DIY Electric Car Forums banner
1 - 19 of 19 Posts

·
Registered
Joined
·
16 Posts
Discussion Starter · #1 ·
Any recommondation? I have this choice for my project based on a Netgain Hyper9 motor. Both sets of moduls are from two 2018 Model S. Mileage differ a bit, but both are less than 100k. Price is nearly the same for each set.
Four moduls would reduce my concerns in regards of less space, but I'm afraid that the 6.x kWh moduls will be very exotic in the future. What should I do if one fails?
The new 2021 Model S uses a completly different form factor that I could never use in my car.
 

·
Registered
Joined
·
7,971 Posts
The 5.x kWh modules would be from the 85 kWh battery, while the 6.x kWh modules would be from the 85 kWh battery (there are 16 modules in both cases). The 100 kWh Model S is not obscure, so future availability doesn't seem like a big problem to me.

Even four classic Model S modules seem like a huge battery for an X1/9; avoiding the need to shoehorn five of them in seems like a good idea.

Of course they're the same voltage (both 6S configurations, 22 V nominal) so the number of modules changes the battery voltage (88 V vs 110 V), and you need at least suitable voltage for your motor... or a suitable motor for your voltage. If 88 V is enough but the controller can handle the fully-charged voltage of the nominally 110 V pack, you can change the number of modules later if desired.
 

·
Registered
Joined
·
16 Posts
Discussion Starter · #4 · (Edited)
Thanks a lot for your input!
Yes, it is a small car, but not as light as a Lotus Elan (Elon). With an ICE it has an 'empty weight' of 970kg (corrected Mar 08, 2021) and the 'total permissible weight' is 1200kg (corrected Mar 08, 2021). The latter must not be exceeded to follow the german TUEV regulations. The TUEV list several requirements for the battery pack enclosure and the mechanical installation in the car. For example they compare the required mounting points for the battery box to mounting points of seatbelts.
I do like Aluminium Construction Profiles. You do not have to weld and the slots give a lot of flexibility. For my 'TwoTeslaModuleBox' I've choosen 240x28mm profiles as sidepanels. Front and back panels have to be mounted with countersunk screws, because the maximum length for the box to fit in my frunk is 791mm. The cover is 'removed' in the picture below and you can see the two Tesla moduls. The orange part is the 'Manual Service Disconnect' on top of the box. The plugs or cable glands are still missing, same for the water connections.
Rectangle Font Cylinder Metal Roof


The gearbox is another part that gives me a headache, but to reduce the complexity of 'step 1' I start with the original one. It will be just a matter of time how long it resists the double amount of torque from the electric motor, but I'm already looking for alternatives.
 

·
Registered
Joined
·
7,971 Posts
... because the maximum length for the box to fit in my frunk is 791mm.
You're putting battery modules in the front of an X1/9? My thinking is that if you're going to do that, you have given up on handling and might as well just convert a pickup truck - there's lots of space in those for battery. ;) But seriously... I suggest working out the weight distribution of this vehicle, before and after conversion.

For those not familiar with the X1/9... it is a very small mid-engine car which doesn't even have the fuel tank or spare tire up front - they are behind the driver and passenger. It's smaller than the weight would suggest - the wheelbase is only 2,202 mm (86.7 in).
 

·
Registered
Joined
·
16 Posts
Discussion Starter · #6 ·
My crazy idea is: First pack with two Tesla moduls in the front trunk (frunk, in Tesla slang) and the second pack on top of the motor in the engine bay. Each battery pack is 75 kg. The rear with 75 kg battery + 50 kg motor should be equal to the current situation with ICE, maybe even lighter without fuel & tank.
You are right, the extra weight of 75 kg is up front and between the front wheels. Only radiator and fan can be removed there. Maybe a smaller 12V battery or an exchange to a LiFePo4 will save a few kg. That's the layout.
Automotive design Automotive lighting Motor vehicle Automotive exterior Vehicle
 

·
Registered
Joined
·
16 Posts
Discussion Starter · #8 ·
I've seen that the Tau software got some really great options to adapt it to the car and I can't wait to play around with them ;-) All the VW Beetle, Renault 4, Citroen 2CV conversions I've already heard of must have similar challenges.
I am not a race driver. I own this X1/9 since 1993. It has been a loyal buddy since and I simply love driving it.
brian_ knows the X1/9 quite well and maybe the sketches promis more than the reality. The gas tank is just a square right behind the driver. The spare tire is right behind the passenger. Both areas are separated. I want to use the space of the spare tire for HV-relays and distribution. I don't find any use for the space of the gas tank at the moment. Maybe in future if I want add more chargers.
Tbh I'm not afraid about the weight distribution with this constellation. It will be more 50:50 than ever and maybe I won't be overtaken by the back on wet / slippery roads any more ;-) Here is the look of the arrangement from above.
Line Engineering Auto part Machine Font

These Tesla moduls have the highest energy denisity on the market. As a Tesla Model 3 owner I trust the company and do not want to use anything else. LiFePo4 cells would be the only alternative, but they are much heavier and bigger.
20 kWh is my goal for this car. This should be enough for 150 km fun in the summer on country roads.
 

·
Registered
Joined
·
3,100 Posts
Tesla modules don't have the highest energy density on the planet...packing circles into squares is far from dense.

In this modern day and age, nobody carries a spare tire. Ditch it and the gas tank, put the heavy "dense" module in that space and put the distribution and relays on top of it in that same space.
 

·
Registered
Joined
·
7,971 Posts
My crazy idea is: First pack with two Tesla moduls in the front trunk (frunk, in Tesla slang) and the second pack on top of the motor in the engine bay. Each battery pack is 75 kg. The rear with 75 kg battery + 50 kg motor should be equal to the current situation with ICE, maybe even lighter without fuel & tank.
You are right, the extra weight of 75 kg is up front and between the front wheels. Only radiator and fan can be removed there. Maybe a smaller 12V battery or an exchange to a LiFePo4 will save a few kg. That's the layout.
View attachment 127899
That's a lot better than putting it all in the front, although the rear pack ends up very high, which is a common conversion problem.

It would be nice to keep the motor compact, and fit at least some of the battery at floor level, between the drive unit and the firewall. That might require a coaxial (with the axle) or rearward motor position. The spare wheel & tire (on the passenger side of this area) can go, with the cavity for it replaced by a flat panel; of course the gas tank (on the driver's side of that space, and visible in the side view drawing) will be gone.

The gas tank is just a square right behind the driver. The spare tire is right behind the passenger. Both areas are separated. I want to use the space of the spare tire for HV-relays and distribution. I don't find any use for the space of the gas tank at the moment. Maybe in future if I want add more chargers.
...
View attachment 127916
The fuel tank has a working volume of almost 48 litres; it occupies a space a bit larger than that. The corresponding space on the other side - occupied mostly by the spare tire well - is similar in volume. Between the two, that's almost enough volume for a 20 kWh pack by themselves. I assume a structural rib separates them, which is unfortunate when using long modules, but with other formats two packs or a pack which accommodates the separation would work.

From a mass distribution standpoint, it would be better to place battery modules ahead of the drive unit and low with HV distribution components over the drive unit, rather than the other way around.
 

·
Registered
Joined
·
287 Posts
Any recommondation? I have this choice for my project based on a Netgain Hyper9 motor. Both sets of moduls are from two 2018 Model S. Mileage differ a bit, but both are less than 100k. Price is nearly the same for each set.
Four moduls would reduce my concerns in regards of less space, but I'm afraid that the 6.x kWh moduls will be very exotic in the future. What should I do if one fails?
The new 2021 Model S uses a completly different form factor that I could never use in my car.
5 modules is ideal because you'll be closer to the maximum voltage accepted by the hyper 9 motor and will get better efficiency and performance than with 4. I'm sure you can find space for a 5th module somewhere. I've managed to fit 10 modules in my 2010 ford escape (which is admittedly a larger car) without compromising any interior space. 3 modules in the gas tank area, 5 above the motor, and 2 where the radiator was behind the bumper.

5 modules will give you more capacity than 4 of the 6.x modules too.
 

·
Registered
Joined
·
61 Posts
I run a Hyper9 with 5 Tesla modules and would not recommend running lower voltage on that motor.

I'd also caution you to consider voltage sag when anticipating your range for this system. The internal resistance of my pack is about 45mOhm, which means at wide open throttle (650A) I'm getting 32.5V of sag! My pack charge cutoff is 125V and my discharge limit is 100V. This means even on a full charge I'm forcing the battery below safe discharge voltage at full power.

