Victron monitor defaults and programming

[booster]

There have been several discussions on Victron monitors, installing, using, etc lately but little discussion about how to know what the monitor is telling you and/or what it is not.

The Victron 712 Smart seems be getting popular because of the Blue Tooth to phones, so I looked up the manual for it recently.

https://www.victronenergy.com/upload/documents/Manual-BMV-700-700H-702-712-EN-NL-FR-DE-ES-SE.pdf

The biggest question to me was what the default values in monitor as shipped were, as that will be a huge determiner of how good the information is that you get out of it. We have heard that the monitor only really asks for battery bank 20 hr capacity, and that the defaults are good to use as is, so looked for the actual numbers.

Here is what it says in the quick start guide.

The factory settings are suitable for the average lead acid battery: flooded, GEL or AGM.The BMV will automatically detect the nominal voltage of the battery system immediately after completion of the setup wizard (for details and limitations of automatic nominal voltage detection, see section 3.8). Therefore the only settings which need to be made are the battery capacity (BMV-700 and BMV-700H), and the functionality of the auxiliary input (BMV-702 and BMV-712)

Here is the specification that I was looking for, describing what determines a fully charged battery and synchronizing of the monitor back to full (100%).

For a reliable readout, the state of charge as displayed by the battery monitor has to be synchronised regularly with the true state of charge of the battery. This is accomplished by fully charging the battery.In case of a 12V battery, the BMV resets to ‘fully charged’ when the following ‘charged parameters’ are met: the voltage exceeds 13.2V and simultaneously the (tail-) charge current is less than 4.0% of the total battery capacity (e.g. 8A for a 200Ah battery) during 3 minutes.

My first impression of this was that the Victron will be telling you that the batteries full when they are quite a ways from being that full. My guess was that they would be at about 80% full, but that was just a guess so I decided to test it on our system, as the controls we have make it pretty easy, but time consuming, to do.

We have a Lifeline bank of four 6v batteries totally 440ah at 12v, charged and monitored on a Magnum MS2000 charger, ARC50 remote, and BMK monitor kit. All the settings and monitoring are done on the ARC50 remote. The Lifeline recommended specs for a full battery are hugely different than the Victron defaults, but very close to what other AGM batteries are. For Lifeline those specs are over 14.3v and under .5% amps to the batteries


I started with the batteries charged fully to Lifeline specs, and then floated for two days, so know full and stable. The monitor was reading the 13.2v float voltage and .1 amps to the batteries, so 1/180th of what the Victron spec would indicate as a full battery, and typical of what we have seen on these batteries since new. That is a lot of difference and really mades me wonder if I am looking at something wrong.

I decided to continue the test anyway, and set the charger to 13.2v bulk/absorption and 13.0v float with 4% (440X.04) tail amps set, which came out 17.6 amps rounded to 18 amps as no decimal programmable for that setting.

I ran the batteries down slowly to 88% SOC and let them rest an hour, then hooked up shore power. It settled at barely 5 amps to the batteries at 13.2v, so way under the 18 amps.

Unhooked shore power and ran the batteries down to 75% SOC and let them rest. Plugged in shore power and within a couple of minutes the amps were under 8 amps at 13.2v, still way under the 18 amps that the full battery point would be on the Victron.

I unplugged shore power and ran the batteries down to 63% and they are now resting, so will know more in an hour.

So far, it appears that the Victron would be indicating a full battery at a much lower SOC than I have guessed, but how low is still to be determined. I will continue dropping SOC until at recharge I get above 18 amps going to the batteries. That 18 amp point at 13.2v would be the SOC on our batteries that the Victron would indicate the batteries being full on a recharge cycle if it is being charged at 13.2v.

 

[markopolo-ClassB]

The only circumstance that using 13.2V & 4% tail current to indicate a full battery works out (that I can think of) is if plugged into the grid and if a properly configured charger continues through it's full charge cycle.

There should be a strong emphasis in the manual about changing the defaults to actually meet full charge criteria to allow the monitor to supply useful information.

The manual does state this:

The BMV synchronises too early

In solar systems or other applications with fluctuating charge currents, the following measures can be taken to reduce the probability for the BMV to reset prematurely to 100% state of charge:a) Increase the “charged” voltage to only slightly below the absorption charge voltage (for example: 14.2V in case of 14.4V absorption voltage).b) Increase the “charged” detection time and/or decrease the tail current to prevent an early reset due to to passing clouds.

However, even that seemingly still places the importance or value of getting the reset above actually getting the batteries fully charged. The point of a battery monitor should be to provide useful information to help you ensure that the battery gets fully charged.

IMO, it would be better to start with 14.3V & 0.5% or 14.4V & 1% and then explain how to lower the criteria if you have a persistent load or don't want to upgrade to charge equipment to actually fully charge the battery.

 

[booster]

Totally agree with Marko's assessment.

I got the final results of the finding the 18 amp charge rate at 13.2v test.

Ran it down to 63% state of charge and amps at 13.2v jumped all the way up to 50 amps, but after a an hour or so started to drop quickly. At 70% SOC the amps read the calculated 4% at 17.6 amps, so that is the point that the Victron would say the battery is full. That is about 133ah down on our system, which is nearly 3 days of power use for us. There is no way that I would consider being that far off as "good" or even "good enough" or "not too bad". Our usable capacity to 20% SOC is 352ah, so we only have 219ah of power at that 70% point. 37% off.

I decided to see what would happen at the correct absorption voltage but still the 4% tail current, so I just turned up the absorb voltage to 14.3v and kept charging. Amps jumped to 69.4 amps at 13.68v (charger set at 70 amps so not to absorption yet). I will let this run until it is at the 14.3v and 17.6amp (4%) and see what the SOC is there. It will be higher, for certain, but how much I don't know.

In the real world of shore charging, what I am testing is likely what would happen in some systems. If the charger is big enough vs the battery capacity, it will go past the 13.2v point without the amps dropping to the 4% and rise until it hits it's setpoint for absorption. The amps will fall until 4% point is reached at that voltage as the battery fills more.

 

[booster]

I just finished the last part of the test, which was to get the charger up to full absorption voltage at 14.3v, and then see how full the batteries were at the 4%C Victron default. This would be a best case scenario of what could happen in the real world, I think. Be aware that the SOC given is likely a bit high, as charge efficiency can enter into it when we get above 80% state pf charge so it could be maybe 5% less than stated, or not.

What I got was the 17.6amps to the batteries at 14.3v, with the SOC at 92%, more likely actual of about 90%. This would leave your batteries 10% undercharged when the Victron said they were full, which is in our case about one day of use. Charging only to there will also cause battery capacity to walk down over time, shortening battery life.

That said, for many folks that may be even better than their charger will reliably deliver if it runs on timers or an algorithm. It will always be above or below optimum because it has no idea what the SOC of the battery is at any given time.

Probably the worst situation would be if you have a smallish charger that would go to float very early, as in that case you would get near only being able to get to 70% SOC before the Victron said you were full.

 

[GallenH]

This is all very useful. NOW I need to figure it all out! But thanks for the detailed post.

 

[JohnnyCLE]

Useful vs Interesting?

Since I am trying to find my way around two of these, one for the 24v ground and the other for the 12v ground on an E-Trek, I found this discussion enlightening. It had never occurred to me that I would find a devices "best guess" at when my batteries were "full" or "empty" as "useful" information, but merely "interesting".

I always expected that my batteries would be "full" when the charging equipment I have at my disposal has stop "absorption" and gone to "float", excluding manual ways to game the timer based system by disconnecting and reconnecting shore power to get proper absorption times. The batteries would be "empty" when my inverter reached it's low voltage limit or the CO monitor started to beep.

What I am finding immensely "useful" is that I now have a reading of how many AH out and back in between these "full" and "empty" events, what the batteries are actually seeing from different charging and demand sources, and are there current "leaks" that I am unaware of.

Data by which I can measure the performance of the system is useful, an algorithms opinion on state of affairs would seem to only be interesting to me. While I have yet to try, I am expecting to "Synchronize" the top limit of my chargers ability as "full" and to substitute my actual measured AH capacity for the rated factory capacity to see more realistic % of capacity left readings as the monitor tracks discharge. All the time I will have in mind that this is an interesting "best guess" perhaps adjusted for temperature and discharge rate, but likely not.

It will be interesting to see how good the algorithms are depending if capacity was trickled out at 12v or pulled out at 110v appliance rates. I guess, with reliable and consistent accuracy over time, I may come to consider the monitors opinion related to capacity left as useful rather than only interesting.

 

[booster]

One thing to remember as you are trying to determine you actual capacity is that you have to get them full first to be able to tell how much they are holding on discharge, and therein lies the problem with the mixed system.



Are you trying to determine you maximum possible capacity, or the maximum that your charging equipment will do in the real world? Those are likely two different numbers, maybe significantly.


Also be aware that you can't check capacity on the recharge cycle because of the charge efficiency which makes you have to put in more AH on recharge than you get from the discharge cycle, which is accurate. Charge efficiency varies with a lot of things, so pretty near impossible to predict. A bit issue is that it varies with depth of discharge before the recharge, so every cycle is different unless discharged the same.


It will be interesting to hear about what you come up with for usable capacity and repeatability, as well as how good or bad the charging equipment is. You may want to start by seeing if you can find someplace the describes the charge cycles and how it is calculated by the chargers, so you know what to look for.


Most of this speaks to why there have been recommendations to ditch the split systems and take it to pure 24v with a b to b charger for the 12v power. That way it behaves like one big battery with one monitor that sees everything, so would much better to be able to know what is going on.

 

[MagicBox]

no charge

Thank you for this thorough research, stressing your batteries in the name of science.
I was about to pull the trigger on a victron because it seemed popular and well received, had bluetooth, didnt look like a cold war era artifact, and only requires an easier hole saw cut iinto a panel instead of a trickier rectangle incision (an actual concern of mine).

Dont know much about SOC.........

Eric

 

[JohnnyCLE]

Attached are the monitor readings from my first load test. Running from "full", as defined by "as full as my current charging system can get it" to the low battery warning and cutoff on the inverter. I was pleasantly surprised by the roughly 900 6v AH I was able to get. I had expected much worse after my neglect of the batteries the past two years.

This load test was done by running the inverter and air conditioner with no other loads other than the 12v battery connect so that the thermostat operated. Started at 3:30 and ran for roughly 3 hours 45 min. There was tree shade so the solar had no effect. It was a hot and muggy day so the AC ran at an amazingly consistent 1400 watts.

The life left on the monitors shows that the Victron dealers best guess on how capacity should be split needs to be adjusted!

 

[markopolo-ClassB]

The 24V voltage indicates less than 30% capacity plus the low battery warning all appear to confirm that the 24V capacity shown at 62% is very wrong.

 

[JohnnyCLE]

Yes, the initial values input for each battery bank were guesses about how rated capacity would flow under load to the respective grounds. The E-Trek has a 12v ground in the engine compartment and a 24v ground in the rear battery compartment, but the batteries are two series strings, cross bridged, and tapped at the two voltage levels. I plan to adjust the projected capacities based on the actuals from this test.

How do you get 30% residual or 70% discharged? Temperature at discharge was roughly 85 degrees F. Averaging values from Appendix C of the Lifeline Technical manual suggests a 3.5 hour discharge rate ending voltage of 11.8v-12.0v. at roughly 50% discharge. The chart graphs are not very precise. The half the 24v reading of 23.49v is 11.75v. I read that as maybe 45% residual or 55% discharged.

 

[markopolo-ClassB]

I haven't tried to confirm your conclusion but suspect that you're looking at SOC data while the batteries are still under load.

The 24V side is not actually at rest as a small amount of current is going into the batteries but the OCV at rest SOC chart indicates even lower SOC than I guessed at.

 

[booster]

Yes, the initial values input for each battery bank were guesses about how rated capacity would flow under load to the respective grounds. The E-Trek has a 12v ground in the engine compartment and a 24v ground in the rear battery compartment, but the batteries are two series strings, cross bridged, and tapped at the two voltage levels. I plan to adjust the projected capacities based on the actuals from this test.

How do you get 30% residual or 70% discharged? Temperature at discharge was roughly 85 degrees F. Averaging values from Appendix C of the Lifeline Technical manual suggests a 3.5 hour discharge rate ending voltage of 11.8v-12.0v. at roughly 50% discharge. The chart graphs are not very precise. The half the 24v reading of 23.49v is 11.75v. I read that as maybe 45% residual or 55% discharged.

I seem to remember seeing in the past that the front batteries were connected to the rest of the bank, but tapped for 12 to the separator? If so, there really is not a difference in the grounds as both would likely be on 24v circuits for analysis stuff, I think. It gets to be tough to understand because the meter is going to read the amps, and has no idea if it is 12v coming through or 24v. With all the cross tie points and connections, I don't know if voltages to determine state of charge of individual batteries would be even close to accurate. It would be a lot of work, but you may need to disconnect all the batteries to see where they sit for voltage, after resting a few hours, you might find some huge variations that are making your other tests even more confusing. Best thing to do probably would be to capacity test them all individually, but that would be even more work and you would need a charger that you could make sure stayed at full absorption voltage for as long as you need to get them full, and a good ammeter.


The voltage to stat of charge charts are, at best, only fair at knowing actual SOC. If you have new batteries, already broken in as they gain a bit of capacity over the first few cycles, the charts will be close, but as the batteries age the charts get more and more inaccurate. I had a small AGM for a burglar alarm, that I also used on the bench for 12v power when needed, that would charge to 12.6v and even hold it a while when resting. Capacity test showed it had less the 20% of it's rating so was totally shot, but passed voltage test at 95+%.


