|
|
Post by servdiv on Nov 4, 2020 18:17:20 GMT
2015 409 with 4 AGM 105AH house batteries, I was not feeling right about the batteries as I ran my frig only for a not plus the standard minot.7 amp draw from other minor devices but when I came back to the boat in the morning I had a low voltage on the display at 12.1v. I bought a battery tester with load testing and SOH features. I check the farthest battery from the negative/positive switches and it came in at 32% SOH and the voltage is right, the 2nd was 34% SOH and the 3rd was 41% SOH and the last one closest to the switches was at 100% SOH. I feel I have to get a new bank as these are originals. Got a price on AGM 105AH deep cycle from West marine for $250 each new form the same lot.
If I missed something please advise as I am only 1.5 years into this...thanks to all...Kevin |
|
|
|
Post by ForGrinsToo on Nov 4, 2020 21:04:46 GMT
Kevin,
If the batteries are the same age (preferably the same lot), that is quite a disparity in the SOH. Assuming they are all 5+ years old, the 100% SOH is pretty remarkable, as well. Bear in mind that the batteries should be fully charged, then fully disconnected from everything, then allowed to rest for as much as 24 hours before you try to test capacity or State of Charge. SOH relies on some estimates of cell impedance if you do not have data from brand new. I think it's really odd that a "set" of 4, presumably wired in parallel, the same age if not the same lot, would exhibit such a range. I think you should recharge, isolate, then re-test.
With 220 AH available on AGM, if the refrigerator runs frequently, I might see 12.2 on the Scheiber panel in the morning, but the battery monitor will show 12.3 or more, especially when the refrigerator stops running. You'd be well off to invest in a good battery monitor, especially if you get a new set of batteries.
You'll also find forum posts on running a dedicated line to the refrigerator directly from the battery bank, with a breaker and solenoid switch in line; the solenoid would be controlled by the Scheiber panel switch. The point being, the refrigerator pulls 7+ A and there is a lot of wire from the battery bank to the panel then to the refrigerator. Wire has resistance; power is lost as heat through all that wire.
Lead-acid battery life in particular is more a matter of cycles and depth of discharge than age, but 5 years is pretty typical.
Geoff
|
|
|
|
Post by vasko on Nov 6, 2020 3:19:55 GMT
|
|
|
|
Post by zaphod on Nov 6, 2020 5:01:53 GMT
Definitely invest in a battery monitor and install it. I was in the same situation as you this past spring. Before replacing my batteries I spent some time getting familiar with the charge and discharge characteristics of my existing bank. In doing so I noticed some strange behavior, which prompted me to take a closer look at the wiring. I figured out that the bank was wired incorrectly which was leading to unbalanced charging.
The monitor I bought was the Victron BMV712. It is a bit pricier than others, but allows me to connect to it with my phone via Bluetooth to configure, monitor amps in and out, SOC etc.
I have since installed 4 Firefly carbon foam AGMs that I am very happy with!
|
|
|
|
Post by vasko on Nov 6, 2020 11:20:42 GMT
I really don’t think a battery monitor can show any meaningful information regarding to the state of charge ...
e.g all battery standard monitors in the end are one voltmeter , one ammeter and very misleading number for state of charge...
I have experience with NASa, Victron and another two hi-end models with charts lcd etc. all fully useless...
|
|
|
|
Post by servdiv on Nov 6, 2020 13:15:58 GMT
Thanks to all, I found it interesting to let the batteries rest before truly seeing if the batteries SOC/SOH data is accurate, interesting. I have the current factory wring set up with a Balmar monitor with wifi which seems to work well. One also noted the panel and the battery monitor are a little different too which they are. I did do a draw test for all components by turning on and off devices, this also revealed that I draw .7 amps with everything off on the panel (still lit). My frig seems to draw about 4.2 amps too.
Again, I have the factory wiring except the battery monitor i.e. the house bank last negative line going through the monitor shunt. West Marine has 105ah group 31 Deca (spelling?) batteries for $250 each all built in the same month I will pick up for thanksgiving.
As a side note, I have a 200W solar blanket with a MPPT controller and use as needed to trickle charge the bank when I am going to be on the hard (Practicing next season). I was able to test it and got about 6 amps out it which works for what I am going to use if for.