What about at lower power? Well I use about 25kW to maintain freeway speeds. So at say, 20% SoC (105V) that'd be about 240A, which gives 12V of sag. So I can't maintain freeway speeds at 20% SoC. In fact, I can't sustain much better than 30mph below about 30% SoC, due to the voltage sag.

This is a limitation of lower voltage systems which I don't see discussed much and did not realize myself before I built my conversion. With a higher system voltage, the required current for a given power is lower, which means less voltage sag. Plus as system voltage goes up so does the spread between charge and discharge voltage, which means the (already smaller) voltage sag is an even smaller proportion of the usable voltage range.
 

·
Registered
Joined
·
16 Posts
Discussion Starter · #13 ·
Thank's a lot for this information @bawfuls! That makes it a lot clearer and it changes everything for me.
I'm amazed about the high internal resistance of your pack. 50 mOhm is the guess I had for a single cell. Calculated for a TESLA 6.x module with 6S86P this should result in 3.5 mOhm, correct? Ignoring the contacts and wires inbetween your pack it should have less than 18 mOhm. Which would be around 12V of sag. But that's just the theory, you provided the facts.
I'm wondering if this can be improved or if the main cause for this is the TESLA modul itself? Do more real world data about internal resistance of a single TESLA module exists? I'm looking for a criterium to select the moduls I'm going to buy.
 

·
Registered
Joined
·
61 Posts
They are in series so 45 mOhm is the full pack (measured/computed value based on observed performance in my car), meaning each module is about 9 mOhm. This is in line with other lithium batteries where I've been able to find internal resistance documented. For example, the OX Power 2.66kWh modules from ElectricGT are reported to have an internal resistance of 2.25mOhms. 16 of those in series for a 355V pack gives 36 mOhm. Sure that's 20% lower than my pack, but the reduction in sag from higher voltage is 300%.

My pack resistance may be a little on the high side (though I have checked the connections with a thermal camera after hard driving and found no hot spots), but even your lower estimates would still be an issue for the reasons I described. Going to a higher voltage pack improves the situation dramatically in two different ways. It is one of several reasons why I'm planning to eventually rebuild my car with a high voltage driveline.

A cursory google finds this link which says the tesla cells are 30 mOhm each which would imply 10.5mOhm for my whole pack, and this link where someone measured 46mOhm for a full 100D pack. Both would suggest mine should be lower than 45mOhm, but in reality I'm getting voltage sag in line with what I posted above. If my pack were only 10.5mOhm then I'd get no power-limiting sag until about 5% SoC. In the real world, if I floor it at full SoC I can pull the pack all the way down to 100V, and the lowest I ever discharged was about 25% SoC which required very careful driving below 25mph to limp home.

These voltages are measured in two different ways on my car. I have a general battery monitor in the dash which displays pack voltage measured at the contactor box, and a Dilithium BMS which monitors all the cell voltages and triggers a dash warning light upon low voltage or other error states. The warning light goes on at the same time the gauge displays <100V so I am confidant at least in the measurement. And since the BMS measurement is happening at the cell level, it should exclude any poor wiring on my part outside the pack.
 

·
Registered
Joined
·
7,971 Posts
I run a Hyper9 with 5 Tesla modules and would not recommend running lower voltage on that motor.

I'd also caution you to consider voltage sag when anticipating your range for this system. The internal resistance of my pack is about 45mOhm, which means at wide open throttle (650A) I'm getting 32.5V of sag! My pack charge cutoff is 125V and my discharge limit is 100V. This means even on a full charge I'm forcing the battery below safe discharge voltage at full power.

What about at lower power? Well I use about 25kW to maintain freeway speeds. So at say, 20% SoC (105V) that'd be about 240A, which gives 12V of sag. So I can't maintain freeway speeds at 20% SoC. In fact, I can't sustain much better than 30mph below about 30% SoC, due to the voltage sag.

This is a limitation of lower voltage systems which I don't see discussed much and did not realize myself before I built my conversion. With a higher system voltage, the required current for a given power is lower, which means less voltage sag. Plus as system voltage goes up so does the spread between charge and discharge voltage, which means the (already smaller) voltage sag is an even smaller proportion of the usable voltage range.
But that's only true if the resistance is due to wire and wiring devices and connections, not internal resistance of the cells.

