I know this is a post that a few people have been waiting for. We actually received our batteries quite some time ago, but I held off writing this as I wanted to work out the bugs and gain experience.
Now, before you read the rest of the article, I’m going to assign some homework. Months ago I wrote 2 posts when I was contemplating these batteries. Re-reading those posts will definitely help at making more sense of this post. So here are the 2 links:
Ok finally, here is a schematic of the final design that I actually installed.
Don’t worry, I will wait for you homework haters to open the second link I posted above so we can compare the original design to what I actually installed…………. Done yet? Nope, ok………….. How about now? Ok.
First lets discuss why I changed the Victron BP device to a BlueSea ML-RBS 7700. Upon further testing with the Victron device, I discovered that if there was current flowing backwards through the device it would never trip off. Of course this was very problematic as this device was supposed to COMPLETELY isolate the battery bank (removing all loads and charge sources). However if a charge source was actively putting power in, then that would never happen. The second problem with the Victron BP device is how it acts in the event of a failure. There have been multiple reports online of the device failing dead short circuit (i.e. just a big wire). In the event that did happen, then the BMS would not be able to disconnect the battery bank, EVER. You also would not know about that failure until it was too late.
Enter the BlueSea RBS devices. These devices are as close as you could get to a true “mechanical latching relays”. They draw almost no current in a resting state (like 0-13ma) and only take a momentary pulse of current/voltage to change states. Unfortunately, they are over double the cost of the Victron devices, but I think they are worth it. However, I made a mistake with the model I picked. Unfortunately my power frugalness got the best of me. I choose the 7700 as it does not draw any current when its in a resting state. Unfortunately it has a fairly big flaw as it will not fail safe in the event of a BMS failure. If the BMS were to fail, it could not send a trigger pulse to the 7700 and the batteries would remain connected. A better choice would have been the ML-RBS 7713. The 7713 requires a tiny current (13ma) to hold it in the closed state. However, in the event of a BMS failure that current would most likely be cut causing a state change and isolating the batteries. Sigh, stupid me gave up a valuable failsafe for 13 measly milliamperes (for those wondering, that is only 3.12ah per day).
The second big change is a control unit between the cell boards and the 7700. Unfortunately my initial design would not have worked as I thought it would. The wire loop between the cell boards didn’t operate as I originally believed. The control board was required to monitor those boards and then send the signal to the 7700 (or any other disconnect device). Unfortunately the control board cost $55AUS and drove the cost of the BMS up significantly. However, the device is open source and came with the source code (written in PICAXE). This allows me to easily add features (like a pack voltage cut) and personally vet the code.
So here is the complete bill of materials for both the batteries and the BMS. I didn’t post prices for the components as it can range wildly depending on where you buy them from (especially the cells). The total cost for our cells and the BMS (including all the shipping, brokerage and taxes) was about $3050USD delivered here in Grenada.
|4||Thundersky Winston WB-LYP400AHA||http://en.winston-battery.com/index.php/products/power-battery/item/wb-lyp400aha?category_id=176|
|1||BlueSea ML-RBS 7713||https://www.bluesea.com/products/7713/ML-RBS_Remote_Battery_Switch_with_Manual_Control_Auto-Release_-_12V|
|4||EV Power CBM400 Cell Module||https://evparts.com.au/ev-power-bms/bms-cell-modules/cm400-cell-module.html|
|1||EV Power BMS Headboard||https://evparts.com.au/ev-power-bms/bcu-micro-08b.html|
Stay tuned for Part 2, where I discuss more of the technical details and specific settings on our devices. I will also detail some of the issues we encountered. For now, enjoy some pictures of the initial installation (it has been changed at least twice since these photos).
I had A LOT TIME while I waited for our batteries to arrive. In the meantime, I cobbled together a test bed so I could ensure everything would work as I designed. It didn’t look pretty, but got the job done.
Woot Woot, they arrived.
Although the company I bought from was horrible, they did pack them well.
Initial commissioning and balancing. Time for more cobbling. (Don’t worry I filled out a JSA before starting the job). Notice the black tape that I used to tape up all my tools. The bottom of the charger was all stainless steel, so we taped cardboard to it (dammit mom, don’t comment that its a beer case). All risks mitigated.
I used a 100W resistor during the initial top balance to bleed energy from the high cells. Bleed, then check with the multimeter (I actually used 2 meters, but one is out of shot). Repeat, over and over until each cell reads 3.650V. Then I reconfigured them in parallel and let them sit for 2 days. After sitting, I repeated the whole process. After I was happy with the top balance, I immediately hooked up an inverter and pulled the pack down to about %60. I now monitor the balance about every 2 weeks. So far they have stayed completely balanced to a second decimal point.
100W is a lot of power. That resistor got really hot and required water cooling. We had to switch out the water twice as it had begun to steam.
Test fit next to the old batteries.
The control board installed in a moisture proof case. It was a tight fit with the switch.
Crimping up the new cables (thanks Dave on Tigres for the loan of the crimper). Everything is double runs of 2/0 gauge. This is how the boat left the factory so I made up like for like.
(If the Fille De Joie crew is reading, notice the lights in the background. One of Rhonda’s favorite items on our boat).
Installed and connected for a test run.
Turned out our boat needed some speed holes.. LOL, JK these are for cooling. The holes eventually got some nice wooden grates. I even installed a fan in the closest vent to exhaust hot air. I don’t think it was required but I went for the belt and braces approach.
Installed and admiring the blinkin lights. As may not be evident in this picture but the batteries are actually physically “floating”. They are sitting on a bed of foam and secured on the sides with foam as well. A nylon strap holds them on the top and the back. This is an attempt to isolate the batteries from vibration as that can cause premature failure.
The results of removing a metric shitton of weight. This is also where my laziness is paying off. Here you can see our old waterline (the high black mark) and the new waterline. The boat rose 3 inches.