Now I had the mounting system, it was time to hook the lithium system up, including the LiFePO4 BMS.
WARNING: I know I do this a lot, but this entry is heavy on the technical details and light on everything else, including excitement, humour and ~fun sailing stuff~. Here we go!
The cells were organised in a way called 2P4S – as in, pairs of batteries were connected in parallel and then these 4 groups were connected via series. Basically like this. I put them in the new space, secured them and then attached the copper busbars being VERY careful not to short them out. I wrapped a spanner in electricians tape to insulate it.
Connecting the Orion JR BMS
Now that the batteries were arranged in their final configuration (and mounted), it was time to connect the LiFePO4 BMS I was going to use to monitor the pack. A BMS basically keeps an eye on the cell voltages – if any start to go too high or low (indicating under/over charge) then the BMS drops a connection, breaking a relay and the charge/discharge stops. More info is in the first post of the series.
Connecting the cell taps
The cell tap connector looks like this
Yeah, that’s a lot of wires.
It’s actually two cell groups, 1-12 and 13-16. Since I just had 4 cells in series I could ignore the 13 to 16 leads and the 13- one. The current shunt leads attached to my current shunt, the cell 1- tape connected to the negative side of the battery, the cell 1 tap connected to the + side of the first (parallel) connected cell pair, cell tap 2 connected to the + side of the second cell parallel pair and etc for cell taps 3 and 4. Cell taps 5 – 12 all had to be connected to the positive of cell 4, as did the total pack+ tap. THAT’S A LOT OF WIRES.
Here is where all the taps went (the tiny numbers written next to the cells) (not shown – the current shunt ones as they went directly to shunt)
Connecting the other connector
Oh you thought that was it? Sadly not. Here is the OTHER connector, that controls all the relays etc
So it’s probably easiest if I go down through each pin in order. HOLD ONTO UR BUTTS
1 – When the BMS decides that the voltage etc of the battery is ok to allow it to charge, this pulls to ground. If there is an over voltage condition, it shuts off. Connected to a RA-700112-DN (70 amp relay) on pin 85.
2 – This is connected to +12Volts and turns on the BMS when connected. I connected it to a switch, and from there to the battery. This is the only thing directly connected to the battery (aside from voltage taps). Fused.
3 – This is the opposite of pin 1, in that it controls the discharge (i.e – stuff running off the battery). It actively pulls to ground, and if there is an error condition or an under-voltage condition (i.e I’ve used too much juice) then it switches off. Connected to a Blue Seas 7713 remote battery switch. Note: this switch requires a 12+ input so I had to do some extra fancy wiring (see below)
4 – Not used. More for electric vehicles to recognise when they are plugged in.
5 – Not used. Same as 4.
6 – Ground wire. Connected to ground. Pretty self explanatory.
7 – Now, this one actually ended up driving the main relay which was a Blue Seas 9012 (also known as the OH SHIT relay as it’s the last line of defense if the charge/discharge enable fails). This wire draws to ground like the others, and when it detects and under/over voltage condition it shuts off, shutting off the relay. This relay is IMMEDIATELY after the battery, and everything that draws power (aside from BMS) connects on the other side of it, isolating any charge/discharge source from the battery. It also is ‘watchdog backed’ meaning it turns off automatically if a BMS fault is discovered.
8 – Ground for the thermistors. Self explanatory. See also 10/12
9 – This basically has the same functions as 7 – it is an output pulling to ground that is configurable. However it is not watchdog backed, meaning I didn’t want to use it for any of the ‘main’ relays. This one I used as part of the state of charge display, it activates when an error is discovered and turns on an error LED.
10 – positive lead for thermistor 1. This measures the temperature around the batteries, as they shouldn’t be charged under zero degrees celsius.
11 – This wire uses a voltage between 0 and 5v to show state of charge of the pack. I ran it to a little SOC display so I didn’t have to plug in the laptop every time to see the SOC.
12 – Thermistor 2. See 10. Basically another one same as 10 so I have some redundancy with the temperature measuring.
13 – This uses a voltage between 0 and 5v to show the current limit. I run it to the SOC display to show if charging has been shut off.
14 – Same as 13 but for discharge. Not used.
15 – This can be used to do a couple of things. Not used.
16 – Same as 10 and 12. Not used.
17 – Same deal as 9, as in same as 7 not watchdog backed. I used this one to drive another RA-700112-DN relay that controlled the power to the alternator regulator. This was set to shut off slightly before the over-voltage protection on pin 1, meaning the alternator would shut off before the charge circuit is disconnected, which would otherwise damage the diodes.
18 – Blank
19/20 – leads for the canbus, which is a protocol that the BMS can use to talk to similarly equipped chargers etc. I don’t have any of these so it wasn’t used.
