JRP3
Hyperactive Member
You might try contacting some of the people on DIYelectriccar.com who might be able to point to to a higher voltage inverter, or build one for you.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Plan: Off grid solar with a Model S battery pack at the heart
- Thread starter wk057
- Start date
-
- Tags
- Energy Environment Policy
Excellent pictures... thanks for posting those. Too bad about the pack not wanting to accept a charge as-is... I suspect unless the BMS is alive to allow it the contractors might not close?
So that's an interesting cell layout in each module. There are four groups of cells in each module, with each group having a common battery contact plate visible on the top. The two larger quadrants each have 148 cells, and the two smaller 74 cells.
Can you tell if the batteries in the larger groups are oriented with half the cells anode-up, and half cathode-up?
That's the only way I can seemingly make the numbers work to come up with a nominal pack voltage of ~390V.
With (rouugh numbers) a cell voltage of ~4v, they cant be series groups of 74 or 148 cells per module, as the voltages would be either too high or too low (not to mention that the common battery plates dictate there is a parallel connection in each group).
Thus, it appears that the 16 modules in the entire pack must be series connected. That dictates each cell module must be ~390/16=24V.
The only way to get 24 volts out of each module with 4v cells would be to have 6 groups of parallel cell sets connected in series. But there are only 4 visible groups.
Given that you would not want unequal groups (so as to keep current capacity even), I'm guessing that each of the larger visible groups is actually two groups of 74 cells, using that common top plate to connect them in series. That means half the cells in those larger groups have to be physically inverted.
It's hard to see clearly through the plastic top, but it appears that some of the battery contact areas slightly triangular (cathode?), whereas others are circular (anode?). The fact that there are "+" signs stamped in to each plate made me wonder if that implied cathode connections only...
If so, then I'd guess that those two larger groups have two smaller contact plates on the bottom of each of those larger groups. That allows 6 cells sets in each module at ~24v@230A, which is the neighborhood of the pack rating.
Interesting...
JRP3
Hyperactive Member
That's what I'm describing (or at least attempted to): Each module has 6 groups of 74 cells in parallel. With 16 modules, that's 96 groups in series. (I had seen pack voltage around 390v, but 400v is just as likely).
What I was attempting to rationalize was how each module would have 6 groups of 74 cells when there were only 4 visible contact plates on the top. I'm surmising that half the cells in each module are "inverted" to make that happen...
fluxemag
Member
You might try ordering inverters from Alibaba suppliers. A lot of these outfits basically have an order form that is a text box. Maybe they will make you the inverter you need. For me, I was looking for a 5-10kW inverter 24V DC input with a 220V AC pure sine output. You might be able to guess what I was trying to do based on those parameters. I gave up on that phase of the project, but it might be worth exploring the Alibaba route.
Attachments
Last edited:
mitch672
Active Member
I wonder if you could get your hands on one of the Tesla/SolarCity backup inverters... They are supplied/built by Tesla, normally uses a 10KW pack, probably could use a larger one... Charges from solar as well. Wonder if they are just using a Xantrex system plus maybe some custom BMS.
mitch672
Active Member
Your probably right. If they used just 2 sets of 73 cell modules, they'd have 5.31KW * 2 (about the 10KW system mentioned), at about 48VDC, so could just be an Outback Radian type of system (those are 4KW and 8KW, maybe they made their own similar product). These can be ganged together (up to 10), is this what you mentioned above? 80KW at 48V max?
Outback Power Inc. - Radian Series GS8048A / GS4048A
mitch672
Active Member
FYI, I have found some "packaged systems" that use the Outback Radian inverters, and a 48V battery made by a company called "Corvus"' it's a lion pack with 7500 cycles to %80 DOD, with a 5 year warranty:
Packaged system: http://www.alpha.com/solar/downloads/specsheets/pvups/pvups_lion_datasheet.pdf
Corvus 6.5KW pack: Corvus Energy - Lithium Ion for Industry and http://www.corvus-energy.com/pdf/2014-06-25_AT6500_web.pdf
No idea on the cost of the above, I'm betting it's more than most residential customers would want to spend, since it's targeted at commercial applications.
Maybe your best bet is to reconfigure the pack as 48V modules, lots of commercial hardware works with 48V
Packaged system: http://www.alpha.com/solar/downloads/specsheets/pvups/pvups_lion_datasheet.pdf
Corvus 6.5KW pack: Corvus Energy - Lithium Ion for Industry and http://www.corvus-energy.com/pdf/2014-06-25_AT6500_web.pdf
No idea on the cost of the above, I'm betting it's more than most residential customers would want to spend, since it's targeted at commercial applications.
Maybe your best bet is to reconfigure the pack as 48V modules, lots of commercial hardware works with 48V
wk057,
I have considered ways to spoof a grid-tie inverter....i.e. get around the anti-islanding algorithms...and I think it's a tough problem. The grid has near-zero impedance so, when tied to the grid, the inverter's output voltage doesn't rise much at all when your local generation power exceeds your local power consumption. With the right kind of charger you might be able to duplicate that condition while charging your S battery but, of course, once the battery got full, there'd be no place for the excess power to go and the grid-tie inverter would shut down. Maybe that'd be OK.....