Kevin |
|
|
|
Post by zaphod on Nov 6, 2020 16:35:01 GMT
I really don’t think a battery monitor can show any meaningful information regarding to the state of charge ... e.g all battery standard monitors in the end are one vomtmeter , ione ampmeter and very misleading number for state of charge I agree, so far I haven't found a lot of use for the SOC function, i find the the Ah counter more useful. |
|
|
|
Post by ForGrinsToo on Nov 8, 2020 17:37:19 GMT
Vasko, Zaphod:
Victron says the BM-712 calculates %SOC based on Ah used rather than voltage. That's why it has a shunt and requires a sync when first powered up, once the battery bank is fully charged, along with an input of the theoretical capacity (220 Ah for 2 6V AGM in series).
I have the BM-712 installed adjacent to the battery switches. Investing in the Bluetooth dongle means I can check from phone or iPad without disturbing the Admiral when she's in the galley. Or in the middle of the night. Or getting on my knees.
Thanks to both of you for making me read the manual carefully this time.
Geoff
|
|
|
|
Post by vasko on Nov 9, 2020 4:49:37 GMT
Vasko, Zaphod: Victron says the BM-712 calculates %SOC based on Ah used rather than voltage. That's why it has a shunt and requires a sync when first powered up, once the battery bank is fully charged, along with an input of the theoretical capacity (220 Ah for 2 6V AGM in series). I have the BM-712 installed adjacent to the battery switches. Investing in the Bluetooth dongle means I can check from phone or iPad without disturbing the Admiral when she's in the galley. Or in the middle of the night. Or getting on my knees. Thanks to both of you for making me read the manual carefully this time. Geoff All battery monitors are based manly on that idea...and all battery monitors have shunts ... BUT unfortunately that is not enough at all - there are huge amount of factors related to state of charge: temperature when charging and when discharging, cables, battery state, battery position , sampling frequency, shunt accuracy(strictly speaking the are arrays of hall effect sensors which are cheaper and million times more accurate then a shunt - shunt is very obsolete technology) etc. million of things to take in account - strictly speaking it is way easier to predict the lotto numbers then s SOC of a battery bank  in brief there are two ways to judge your battery SOC: 1. Simple one : monitor the amount of amps your battery bank can accept as a charge < 5amps almost fully charged , <10amps well charged > 10amps you no still continue charge it to prolong the battery life 2. Complicated one : you need a system that plots on a chart your battery bank behaviour over 24h period and have in mind what you have ON on the boat, type of chargers (solar wind, mppt, engine, Are they smart ? Etc.) and know the standard behaviour of your battery banks - and when it deviates from it you need to find the reason : you can see the chart if my boat battery banks here (my one are LiFePo4) : www.boat-vitals.com/display.html?link=11-101-ec2c284ef0-Celcius-dd_MMM_HH.mm-GPS5 Unfortunately this is |
|
|
|
Post by goody on Apr 11, 2021 4:44:43 GMT
Hi all,
I have enjoyed a new Jeanneau 51 for the last 3.5 year and I went for the additional betteries for the house bank giving 720amp hours.
i have the standard factory chargers plus solar and these worked really well for a couple of year until one of the batteires faulted and was giving off sulphar smell, battery swelled and was very hot. I isolated this one and all was good but we noticed the reduced value. Then last week it happened again and I dont like having risk so I am going for a full LI replacement. Eye watering in expense but I will have 600 amp hours available and more importantly will have 400+ useable at a full 13.2 volts and each one has its own BMS and safety gates. Even with my 720amps on the old system at best there was only 140 amps available when they were new.
These batteries are safer and have black ship protection and alarms at state of charge.
I wnat piece of mind when i am out and I dont believe i will need to connect to shore power any more. I have the Balmar unit which means no further expense just some setting changes and some protection for the alternator.
I am really looking forward to this as I am not a hands on mechanical or electrical person so being ahead of failure is critical to me.
|
|
|
|
Post by vasko on Apr 11, 2021 14:27:31 GMT
I’ with 260 LiFePo4 with 300amps BMS two full seasons and starting the third one - daily actual electric usage about 220ah - cost about 700EUR all works perfect...
|
|
|
|
Post by blade on Apr 11, 2021 15:02:10 GMT
Thanks to all, I found it interesting to let the batteries rest before truly seeing if the batteries SOC/SOH data is accurate, interesting. I have the current factory wring set up with a Balmar monitor with wifi which seems to work well. One also noted the panel and the battery monitor are a little different too which they are. I did do a draw test for all components by turning on and off devices, this also revealed that I draw .7 amps with everything off on the panel (still lit). My frig seems to draw about 4.2 amps too.
Again, I have the factory wiring except the battery monitor i.e. the house bank last negative line going through the monitor shunt. West Marine has 105ah group 31 Deca (spelling?) batteries for $250 each all built in the same month I will pick up for thanksgiving.
As a side note, I have a 200W solar blanket with a MPPT controller and use as needed to trickle charge the bank when I am going to be on the hard (Practicing next season). I was able to test it and got about 6 amps out it which works for what I am going to use if for.
Kevin Interesting that you should see 0.7 amps draw with everything turned off. On my 2012 36i I also have about 0.5 - 0.6 amp draw with everything turned off, do you have any ideas what is causing this draw, or anyone else have any ideas. Doesn't seem much of a draw but over 24hrs 7 days a week it can be as much 100 amp hrs, enough to eventually drain the batteries. |
|
|
|
Post by zaphod on Apr 11, 2021 17:00:09 GMT
Thanks to all, I found it interesting to let the batteries rest before truly seeing if the batteries SOC/SOH data is accurate, interesting. I have the current factory wring set up with a Balmar monitor with wifi which seems to work well. One also noted the panel and the battery monitor are a little different too which they are. I did do a draw test for all components by turning on and off devices, this also revealed that I draw .7 amps with everything off on the panel (still lit). My frig seems to draw about 4.2 amps too.
Again, I have the factory wiring except the battery monitor i.e. the house bank last negative line going through the monitor shunt. West Marine has 105ah group 31 Deca (spelling?) batteries for $250 each all built in the same month I will pick up for thanksgiving.
As a side note, I have a 200W solar blanket with a MPPT controller and use as needed to trickle charge the bank when I am going to be on the hard (Practicing next season). I was able to test it and got about 6 amps out it which works for what I am going to use if for.
Kevin Interesting that you should see 0.7 amps draw with everything turned off. On my 2012 36i I also have about 0.5 - 0.6 amp draw with everything turned off, do you have any ideas what is causing this draw, or anyone else have any ideas. Doesn't seem much of a draw but over 24hrs 7 days a week it can be as much 100 amp hrs, enough to eventually drain the batteries. Mine is the same. I assume it is the electronic breaker panel. .7a seems a bit high to power up an lcd display, but there is probably some other stuff going on like tank level monitor circuitry, and of course, my battery monitor. |
|
|
|
Post by andreshs1 on May 5, 2021 3:51:16 GMT
I’ with 260 LiFePo4 with 300amps BMS two full seasons and starting the third one - daily actual electric usage about 220ah - cost about 700EUR all works perfect... Hi Vasko I am really interested in your full setup, can you please share the full details? I am looking at the following cells, but stuck on the BMS to use, as I would like one that can cope with heavy loads, as I would like to avoid a bank just for engine and genset starter these are the cells I am considering, 560AH will work out around 900USD: EVE 3.2V 280AH Grade A Lifepo4 Battery Cells Fully Matched (https://www.alibaba.com/product-detail/Free-Shipping-4pcs-pack-Grade-A_1600071693145.html) Cheers |
|
|
|
Post by vasko on May 5, 2021 4:25:35 GMT
I’ with 260 LiFePo4 with 300amps BMS two full seasons and starting the third one - daily actual electric usage about 220ah - cost about 700EUR all works perfect... Hi Vasko I am really interested in your full setup, can you please share the full details? I am looking at the following cells, but stuck on the BMS to use, as I would like one that can cope with heavy loads, as I would like to avoid a bank just for engine and genset starter these are the cells I am considering, 560AH will work out around 900USD: EVE 3.2V 280AH Grade A Lifepo4 Battery Cells Fully Matched (https://www.alibaba.com/product-detail/Free-Shipping-4pcs-pack-Grade-A_1600071693145.html) Cheers I've ordered prebuild 2 x120 lifepo4 - each with 150 amps BMS in directly from aliexoress (asked the seller to upgrade the BMS from 120 to 150) and extreamly happy so far- my only regret is that I did not get a BMS with BLE - e.g can monitor it from my phone For starting you need batteries without BMS e.g shop.gwl.eu/LiFeYPO4-batteries-12V-1-1/Lithium-Battery-LiFePO4-12V-20Ah.htmlAlso important in case your alternator is NOT smart that supports li batteries profile you need to connect one wet/AGM/gel/efb battery in parallel to the li once to prevent damages to the ternator when the ki batteries are fully charged and the BMS cut them off |
|
|
|
Post by Trevor on May 6, 2021 3:02:50 GMT
Hello all, A lot has been said about the poor old battery monitor but let’s investigate how a modern battery monitor works. Victron has been mentioned so perhaps that is typical of what we can investigate. Others do work in much the same way but sometimes their description of the coefficients, which the user sets, can be a little different. Batteries do lose capacity with age but we need to provide some starting point and the published battery capacity is the best place to start. In some cases the battery published capacity is actually slightly lower than the actual battery capacity. That allows a little bit of leeway for age related capacity reduction. Ok, let’s dig into it. Batteries have some variables that make estimating SOC difficult but by setting the coefficients relating to those variables, we can have a reasonable chance of estimating our state of charge. Importantly, when the battery is 100% fully charged (or very close to that) if we synchronize the battery monitor to indicate “fully charged” we go back to a “known state” in which to estimate other parameters. A bit like Dead Reckoning, it’s not perfect but good enough. The automatic and periodic reset to "fully charged" give us a starting point so any inaccuracies in our parameters do not accumulate over many charge/discharge cycles giving bad results. Batteries hold energy, allow us to draw that energy and then later allow us to recharge that energy. The storage and retrieval is not 100% efficient. In other words electrons in does not exactly equal electrons out. We have to provide a little more energy with each charge than we can extract on the next discharge. In some cases it is close to 100% but it cannot ever actually be 100% In the case of Victron this is called “charge efficiency”. Depending upon the battery this can be set pretty high..... let’s say between 95 and 99%. The battery capacity does change depending upon just how much current is being drawn. In other words the capacity changes a bit depending upon if we are discharging the battery fast or slow. That is why batteries are sometimes rated for discharge over a period off 20hours. It sets a discharge rate for a certain time because capacity changes depending upon the discharge rate. If we increase the discharge current we slightly reduce the apparent capacity of the battery. Just to add insult to injury, the degrading of the capacity is not linear. It is exponential. Luckily with smart battery monitors we can use an exponent to set that de-rating factor. This is known as the Peukert factor. The Peukert factor is usually somewhere between 1.05 and 1.2. It is an approximation and not particularly accurate at higher currents but better than nothing. As previously mentioned in this thread, observing a low charge current may indicate a fully charged state. It could also indicate a low power charging source so in the case of Victron, three factors are observed before 100% charge is declared. The charge current is observed and is known as the “trailing current” set as a percentage of battery capacity. For instance a 100Ah battery with 4% trailing current has a trailing current of 4 amps. Current alone is not enough to satisfy the declaration the battery is fully Fully Charged. Imagine a battery with a charging current of 4 amps but a terminal voltage of 10 volts. Yes the charge current is low but the power supply feeding it is not powerful enough to satisfy its charging needs. The Victron unit also measures battery voltage to ensure we are not seeing low charge current because of a low power charging source. That is set as “charged voltage”. They also have a setting for how long those parameters must be observed before the battery is declared as 100% full. It is called the “Charged Detection Time” and the user sets it as long as you wish but simply defaults to 3 minutes. Once these three parameters are satisfied the battery monitor synchronize the system as 100% full, ready for the next discharge cycle. Once the battery monitor declares a fully charged state it waits for the current leave the battery, monitoring and reporting the amount of current that has been consumed. One Amp for one hour means 1 Ah has been consumed. If the battery is a 100Ah battery we have 99Ah to go. We wont talk about minimum recommended discharge level here. Another factor that can be set is the temperature coefficient for temperatures under 20 degrees C. This allows for de-rating the battery capacity for cold weather as a certain percentage per degree below 20 degrees. That can be set by the user but certainly in my region of the world we don’t bother. Battery monitors can also give very many alarms depending upon high as low battery temperatures, high and low battery voltages, and low SOC. They can be used to start generators at certain SOC to ensure the battery bank does not discharge to the point of damage. Mastervolt and Victron have several preset points for SOC to enable relays to alarm, load shed or activate certain events based on SOC including generator start. The beauty of the modern battery monitor is the ability to check parameters and declare the battery full based on those parameters. In that way the accumulation of errors does not occur beyond a couple of discharge cycles. The system on my boat reaches float and fully charged status nearly every day because of the solar panels. It doesn’t have a chance to accumulate errors. The difference between shunts and hall devices was also mentioned. My experiments with Hall devices have shown numerous shortcomings that are not evident with shunts. Shunts are linear devices (complying with simple ohms law) and are extremely precise and hence quite accurate. They may suffer in very extreme cases from very slight temperature drift but it is minimal. Unfortunately they do have to be cut into the circuit to be measured. That is a bit inconvenient. Hall effect devices on the other hand are many times less precise and hence less accurate than shunts. They are brilliant for simple switching or tach pulse generation but the ratiometric ones used for current measurement have some inherent shortcomings. They suffer from temperature drift at the crossover point but the most annoying characteristic is the remanence of the core causing the crossover point to change if large currents are passed through the device. The answer is to remove or reduce the magnetic core but that reduces the sensitivity to an unusable degree. Providing a close loop system can reduce it but we introduce greater complexity. The device can be reset periodically to zero the crossover point but at this stage they are not used for critical applications because they don’t have the precision of a shunt. They are very convenient if they have a split core that can be simply closed around the wire carrying the current to be measured. That is absolutely brilliant but they are not in the same league as a shunt for precision and hence accuracy. Just more grist for the mill..... Trevor |
|
|
|
Post by vasko on May 6, 2021 3:43:19 GMT
Tested multiple battery monitors , NASA, Victron, two models very cheap Chinese no name from AliExpress - all of them have pretty much random readings for % power available (actually the Chinese ones was way better then the two brand names) - as result and frustration I ended up building my one based on arduino with very fine tuning and multiple measuring spots - had fun for sometime - next the project evolved to www.boat-vitals.com (www.boat-pulse.con) for north America In brief battery monitor are completely useless - they can be used as a good voltmeter and ampermeter but thats all in the end.... SoC In the end is kind of a number which is like the electron - as more accurate you think it is a more far from the true it is... |
|
|
|
Post by zaphod on May 6, 2021 18:48:52 GMT
I'm not sure I would agree that battery monitors are "useless". I would say that the SOC function is not terribly useful, but everything else is still good information. It is pretty valuable to be able to see cumulative amp consumption when it comes to monitoring power consumption. Once you get to know your daily average ah consumption you are able to manage your power consumption better. If you see a sudden increase in power consumption it may give you advanced warning of a problem with one of your systems, be it a pump that is running too long, a refrigeration system that is not cycling normally etc.
|
|
|
|
Post by goody on May 23, 2021 2:15:55 GMT
Yes I agree state of charge through a shunt dosent give a good view over time. The LI batteries I have just installed measure direct from the battery to the monitor from each battery BMS so is very accurate. These batteries stay at 13.2 volts or higher all the way through the state of charge and draw so great clean output. I also have the .7 draw at all times, that is the monitor, the bilge pump monitor and various other small systems that keep checking on your boat safety.
My LI have lived up to expectations so far but it will be time and extended use that makes the difference. Note that they also charge myuch faster for me, under motor I get a full 90 amps back in an hour as they run generally coooler and check health at individual batteries real time. They dont go to bulk until 95% is reached under alternator and solar seems to have a better input for them.
|
|
|
|
Post by andreshs1 on May 23, 2021 7:19:03 GMT
Hi goody
can you please share full details of your setup?
cheers
Andres
|
|
|
|
Post by goody on Jun 14, 2021 8:55:24 GMT
Hi Andres,
I have put in 3x200 amp juice brand batteries each with their own BMS and a monitor whch is also juice brand. The alternator has a tempreture setting to stop it overheating and regulate the charge as this can be an issue for a straight installation, my current regulator was good enough to do it. I have 4 x 200w solar panels to support this but with 600amps (480 available) i dont have to move for 5 days or more in terms of motoring. I run a standard gel for the start battery but dont need any system alteration to get this working.
It was done in a day and I have been out for about 9 days since install and no issues to date, the biggest problem is letting the capcity to go below 80% it takes a bit of getting used to but I run two fridges full time so I use quite a bit of power.
|
|