Without full instrumentation on it, like ammeters and voltmeters on every cable and connection, it may be near impossible to truly figure out what exactly is going on.


One question I have kept forgetting to ask. Is the separator connection to the front batteries the only connection to second alternator?

 

[JohnnyCLE]

Yes, the batteries were definitely not "at rest". The 12v loads were still connected and the inverter was still on and chirping, but in a "shutdown" state.

It is a mystery as to where the positive current flow was coming from. There was no shore power or solar. Perhaps from some of the cross bridging or just erratic data. The inverter is programmed to shut off at around 52% discharged. That is what it's LCD display was showing and the lights above the slider were display 2 out of 4 lights still on. The solar controller voltmeter was behaving erratically, displaying a succession of voltages, some which correlated with the Victron.

 

[JohnnyCLE]

Yes, the "Smart Solenoid" is the only connection to the chassis battery and it is connected to the 12v tap in the engine compartment. I would have expected to see some current flow into the 12v tap through this "Smart Solenoid" separator connection. Instead there was continued discharge at the 12v point and positive flow at the 24v point.

I did not wait around long to see how this state would persist over time. I started the engine to capture the aux generator flow and then repositioned the RV to plug into shore power.

One positive side effect of this exercise was that, probably for the first time, these batteries received a proper absorption charge. Normally when connected to shore power the batteries would have been fully bulk charged by the aux generator and the charger would default to one hour of absorption charge before going to float.

 

[booster]

Yes, the "Smart Solenoid" is the only connection to the chassis battery and it is connected to the 12v tap in the engine compartment. I would have expected to see some current flow into the 12v tap through this "Smart Solenoid" separator connection. Instead there was continued discharge at the 12v point and positive flow at the 24v point.

I did not wait around long to see how this state would persist over time. I started the engine to capture the aux generator flow and then repositioned the RV to plug into shore power.

One positive side effect of this exercise was that, probably for the first time, these batteries received a proper absorption charge. Normally when connected to shore power the batteries would have been fully bulk charged by the aux generator and the charger would default to one hour of absorption charge before going to float.

So the two front batteries are connected to the starting battery and alternator, and the coach batteries are connected to second alternator?

 

[booster]

You may want to start a new thread on your troubleshooting so it doesn't get all mixed in with this one which will hopefully be able to be used by others that have a more standard monitor installation process going on. I think it would be best for both discussions.

 

[markopolo-ClassB]

In terms of energy used (watt hours) during the air conditioner run, it looks like 1/3 came through the "12V" tap point ground and 2/3 through the 24V ground.

For the monitor settings, capacity might be calculated as follows:

Using 220Ah capacity rating at 6V x 8 batteries -> 10,560 watt-hours

7040Wh through 24V ground point
3520Wh through 12V ground point

7040/24 -> 293Ah 24V capacity
3520/12 -> 293Ah 12V capacity

You could reduce those equally by the expected capacity lost due to age & previous care.

The 1/3 & 2/3 usage probably should be confirmed using a lower rate of discharge than the air conditioner to see if it is consistent.

 

[markopolo-ClassB]

Good idea to merge all the etrek capacity/monitoring posts into one topic.

 

[booster]

Thank you for this thorough research, stressing your batteries in the name of science.
I was about to pull the trigger on a victron because it seemed popular and well received, had bluetooth, didnt look like a cold war era artifact, and only requires an easier hole saw cut iinto a panel instead of a trickier rectangle incision (an actual concern of mine).

Dont know much about SOC.........

Eric

There is nothing wrong with Victron monitor as far as I know, the big issue is with the settings that they put in for the default and the fact that they say those settings are good for nearly all applications.


Monitors are great things, IMO, and in general a top need in an RV, but they have to be setup to match the system well or the information they give will likely be very far off of accurate. Secondarily, when set up to match the system properly, it will not improve the charging of the charging systems but it will tell you how good or bad a job they are doing (the exception would be if you have a charger you can manually override to improve the charging).


The big question is how to set up the monitor so it gives the best information based on the quality of charging the other systems are capable of. It is highly likely that in many systems, the "full battery" indication will not really be a full battery because the systems are capable of charging to true full. What this will do is show you that you have less capacity than the battery rating (but not change that as it has always been the way, but now you know it) and will likely shorten battery life. I will be putting up some more information to get let us try to find out what the best way to do that would be based on what kind of information the various users would like to see, and how they would use it.