Thanks for the photos! I looked carefully at the module photos but could not discern the nature of the heat exchanger between the coolant and the individual cells. Is it a flat metal tube that runs down the center of a double row of cells with the cells pressed against it?
- Absolutely delighted P85 owner -
I have considered ways to spoof a grid-tie inverter....i.e. get around the anti-islanding algorithms...and I think it's a tough problem. The grid has near-zero impedance so, when tied to the grid, the inverter's output voltage doesn't rise much at all when your local generation power exceeds your local power consumption. With the right kind of charger you might be able to duplicate that condition while charging your S battery but, of course, once the battery got full, there'd be no place for the excess power to go and the grid-tie inverter would shut down. Maybe that'd be OK.....
Thanks for the photos! I looked carefully at the module photos but could not discern the nature of the heat exchanger between the coolant and the individual cells. Is it a flat metal tube that runs down the center of a double row of cells with the cells pressed against it?
- Absolutely delighted P85 owner -
I sat and looked at the module photo for a while and figured it out and I see 6 groups in there now. From the top each long plate (2/3rds wide) are the series links between two group hence half the cells are inverted. There are 3 plates below, two long plates but now on right side (looking from top) and now a different plate on the left (bottom side) side that links the top row to bottom (not two smaller plates as on the top). It would have a vertical orientation looking from the top. What you have is seven plates in total. The top right hand small plates are the outputs. The five others are the series linking plate between the six cells (five links are needed to link six cells). The 5 linking plates have 148 cells connected, yes, but half are orientated the opposite way i.e. 2 x 74.
Top right of pic is the output positive terminal as pictured with the multimeter and bottom right (of picture) is negative as shown by the OP. So yes six cell in series.
Sequence goes...(looking at the overhead pic) - it's counterclockwise starting from the top right of pic (top plate) ending at the bottom right (of pic) on the top plate. Count the cell groups (of 74 cells) from 1 to 6 counter clockwise starting from the top right.
First cell plate of 74 (top right plate) goes through to the bottom long plate heading left and through the long plate underneath, back up through cell/group 2 to the top side in middle group on the top row. Back along this top plate to left and back down below through cell 3. From cell 3 there is plate on the underneath that now links to cell/group 4 on the bottom row. (bottom left of pic). Back up through cell 4 to the top side again. Head right along the long plate to the cell/group 5 (middle of bottom row) - though cell 5 to the bottom again. Another long plate on the bottom heading to the right which joins to the last group (cell 6) - back through cell 6 to the top again on the final output plate at the bottom right of the pic. This is the negative side.
Too easy :tongue:
OK, so, I finally found a few companies that may be able to get an inverter that will work from this battery. I'm waiting for them to get back to me with exact specs, price, lead time, etc.
In other news, I did get this pack charged using an AC->DC charger. Cell voltage is sitting around 3.8V now, which I'm happy with.
More to come as I make progress.
I did post some more pics and info in my other thread about the pack itself: Pics/Info: Inside the battery pack
In other news, I did get this pack charged using an AC->DC charger. Cell voltage is sitting around 3.8V now, which I'm happy with.
More to come as I make progress.
I did post some more pics and info in my other thread about the pack itself: Pics/Info: Inside the battery pack
It seems it is easier to take apart a Leaf pack to rearrange for use at 48VDC nominal than the Model S pack. See this nicely done system at My Nissan Leaf forum.
Turns out I was able to pick up an Outback Radien GS8048A inverter for pretty cheap.
I also just struck a deal for the remaining solar panels for my project, 30kW worth.
So, it's looking like I'll be dismantling the Tesla battery after all. I have one final lead to follow up with regarding a high input voltage inverter, and assuming it strikes out like all of the others, I'll be reconfiguring the modules to work as a ~44V pack (2 modules in series with 8 sets of these in parallel) so that I can utilize the stackable Outback inverters... supposedly they will work at the 12-cells-in-series voltage (44.4V nominal instead of 48V like it expects).
Wonder how much coolant I need to drain to dismantle the pack
I'm considering custom building some additional bricks of 74 similar cells in parallel to add a 13th cell to each series, or another 592 cells (~7kWh). This would bump the nominal voltage up to 48.1V (3.7*13), and ~low SoC to 41.6V, charged to 53.95V... which would be closer to lead acid I suppose. People seem to use 14 cells in series, which could work also, but would mean me building 16 packs of 74 cells, 1184 additional cells. This would drop my amperage draw significantly, though.
I don't plan on dismantling the Tesla pack modules, which are 6 cells in series, so, stuck with those increments.
I also just struck a deal for the remaining solar panels for my project, 30kW worth.
So, it's looking like I'll be dismantling the Tesla battery after all. I have one final lead to follow up with regarding a high input voltage inverter, and assuming it strikes out like all of the others, I'll be reconfiguring the modules to work as a ~44V pack (2 modules in series with 8 sets of these in parallel) so that I can utilize the stackable Outback inverters... supposedly they will work at the 12-cells-in-series voltage (44.4V nominal instead of 48V like it expects).