If the same cells were arranged with half as many cells in each parallel group, and so twice as many of those groups in series, they would have the same energy capacity and same power capability, but at twice the overall battery voltage. The internal resistance of each of the groups of parallel cells would be twice as high, and with twice as many groups in series the total internal resistance of the battery would be four times as high.
  • modules as-is:
    • total internal resistance of five modules: 45 mΩ
    • internal resistance of one modules: 45/5 = 9 mΩ
    • internal resistance per cell group: 9/6 = 1.5 mΩ
  • doubled voltage version:
    • internal resistance per cell group: 1.5 * 2 = 3 mΩ
    • total internal resistance of battery (60 groups): 180 mΩ
Fortunately, for the same power requirement the current would be half (simply power = voltage * current) - that's 325 A at wide open, and 120 A at 25 kW and 210 V. As a result, the voltage drop for the whole battery would be only twice the present situation - that's about 60 V at wide open and 24 V at 25 kW. Twice as much voltage drop sounds bad, but it's a drop from a starting voltage twice as high, so the sag as a fraction of battery voltage is the same (12/105 = 11.4% and 24/210 = 11.4%). The sag is the same proportion of battery voltage and the same proportion of the usable voltage range, for the same power requirement.

Higher battery voltage is good to reduce the size of conductors and wiring devices required, but doesn't matter to voltage sag due to internal resistance. If those conductors and wiring devices are sized to suit the current there is no benefit in voltage sag of higher system voltage; if conductor size is limited then there is an advantage to reducing the amount of current squeezed through it.

Voltage sag is reduced by using a larger battery, relative to the power demand, regardless of series and parallel configuration. As they say "there ain't no free lunch" - you can't fix battery sag by shuffling connections within the battery.

The benefits of higher operating voltage are in the motor and inverter, not the battery.
 

·
Registered
Joined
·
61 Posts
Voltage sag is reduced by using a larger battery, relative to the power demand, regardless of series and parallel configuration.
This is a fair distinction, though in reality with what's available to the DIY crowd, when you go to a higher voltage battery you are likely forced to go to a larger battery as well.
 

·
Registered
Joined
·
7,971 Posts
Voltage sag is reduced by using a larger battery, relative to the power demand, regardless of series and parallel configuration.
This is a fair distinction, though in reality with what's available to the DIY crowd, when you go to a higher voltage battery you are likely forced to go to a larger battery as well.
Yes, finding the module which has a configuration which works to reach the targets for total battery energy and voltage while being packaged in a workable module size and shape can be a huge challenge.
 

·
Registered
Joined
·
16 Posts
Discussion Starter · #18 ·
Next days I will visit a used part dealer who specialized himself on higher price cars, including TESLA. He has several modules from different Model S on stock. It doesn't look like he is performing any tests with them to find out the capacity or SOH of the module. He just provides mileage and year of construction of the donor vehicle.
For my own confirmation I'm going to use a Battery-Tester that determines Voltage and internal Resistance in one step. The DIYSolar guys are using them to select their batteries. There is even a database for cells available that gaves a prediction of SOH of a single cell based on an AI system Vorhersage. They plan to expand the database with values for complete EBike or Laptop-Packs. Maybe they are interested to add Tesla modules... I would expect for the 6S74P 5.3 kWh module:
best = 2.43 mOhm
used = 3.24 mOhm
old = 4.46 mOhm
 

·
Registered
Joined
·
16 Posts
Discussion Starter · #19 ·
Here are the results for the five modules I finally bought: 3.41, 3.36, 3.36, 3.33 and 3.36 mOhm. Let's say a 'used' with a minus in regard of my valuation in the above post. Or maybe the proof that calculating just the cells and ignoring all the other internals of a module can't be the whole truth. Voltage was 23.6 V on each. The donor was a 2017 Model S with just 60.000 km.
After now, the starting point for total internal resistance of the five modules forming my battery is 16.82 mOhm. Let's see how much it will change after adding conductors, cables, connectors, etc.. I will watch it closely.
This is the offical start of my project. I will report more in the all-ev-conversions-and-builds section soon.
 
1 - 19 of 19 Posts
Top