Subaru e-mpreza (Impreza) 96 - AWD with 2 Outlander rear motors

[Zapatero]

The cooling plates are welded now.

PXL_20221006_104317592.jpg

First i made end pieces out of a standard u shaped aluminum profile:

PXL_20220903_221203911.jpg

Then they have been welded on the end of the cut cooling plate, so the water can circulate again:

PXL_20221006_104337975.jpg

Fits nicely in the Box!

PXL_20221006_104423414 (1).jpg

[Pete9008]

Very nice

[EV_Builder]

Nice work! But are you sure this works? Looks like the flow is quite limited by the deformation in the middle? Very cool how you shaped the U but I would have expected that in the middle you close that dent in the front. Maybe it's the picture fooling me, can't see the dept.

Just to help you ok?

[Zapatero]

Nice work! But are you sure this works? Looks like the flow is quite limited by the deformation in the middle? Very cool how you shaped the U but I would have expected that in the middle you close that dent in the front. Maybe it's the picture fooling me, can't see the dept.

Just to help you ok?

No i'm not sure this works , but the picture is irritating Its just the front that's bend down, in the rear it's still all the way open. They have been pressure tested at 2 bar, but no flow test have been made. I checked resistance by blowing air through them, which worked very good. I'll try with water next.

[EV_Builder]

Ah ok...you could improve flow by welding U shaped round tube on the top. Over the middle part. Another option could be to add 2 90degrees elbows with some hose if you have wall thickness for thread.

[Zapatero]

Question:
If I have two inverters like in this case, does it matter if the cable length for the first one is significantly shorter than for the second inverter? Assuming i want 50/50 performance.

I'm thinking of connecting the DC cables in a row, meaning: battery - inverter 1 - inverter 2.

What effects could there be?

I use 30mm² wires as original in the outlander. Could also use 50mm² cable ...

[marcexec]

I'd wire it
battery -50mm2- inverter 1 -30mm2- inverter 2
in that case, as the first leg has to carry current for both inverters.

[Zapatero]

I should also mention cable lengths:
Battery to inverter 1: 50cm
Inverter 1 to inverter 2: 400 cm

[marcexec]

I should also mention cable lengths:
Battery to inverter 1: 50cm
Inverter 1 to inverter 2: 400 cm

Yep, makes no sense to carry that extra copper on 8m of cable, even though I'm a friend of overkill wiring. I have 35mm2 on the bike and the controller peaks at 150A...

[Zapatero]

I'm using 50mm² cable in the Toyota GT86 conversion at 130kw / up to 400A. The cables never got really hot.

In the Subaru I'm pushing 80kw for each motor, means the 4m of cable have to safely transport 300A.

Isn't 30mm² too small? Although I'd like to be wrong. Obviously in the outlander they only use 30mm² cable which is also almost 2m long!

[arber333]

I'm using 50mm² cable in the Toyota GT86 conversion at 130kw / up to 400A. The cables never got really hot.

Ampera main fuse is 350A, battery cables are 35mm2 and each motor phase cables is about 16mm2 with phi8mm circular point of contact. And inverter has 600A IGBT for each motor. Its not for the want of amps... i think 35mm2 is enough for HV at 400A. You rarely use its max value.

Gen1 Leaf motor otoh has 60mm2 phase cables with 800A IGBT brick. Correct me if i am wrong. I dont think dc link has more than 35mm2...

[EV_Builder]

Crossection capacity depends heavyly on distance and temperature...

[Zapatero]

I think the main issue here is cable length: 400cm one way
So the main question is: can a 35mm² cable transport 90kw at nominal 310v, 268 min voltage, 350 max voltage which results in max 335A?

[Zapatero]

Is that calculator correct or which one do you guys recommend?

https://www.js-technik.de/en/calculate- ... -dc-motors

[arber333]

Is that calculator correct or which one do you guys recommend?

https://www.js-technik.de/en/calculate- ... -dc-motors

Well probably, but then at V=400V dV of 1% voltage drop at 400A should mean 1600W of heat generated within that string!

[EV_Builder]

335A on 35mm2 defenitly not nominal on that distance.
So the question then is how often...