Wonder how much coolant I need to drain to dismantle the pack
I'm considering custom building some additional bricks of 74 similar cells in parallel to add a 13th cell to each series, or another 592 cells (~7kWh). This would bump the nominal voltage up to 48.1V (3.7*13), and ~low SoC to 41.6V, charged to 53.95V... which would be closer to lead acid I suppose. People seem to use 14 cells in series, which could work also, but would mean me building 16 packs of 74 cells, 1184 additional cells. This would drop my amperage draw significantly, though.
I don't plan on dismantling the Tesla pack modules, which are 6 cells in series, so, stuck with those increments.
Last edited:
So, working on a test-bed setup for my project now that I have the pack disassembled (see pics in my other thread).
Trying out this reasonably sturdy plastic rack (rated for 1000 lbs, modules are < 100 lbs each) for now and have half of the modules sitting in it. I linked the pairs using the salvaged bus bar connectors since they can obviously handle the loads well enough. I'll pick up another one of these racks for the other eight modules tomorrow maybe.
Going to work on setting these up in parallel when my cable and connectors arrive. Will be using 4/0 cable (overkill I think) to lower overall resistance.
Also going to remove the Tesla BMS boards and start working on a custom replacement probably, eventually.
- - - Updated - - -
My first 48V inverter should arrive Monday and I should have everything I need by then to give it a test run.
Since my pack won't actually be 48V nominal and will actually be 44.4V I'm going to full charge the pack to just under 4.15V per cell (49.8V for my configuration) and then run the inverter with a reasonable load, say, 5kW (pack side after inverter losses, so, just over 112A draw)... possibly charging my Model S from the new pack for fun just to be able to say I did so, lol.
This should run for roughly 15-16 hours if I'm able to use the entire capacity of the pack before I hit the inverter's low cut off voltage. 3.2V per cell would be 38.4V, which is a hair below the inverter's low input rating. This is probably good, since it will be an added "bonus" pack discharge protection... but I won't know how much buffer that will equal until I do a monitored test.
I'll keep an eye on the pack voltage near the end as I actually don't want to drop below 3.2V per cell. I'll also monitor the pack temp with the FLIR cam since the active cooling components are gone for now.
Should be fun!
Trying out this reasonably sturdy plastic rack (rated for 1000 lbs, modules are < 100 lbs each) for now and have half of the modules sitting in it. I linked the pairs using the salvaged bus bar connectors since they can obviously handle the loads well enough. I'll pick up another one of these racks for the other eight modules tomorrow maybe.
Going to work on setting these up in parallel when my cable and connectors arrive. Will be using 4/0 cable (overkill I think) to lower overall resistance.
Also going to remove the Tesla BMS boards and start working on a custom replacement probably, eventually.
- - - Updated - - -
My first 48V inverter should arrive Monday and I should have everything I need by then to give it a test run.
Since my pack won't actually be 48V nominal and will actually be 44.4V I'm going to full charge the pack to just under 4.15V per cell (49.8V for my configuration) and then run the inverter with a reasonable load, say, 5kW (pack side after inverter losses, so, just over 112A draw)... possibly charging my Model S from the new pack for fun just to be able to say I did so, lol.
This should run for roughly 15-16 hours if I'm able to use the entire capacity of the pack before I hit the inverter's low cut off voltage. 3.2V per cell would be 38.4V, which is a hair below the inverter's low input rating. This is probably good, since it will be an added "bonus" pack discharge protection... but I won't know how much buffer that will equal until I do a monitored test.
I'll keep an eye on the pack voltage near the end as I actually don't want to drop below 3.2V per cell. I'll also monitor the pack temp with the FLIR cam since the active cooling components are gone for now.
Should be fun!
Concerning the functions of the BMB:
TI has patents and device datasheets for monitoring Li battery voltage which are freely available on the world wild websticle. The purpose of the 'bleed' resistor network switched by the transistor is not for bleeding or balancing the cells, but rather for reading the near-end-of-charge cell voltage to determine the state of charge using secret calculatus found in the bowels of their patents.
While it sounds like a novel approach and idea, there is no basis in fact for the charge shuttling or impedance balancing theories. The FETs are not used in the linear region for any sort of impedance regulation--there is no provision other than ON or OFF for those FETs as driven by the cell balance pins of the TI chip (please RTFM).
TI has patents and device datasheets for monitoring Li battery voltage which are freely available on the world wild websticle. The purpose of the 'bleed' resistor network switched by the transistor is not for bleeding or balancing the cells, but rather for reading the near-end-of-charge cell voltage to determine the state of charge using secret calculatus found in the bowels of their patents.
While it sounds like a novel approach and idea, there is no basis in fact for the charge shuttling or impedance balancing theories. The FETs are not used in the linear region for any sort of impedance regulation--there is no provision other than ON or OFF for those FETs as driven by the cell balance pins of the TI chip (please RTFM).
Last edited:
