type of battery monitor works in a very different way. It does not measure current or temperature only voltage. Wiring is simple and there is no need to enter many details about the battery. Exactly how these monitors can convert simple voltage information into a meaningful display of the state of charge is never fully explained by the manufacturers, for obvious reasons. There are several possible mechanisms that could be used, and it is important to realise that these are not simple voltage displays, but devices using complex algorithms to determine the state of charge. The main advantage is that they are very simple to install and use. They take some time to "learn" any particular battery installation, but the manufacturers point out this only means the accuracy of the reading is constantly improving. The biggest drawback is there is no current display, which is unfortunate as this is very useful information in its own right. The most popular model that works in this manner is the SmartGauge.

Both these very different types monitors have very adamant supporters.

The obvious question is why not combine these different approaches in a single monitor? This has the potential of producing the most accurate battery monitor and overcoming some of the limitations of both approaches.

Surprisingly no one, to my knowledge, has previously combined these two methodologies. A couple of companies have hinted their products incorporate some "special" technology, which makes them different from conventional amp hour counting monitors, but I suspect most of this is advertising blurb.

The Pico is a little different. It has what appears to be a very sophisticated conventional amp hour counting technology, but will also work as a battery monitor with no shunt attached at all. You would be silly not to use a shunt. Simarine point out if the monitor is used in this way the results will be less accurate, but the fact that it works at all indicates it is also capable of monitoring the battery state of charge using the second methodology. Simarine also indicate their monitor works better as it "learns" the battery system, once again indicating the monitor is using elements of the second approach. Intriguing.

A battery monitor that uses both methodologies, the best of both worlds?

Perhaps this is the answer to settling those forum threads on battery monitors that can sometimes become a little too passionate?

Anyway, at this stage I have been very impressed with the monitor, but a propper evaluation will take some time. As I use the equipment more it might also be clearer exactly how the battery algorithms function so take the above comments as provisional rather than definitive.

Also keep in mind you can still monitor batteries with a simple dumb voltmeter. This method has some limitations, but is by far the cheapest option. If you have a simple electrical system it is an option worth considering.

[Viking Sailor]

A nice aspect of using the SG with an Ah counter is the easy of calculating the Ah capacity of the battery.

If for example you zeroed the Ah counter when the SG showed 100% SOC. Then when the SG showed 50% SOC, the battery capacity would be double the Ah counter reading.

Even better, with the SG you can calculate the battery capacity knowing only the change in SOC (Delta-SOC) and the Amp-hours consumed (Delta-Ah).

Ah-bat = Delta-Ah * (100 / Delta-SOC)

Should you not have an Amp-counter handy - you could apply a fixed load to the battery until there is a 10% or so change in SOC. By measuring the time (in minutes) for the change in SOC to occur, and knowing the Amp's the load draws the Delta-Ah value can be calculated.

Delta-Ah = Amp-Load * (Time / 60)

This Delta-Ah value can then be used in the above calculation to get the capacity of the battery bank in Amp-hours.

[sailinglegend]

Originally Posted by Viking Sailor

A nice aspect of using the SG with an Ah counter is the easy of calculating the Ah capacity of the battery.......

This is exactly what Merlin are doing with their Data Cell 11 technology on the Abrahms tanks. The SmartGauge measures the SOC and a shunt based monitor allows them to calculate the State of Health (SOH). SOH has been almost as important as the SOC because the inconvenience and cost of replacing a battery in the battlefield is hugely expensive.

With an Ah count you have to be sure that the Peukert Constant in the BM is roughly correct for the age of the battery. This is difficult to know correctly and because it's a logarithmic value a small error can make a big difference. A 20 hour discharge gives a more accurate figure but your method will give an idea of any serious capacity loss.

Knowing the SOH at the beginning of the season is also important for cruisers because whenever a bank dies its usually difficult or very expensive to get the batteries you really want.

[Viking Sailor]

Originally Posted by sailinglegend

This is exactly what Merlin are doing with their Data Cell 11 technology on the Abrahms tanks. The SmartGauge measures the SOC and a shunt based monitor allows them to calculate the State of Health (SOH). SOH has been almost as important as the SOC because the inconvenience and cost of replacing a battery in the battlefield is hugely expensive.

With an Ah count you have to be sure that the Peukert Constant in the BM is roughly correct for the age of the battery. This is difficult to know correctly and because it's a logarithmic value a small error can make a big difference. A 20 hour discharge gives a more accurate figure but your method will give an idea of any serious capacity loss.

Knowing the SOH at the beginning of the season is also important for cruisers because whenever a bank dies its usually difficult or very expensive to get the batteries you really want.

The Peukert Constant is only relevant during charging. It accounts for the chemical energy needed and heat energy generated during charging.

The method I described applies only during battery bank discharge. Thus, the Peukert Constant does not apply to this method.