I have a setup on my boat with Venus OS running on a Pi, connected with USB VE.direct cables to a BMV-712, and two Smartsolars. Currently the smartshunt is using the Aux input for monitoring a 2nd battery bank V, so there is no temp sensor input, which is not ideal, since the banks are lead acid, and the boat is in the tropics. I am monitoring the boat remotely until I can return in a few months, and am using the MPPT internal temperature sensors to allow them to reduce the voltage based on temperature compensation. The devices are configured to share data using VE.Smart over BT, and DVCC is disabled. The Venus OS/VRM is just allowing me to monitor things and adjust settings. Although I can’t find a way to view the measured temperature at the MPPT, I can see that it varies during the day, and I can calculate it based on the float or absorption voltages I am seeing - the temperature is usually 40-48C, with ambient temp of 30-35C.
My question is what happens once I add a wired temp sensor to the shunt, and then enable DVCC? I think it will override VE.Smart networking on the Smart Solars (and the Orion XS which I plan to install as well). But what happens if the Venus OS Pi is turned off? Will the devices revert to using the VE.Smart network to share temp, V and I from the shunt automatically? This is what I am hoping, as I would like to have the system function independent of the Venus OS when I am sailing, as I don’t usually have a need to keep the Venus running while on board; the BMV display and Victron Connect on BT is enough to monitor things.
I will test this of course once I return, but I am hoping to have all the necessary components with me - at this point I think a BMV temperature sense cable is all I need.
I have read the manuals for DVCC and VE.Smart networking, and I think this excerpt explains that it should work as I hope, but I am wondering if anyone has actually tried anything similar?
“In systems in which the GX device is used for logging purposes only, VE.Smart Networking can be used to allow chargers to synchronise, or even receive information from sensors. Keep in mind that if, for some reason, the same information (i.e. voltage sense) is being received by the charger over BLE and VE.Can/VE.Direct, the information coming over BLE (through VE.Smart Networking) will be ignored.”
I have not tried it as you have described. What I do know from spending many years in the community is that having both a VE Smart Network and a wired network can at times result in erratic or unexpected behaviour, some people have no issues, others do. The usual advise is to choose one or the other. If you have wired connections and a GX or similar device then use this as it should be more robust.
it is not nessessary to enable DVCC. The Sensor adjusts the voltage of all components via VE.Networking. Independent of Venus. Without a temperature sensor the MPPT uses the temperature in the mornig at sunrise. But it is not the temperature of the whole day. If the battery temperatre rises, the mppt uses the wrong temperature! It is better to use the Aux Input for a Temperature Sensor. In a well wired system is no critical standby load at the starter battery. In a System with lead batteries DVCC is only usefull if the battery ist small and PV ist strong. So you can limit the Battery current without limiting the possible load current. But i think in your boat the battery ist big and th PV is small.
Gel batteries are used. look at the absorbtion voltage peeks of the blue line in teh left diagram. They are highter if the temperature in the right diagram is low. In the first day i marked the start of float phase with the blue poiter in both diagrams. In the right diagram you see that the temperature rises about 3 degrees during float. And the float voltage decreases by 10mV. Without temperature sensor the MPPT would use 29°C for the whole day.
Thank you. I have not enabled DVCC, and it sounds like I may not need to, even after I have added a temp sensor to the BMV-712, as all the devices including the Orion XS will communicated using VE.Smart over BT. I will try to replicate all the scenarios once I return to the boat.
Thank, that is very helpful. Yes, I am definitely not giving my batteries the best treatment, which is why I am trying to monitor them closely until I can install a proper sensor. Without a battery temperature sensor, I am forced to reply on the MPPT for now, and am just trying to keep the Abs. and Float voltages as low as possible while still allowing full charging. The batteries are below the waterline so they will remain cooler than the MPPT providing they are not being charged too agressively. I have tried turning off charging mid day when adjusting voltages, and can observe that the MPPT will start the bulk-abs-float cycle with a new temperature reading and compensated voltages. One problem that I have noticed is that the BMV-712 does not seem to use compensated voltages, so if the Absorption voltage is reduced, it will not trigger the SOC to 100% based on the value at the BMV. Because my solar panel output can be quite low on cloudy days, I have set the tail current quite low to ensure the absorption cycle completes, as it will otherwise drop to float too early, and the BMV indicates it is less than 100% based on the Ah counter (although only about 2-3Ah short usually). My daily overnight consumption is only 12Ah at present.
It sounds like adding a temperature sensor to the BMV-712 and mounted at the batteries will be sufficient so I don’t need to bother with DVCC as you have suggested. I have 660W of solar supplying 210Ah of flooded lead acid batteries, but the panels are heavily shaded during storage, so I usually only see 75-100W of charging input. Once out of storage, the 660W solar will feed a 650Ah LiFePO4 house bank instead, with the lead acid acting only as a serial backup charged with the Orion XS from the lithium bank.
You can always reduce the charged voltage in the BMV-712 to force a sync at higher temperatures and when you are cooler and absorption is st a higher voltage, the tail current will still limit, but there is increased risk of spurious early synchronisation if you go too low.
Priciple a battery is full charged if the mpp is in float and the current has avalue og about C/50 to C/100. In your case about between 2 A and 4.5A. An example:
In the left diagram the jump to 100%SOC is marked. In the middle diagram this event is at 5,3A. The voltage is reduced to float. The “Charged Voltage” of the shunt has to be a little bit lower than the lowest float voltage. A Gel Battery 460AH is used. The absorbtion time ist is set to “adaptive”. So the MPP reads at sunrise the resting voltage to calculate the absorbtion time. This is usefull if the SOC ist different from day to day. Problem ist, when sunset slows down the charge current: There is a space of time that matches “Full”. To prevent synchronisation the “Charged detection time” should be longer.
