mazlak
New Member
Wanted to shared the work I did on my LV system for the sprinter van I am converting into a camper. The objective of this project was to have a stupid amount of power available with little loading effect, a high Energy to size ratio, field serviceable, and low cost. I have never done a system like this before but my background is an electrical engineer who does advising type work in commercial device manufacturing so I like to think I know my way around not blowing myself up and not setting myself on fire.
(Main) Parts Used:
3 - Tesla Model S Battery Module 5.2KwH
Victron Energy MultiPlus-II 24v
Overkill solar Tesla BMS kit – Plug-and-Play Retrofit Kit
Victron Energy Orion 24/12-70A DC-DC Converter
4AWG stranded pure copper wire (stripped out of a retired tesla supercharger station)
Metrics:
Battery box weight: 200lb
Total yield: 13-14 kwH (from initial test depleting 80-20% still benchmarking)
Batt box cost: $2000
System cost (including excluding solar parts): $3500
Labour time: (excluding solar work) 64.5 Hours
Explanation:
I Plan on doing a circuit diagram and full part list in a later post. But wanted to show the finished pictures and some of the construction shots I had to get feedback and comments on the forms.
The innovative thing I did here was really a racking system for the tesla battery modules. The intention was originally have support for a hot spare. but after realizing the 660W of solar panels I have mounted on the roof (plan to show that project in a different thread) has a less than expected average yield all 3 modules are relatively permanently installed. Batteries are mounted on 2ft long 1" 8020 extruded aluminum. I found that the extruded aluminum formed the best ridge for sliding the battery in and out of, as tesla's modules, as you can see by other builds do not have a good place to rack and mount really.
LiPo type batteries are typically discouraged in use with DIY vans so I had to factor in multiple failsafes.
Each module has a 150A block fuse, and the Overkill solar BMS that has 100A Efuse.
Each battery is then wired into a shutoff switch. This was an expensive edition with the partial operation under a single modules failure in mind, and removing batteries to charge them not with the van. Now that the sub-project is done, not sure if it was worth the hassle and wire.
Positive and negative of the battery box go into a box-side bus bar with 5 foot of extension into where the I/O is to be.
On the Inverter side, I have a MultiPlus-II 24v which should give me all the 120V power needed for electronics in the system and can do the 70 Amp shore connection charging. batt Box is connected to a 3 Pol switch so that the LV system and inverter can be switched on or off separately of each other.
There is then a bus bar for servicing the 24v LV appliances that go into the van and a Victron Orion 24/12-70A DC-DC Converter to step down the native 24v so that 12v appliances will work as well.
Things I am glad I did:
Use at least 24v, 12v systems are terrible because so much power is lost in transmission due to the twice as high current for the same power output. The $100 step down was an easy fix to not have to do a 12v system. I maybe would have done even a 48v if there was no solar involved, but that would need an even number of tesla modules.
Using the tesla modules, the yield on these is pretty friggin good considering the cells themselves cost me $300 a peice. They are IMO far safer than doing a DIY LiPo cell pack like I originally planned.
The 150A block fuses were cheap to add on, and gives me piece of mind that if a tool or the module itself fell in a way to make a full short outside the BMS's Efuse or if the Efuse failed, I would not incinerate my van.
Things I learned:
I got the overkill solar BMS for peace of mind, The Efuse gives me the assurance that from the lowest level the battery will shut off under a fault and I can better monitor the performance of each module. I originally did not design for it and could have done without that particular part it to save $600 on the batt box. I am not the most thrilled with the COTS-chinese main BMS with soldiered on wire but I will admit the new Pathfinder that is replacing this model looks like it is worth it and overkill benchmarks each of the COTS part of the BMS and provided a test report in the packaging. It really is just an optimization if I could have also done this part myself as it pained me to use a $200 BMS with each $300 battery pack. But I figure I could have at most saved $300 here after the cost to do something else not even to mention the time.
The per module disconnects I had big intentions for. The only time it came in handy was bringing up each module to synchronize voltages after they had drifted from keeping the batteries disconnected durning assembly. It was a pain to cap more wire, using the wire, and installing them. A nice to have feature, did not need to be done in phase one here.
I did not have access to a laser engraver to cut the acrylic so I did it with drills, circle cuts and a jig saws. I am glad this part is acrylic, but I will probably completely redo those with a laser engraver. It was a big waste of time and took away from the polished look of the batt box.
The extruded aluminum cost like $400, because I was impertinent and used mcmaster, I really could have used a cheaper vender with a longer lead time.