This is a Blue Sea Systems 9012 relay.
It’s sat between the battery and the main charge bus, meaning that EVERYTHING is on the other side of it. The only things that are powered not through the relay are the BMS (which will which to a low power mode drawing 0.1 watt if it detects a low voltage condition) and the positive of the 9012 relay coil. The negative of the relay coil is connected to the BMS, see above. Positive coil was fused and connected directly to battery.
This was a Blue Sea 7713 Remote Battery Switch. I used this instead of another 9012 as it used less power to keep engaged and also had a manual override, for when I REALLY needed that last 20% of battery. It sat between the main bus and the discharge bus (that had almost all the loads on it). In an under-voltage condition, the BMS stops pulling to ground on the control wire and the discharge bus shuts off. HOWEVER
As mentioned above, it needed a +12v source, so I couldn’t just hook the BMS up to the negative coil side like the other relays. I used a P‐Channel MOSFET (STP10P6F6) and a 100k resistor to ‘invert’ the output from pulling down to ground to a +12v output. This was then mounted in a case. Wiring was as follows
Gate – attached to the discharge control wire from the BMS
Source – connected to the main charge bus (+12v)
Drain – connected to the +v side of the 7713 coil control wire.
100k resistor connected between source and gate.
Note shitty soldering job – after this photo was taken I got a little solder stand that made it way easier and I redid it a lot better – but forgot to take a photo. So here we are.
This is me using the stand to solder the thermistors. This was also annoying.
Once this was done, one main lug was connected to the main bus, the other was connected to the discharge bus, the +12v coil wire was connected to the drain of the MOSFET (and fused) and the ground coil wire was connected to ground.
This was a RA-700112-DN 70A Relay that was connected before the charge bus that all my charging sources lead to. In an over-voltage condition, the BMS stops pulling to ground on the control wire and the bus switches off. Simple.
Except the connectors on the main contacts were this weird large tab that I couldn’t find any female connectors for so I ended up having to drill a hole and bolt on ring connectors to. Like below
The other thing to note is that this relay has a diagram on the front which DOESN’T correspond to where the pins are. Pin 85 and 86 are on opposite sides compared to where they are on the diagram, so I ended up plugging them in reversed and blew a fuse. Oops. CORRECT wiring is as follows
Pin 30 – Connection to main bus
Pin 87 – Connection to charge bus
Pin 85 – Connection to charge enable wire on BMS
Pin 86 – Connection to main bus (+v). Fused.
I kinda wish I had used the SPDT version (this one is a SPST) as I could have used the other throw to activate a buzzer that would sound an alarm if the charge level got too high. No matter!
Alternator Regulator Power Relay
Snappy title hey? This is the same RA-700112-DN 70A Relay that I used on the charge bus, complete with the same dumb tabs on the main contacts.
Pin 30 – alternator regulator power wire
Pin 87 – 12v+ coming from ignition power circuit (so it turns on when the engine key is turned)
Pin 85 – Connection to multipurpose output 2 wire on BMS
Pin 86 – Connection to charge enable bus (+v). Fused.
The idea here is that the voltage settings for the Multipurpose Output 2 wire from the BMS are set slightly lower than the charge cutoff, so this relay will trip before the charge one does, killing power to the alternator regulator and hence stopping the alternator charging the batteries safely. If we just cut the charge bus without doing this first, you could damage alternator diodes.
The various voltage levels at which stuff activates are above. Some of the stuff relies on cell voltages, other on pack voltages – I put both in to make for easy comparison, but the ones in brackets aren’t actually used, just included for reference. Aside from the stuff in green, all the red/yellow stuff is controlled by the LiFePO4 BMS and should (theoretically) never happen, as the other charging sources should always keep the voltages in the green.
You may notice the charger settings are SUPER low. This is so that the battery bank is kept between 30% – 40% charge (roughly), meaning I can leave it a week or two without worrying that I am destroying the batteries. The night or morning before I head out, I’ll bump the settings up to 13.8 so I’ll leave with a full charge.
To provide an example of what this all looks like, here is around 70% of the system wired up. Note that none of the BMS wiring is in place yet, but you can see where it would attach on the relays and terminal blocks. You can just see the edge of the main ground bus on the right of the picture.
I mentioned the alternator regulator a few times but haven’t provided much details. That will be in a later post.
The LifePO4 BMS is programmed through software on a laptop that you connect into with a serial cable. The Orion JR BMS is pretty easy to configure, the only thing that confused me was the way you set the charge and discharge limits – its on the ‘cell settings’ tab. Everything else is done on the page for the actual output.
If you made it this far – congrats! And good luck with your own LifePO4 system!