1600Watts on 4meter of 35mm2 during the period of pulling.
Couple of times will be ok but i bet that the cable will heat up after repetitions have been done.

If you know the weight of the cable it might give you a feeling for time. If it's 10kgs it will take a good while to heat up etc.

[Pete9008]

This talk about cable ratings got me curious so decided to do a few sums to see what the rate of temperature rise might be. Came up with this spreadsheet which allows the voltage drop, power dissipated and rate of temperature rise to be calculated. Thought it worth posting as the results surprised me a little (so let me know if you find any bugs in it) - the rate of temperature rise being lower than expected.

Here is the result for 4m of 35mm2 at 500A:

TempRise.png

So a voltage drop of 0.96V (0.24V/m) and a power loss of 480W (120W/m).

The bit that surprises me is the rate of temperature rise, only 1C/sec. It's worth noting that the spreadsheet doesn't allow for the effect cooling at all (from cable or to end fittings), so it only relevant for short time periods and longish lengths, but at this level it looks like bursts of 180kW would present no problem at all.

Just as a sanity check I compared it to this data sheet https://hilltop-products.co.uk/amfile/f ... uct/30146/ which indicates that 35mm cable can handle 2000A for 3sec (shortest duration should give the best agreement as spreadsheet doesn't include effect of cooling) at 140C. The spreadsheet gave a temp rise of 16C/sec so 4x16=48+140C = 188C which is pretty close to the 180C max operating temperature for that cable so it seems to add up

[mjc506]

Is the cost saving of the smaller cable worth the risk of having to redo it later? Or worrying about "how hot are my cables getting while I'm flooring it up this hill"

[Pete9008]

Despite what the sums came up with I'd probably still use a bigger cable - but then I tend to over engineer stuff

[Zapatero]

As I'm in Vietnam at the moment and only have my phone, could you be so kind and compare the upper calculations with a 50mm² cable? How big are the differences? Thanks so much

[Zapatero]

Despite what the sums came up with I'd probably still use a bigger cable - but then I tend to over engineer stuff

The 35mm cables are just in there because the Mitsubishi outlander inverter has them from factory. I'm wondering how they limit Power over time?

[bexander]

Is the cost saving of the smaller cable worth the risk of having to redo it later? Or worrying about "how hot are my cables getting while I'm flooring it up this hill"

Well, flooring it with 200kW, up hill or flat road will accelerate a 1500kg car rather quickly to way beyond road speeds.
If adding a 1500kg trailer for a total fo 3000kg you would still only need 140kW to maintain 100km/h on a 12% incline.

However if you are to race the car on track, then it is another matter. Constant acceleration and deaccelerations would leed to an almost constant 200kW of powerflow in one or the other direction. Would require a quite hefty batterypack to sustain this load constantly.

[Pete9008]

As I'm in Vietnam at the moment and only have my phone, could you be so kind and compare the upper calculations with a 50mm² cable? How big are the differences? Thanks so much

Here you go:

35mm2:

TempRise.png

50mm2:

TempRise_50mm2.png

[johu]

On the lower end it's me with 16mm² for sustained 10 minutes of HPC CHAdeMO (200A) and then still above 100A for another 15 minutes.

Charging loads down the cable way more than (road!) driving because it's "permanent". As bexander says there may be short power bursts but it is impossible to sustain high power for a long time because you run out of road and last but not least because the battery runs dry. 180 kW, 60 kWh -> 20 minutes
That's why OEMs never lay out cable for sustained peak power.
If I remember correctly the WLTP driving cycle has an average power draw of 7 kW or something on the average car.

[EV_Builder]

Nice sheets! I do want to point out the following. The cross section and power of the cable should carefully match with the fuse. I know I'm adding in another choice but remember that if the cable get hot the resistance goes up and the current might drop and then the fuse won't trip. You basically burn/melt down your wires then...
The excel could be expanded with a graph and wire starting temperature. That graph could be overlayed with the graph of the fuse. Fuse should be weakest link. Re-useing vehicle cross and fuses should be safe except if you change length.