Things TODO:
(Main) Parts Used:
3 - Tesla Model S Battery Module 5.2KwH
Victron Energy MultiPlus-II 24v
Overkill solar Tesla BMS kit – Plug-and-Play Retrofit Kit
Victron Energy Orion 24/12-70A DC-DC Converter
4AWG stranded pure copper wire (stripped out of a retired tesla supercharger station)
Metrics:
Battery box weight: 200lb
Total yield: 13-14 kwH (from initial test depleting 80-20% still benchmarking)
Batt box cost: $2000
System cost (including excluding solar parts): $3500
Labour time: (excluding solar work) 64.5 Hours
Explanation:
I Plan on doing a circuit diagram and full part list in a later post. But wanted to show the finished pictures and some of the construction shots I had to get feedback and comments on the forms.
The innovative thing I did here was really a racking system for the tesla battery modules. The intention was originally have support for a hot spare. but after realizing the 660W of solar panels I have mounted on the roof (plan to show that project in a different thread) has a less than expected average yield all 3 modules are relatively permanently installed. Batteries are mounted on 2ft long 1" 8020 extruded aluminum. I found that the extruded aluminum formed the best ridge for sliding the battery in and out of, as tesla's modules, as you can see by other builds do not have a good place to rack and mount really.
LiPo type batteries are typically discouraged in use with DIY vans so I had to factor in multiple failsafes.
Each module has a 150A block fuse, and the Overkill solar BMS that has 100A Efuse.
Each battery is then wired into a shutoff switch. This was an expensive edition with the partial operation under a single modules failure in mind, and removing batteries to charge them not with the van. Now that the sub-project is done, not sure if it was worth the hassle and wire.
Positive and negative of the battery box go into a box-side bus bar with 5 foot of extension into where the I/O is to be.
On the Inverter side, I have a MultiPlus-II 24v which should give me all the 120V power needed for electronics in the system and can do the 70 Amp shore connection charging. batt Box is connected to a 3 Pol switch so that the LV system and inverter can be switched on or off separately of each other.
There is then a bus bar for servicing the 24v LV appliances that go into the van and a Victron Orion 24/12-70A DC-DC Converter to step down the native 24v so that 12v appliances will work as well.
Things I am glad I did:
Use at least 24v, 12v systems are terrible because so much power is lost in transmission due to the twice as high current for the same power output. The $100 step down was an easy fix to not have to do a 12v system. I maybe would have done even a 48v if there was no solar involved, but that would need an even number of tesla modules.
Using the tesla modules, the yield on these is pretty friggin good considering the cells themselves cost me $300 a peice. They are IMO far safer than doing a DIY LiPo cell pack like I originally planned.
The 150A block fuses were cheap to add on, and gives me piece of mind that if a tool or the module itself fell in a way to make a full short outside the BMS's Efuse or if the Efuse failed, I would not incinerate my van.
Things I learned:
I got the overkill solar BMS for peace of mind, The Efuse gives me the assurance that from the lowest level the battery will shut off under a fault and I can better monitor the performance of each module. I originally did not design for it and could have done without that particular part it to save $600 on the batt box. I am not the most thrilled with the COTS-chinese main BMS with soldiered on wire but I will admit the new Pathfinder that is replacing this model looks like it is worth it and overkill benchmarks each of the COTS part of the BMS and provided a test report in the packaging. It really is just an optimization if I could have also done this part myself as it pained me to use a $200 BMS with each $300 battery pack. But I figure I could have at most saved $300 here after the cost to do something else not even to mention the time.
The per module disconnects I had big intentions for. The only time it came in handy was bringing up each module to synchronize voltages after they had drifted from keeping the batteries disconnected durning assembly. It was a pain to cap more wire, using the wire, and installing them. A nice to have feature, did not need to be done in phase one here.
I did not have access to a laser engraver to cut the acrylic so I did it with drills, circle cuts and a jig saws. I am glad this part is acrylic, but I will probably completely redo those with a laser engraver. It was a big waste of time and took away from the polished look of the batt box.
The extruded aluminum cost like $400, because I was impertinent and used mcmaster, I really could have used a cheaper vender with a longer lead time.
Things TODO:
- The Fuse blades are just running the lights. These will be moved and run off a more proper fuse box.
- The inverter board will be propped up with more finished dow rods. I just have the 2x4s there now.
- Need to do a full benchmark on power, from a full 100% charge to 0% discharge.
- Need to add cross bar to top of batt box so movement will not slide the very thin rails off. Though it is very sturdy as it is.
- Need to do a thermal analysis at high current to see if there are any issues with parts.
- replace acrylic with laser cut for clean finish, maybe add some RBG leds for the meme.
, and set the upper bound at 50% of that. So that gave me a budget about 2.5k for the batt box. I ended up going with the Tesla NCA because density per foot was very important to me with it in a van and I thought only 2 batteries would already be plenty enough, or for $300 a piece I could swap the failing component (hence the original hot spare plan). but yes with the effective yield loss LFP would have actually had the same density per foot.