I am sorry but I am going to disagree here, with solar charging the charged voltage on the shunt has to be slightly below the absorption voltage not the float voltage. If you use the float voltage then you can get spurious synchronisation when you have low sunlight because the shunt can not tell between full battery causing low tail current and low current due to low sun which is enough to get to the float voltage.
I am also going to disagree with your statement that a full battery is float and 1 to 2% tail current, a battery is charged when it is at absorption voltage with a tail current of 1 to 2%. The MPPT uses absorption voltage and tail current to go from absorption to float, hence if the tail current is 1 to 2% at end of absorption it will be much much lower when the voltage drops to float.
Perhaps it is simply thst you wrote float when you meant absorption.
At float voltage minus 0,3V the charge current of a not fully charged battery can not be lower than a accurate value of the tailcurrent. Look to the Manual:
Background is the generation of sulfuric acid by charging a lead battery. The highter acid concentration near the electrodes rises the voltage of the battery. So it is important for aceptable current to rise the charge voltage too. A fully charged battery generates no sulfuric acid. If you try to charge it with absorbtion Voltage it will generate hydrogen and oxigen. Then the battery looses water. And if it is a VRL battery ist will go dry. It is naot a good idea to charge with absorbtion voltage to 100%SOC, especially if the battery is only partially discharged.
Close to sunset or if a cloud appears the current can immediately fall to the tailcurrent while the voltage drops slowly. So the Smartshunt will indicate"full" too early. But this situation last only a few minutes. By setting the “Charges detection time” to a minute more, it will not synchronisize.
And the manual also has the following caveat for solar systems because the section you quote is really for chargers able to always deliver a fixed high current and Vicyron recognise that for solar systems, something different is required.
It is also possible that the battery monitor synchronises too early. This can happen in solar systems or in systems that have fluctuating charge currents. If this is the case change the following settings:
Increase the Charged voltage to slightly below the absorption charge voltage. For example: 14.2V in case of 14.4V absorption voltage (for a 12V battery).
Increase the Charged detection time and/or decrease the Tail current to prevent an early reset due to passing clouds.
Whilst float charging has its benefits, many installs, especially mobile (and this was a bist) will not be able to daily recharge at float voltage and need the higher recharge rates from absorption. If not, the batteries never reach 100% and operate in a partial state of charge which leads to sulphation.
You will find this use of absorption minus a small margin is recommended by most of the experienced users here.
I do wish the manual would be updated to reflect that, as -literally- 99%+ of the problems with SOC readings I see from customers are caused by setting their charged voltage to just below float rather than just below absorption.
This is useful information, as one reason for monitoring closely is to avoid losing too much water. Since I don’t have a temp sensor at the battery yet, I am trying to avoid too much off-gassing, which was occurring before I enabled temp compensation based on the MPPT temperature. The default -16.2mV/C (not sure where they get this from) is much less than the -30mV/C recommended by Trojan and Interstate for their flooded lead acid batteries, and mine were off-gassing too much and triggering my CO alarm. Once I reduced abs and float to 14.2/13.6 and -30.0mV/C for compensation, the off-gassing stopped. I have had to fiddle with the float voltage, gradually increasing it until the solar carries the load with a few extra watts going to the batteries. At less than 13.6V (about 13.1V with compensation), the batteries were slowly losing charge in the afternoons.
As pwfarnell has noted, I will use a voltage between float and abs for my BMV charged setting, since my panels are often shaded throughout the day due to obstructions in the marina or cloud cover, so it may trigger false SOC syncs.
Thanks for pointing this out, as it definitely applies to my situation, with cloud cover and obstructions affecting my panels often during the day. While in storage, my SOC only goes to about 97% each night, so having precise SOC isn’t too important, but it is good to understand how to get this working better once we’re back on board and drawing loads. Then we will have a different challenge - the main house bank will become the lithium, with the current lead acid bank kept topped up using an Orion XS from the lithium. I have a Smartshunt to be used for monitoring the lithium SOC separately, and I will need to find the right value for charged voltage.
This has been an informative thread. One more mystery is why my DC system load during the day is about 2x the overnight load. I suspect it is something to do with the Venus OS and when I connect remotely, it triggers realtime updates (and sometimes I will connect to MPPTs remotely), but the load increases during daylight hours each day. The only devices active: Pi 3B+ with 7" display running Venus OS, a MOFI ethernet router, MicroTik WiFi access point, and the Victron BMV-712 and two SmartSolar chargers.
Do you have a shunt measuring the DC loads or just rely on Venus OS to estimate it. If the latter then see the linked FAQ. It is probably down to small errors / voltage drop / inefficiencies whilst the solar is active.
There are other useful FAQs, one covering why to set the shunt voltage based on absorption.
I have a BMV-712 configured as Battery Monitor measuring all loads, and nothing bypasses that common ground bus. It is wired to the Venus OS Pi with a USB VE.direct cable, along with the two SmartSolars. The System Battery power value shown in VRM and on the Venus display match the BMV current reading. The DC System value in W does appear to be correctly calculating the difference between Solar yield and System Battery power. Just as I took this screen shot, the chargers dropped into float - they were previously showing Solar 47W, DC system 22W, and Battery 25W. I notice that the DC system dropped to 17W when solar went to 0, so your suggestion about inefficiency may explain this.
