State of Rex hacking?

[Scrappyjoe]

Hello,

An i3 Rex has appeared in a junk yard in my country (battery gone unfortunately). I have a goal to create a 'standalone' range extender that I can put in the back of my (not yet built) Outlander-driven EV conversion to be able to drive further now and then. I'm pretty sure I can take my pick of the i3 components.

So, I want to ask the community. What is the state of i3 Rex hacking? Is it at the point where I could pull the parts from an i3, mount it in a steel chassis along with a Zombieverter, and fire it up at will? I haven't figured out how it would interface with the HV bus on an EV, but a good starting point would be to try replicate how it works on the i3.

[Woodfie]

Have just purchased a REX, to figure out how to use it..
Will be staring from basics, I.e, an engine, and a generator, then see what is required to transfer charge to a 330v system , as a range extender..
Learning curve start point.

[Woodfie]

Have collected the i3 REX, which is covered with lots if ancillary hardware. First step will be to mount on a frame, and take off the unneeded hardware, to simplify the rig.

Does anyone have any Specifications on the Valeo generator, 2621582 ? Still searching.

The engine is reported at 25Kw, but enquiries about the generator , reveal that in operation the i3 will drive at 60 to 70 kph on the flat...
This would indicate say 6KW ??? Which seems low for the size and about 80kg bare or 100kg in vehicle specs.

[Woodfie]

Have removed some of the conglomeration of cooling plumbing, and removed the airbox with some connectors being rather awkward.. this has revealed the spark plug igniters, which could be useful, and the fuel rail is in easy view..
So far no specs located on the valeo gen/starter.
So have proceeded to separate from the engine . A very grudging extraction, seemed like the spline was locktited in... the housing has threaded holes, which appear yo be designed for Jack screwing the generator off. Suitable Screws and spacers allowed removal of the spline which uses a Hex ended torsion rod , to anti backlash, and possibly work as a dampener connecting the engine to the generator, which is effectively its flywheel.
Of electrical interest, spun the rotor with a drill, at nameplate 1500rpm, and produced AC 33V rms.
There is a 5 lobe resolver, and will do a locked rotor sequence to determine the pole pair count. If this is 10, or 5 pole pairs, then it maybe possible to use a Nissan leaf gen1 inverter and Thunderstruck vcu, to start and charge to the bus voltage.

[barracuda816]

This is really interesting work.

is the generator cooled also? If not that could account for the low kw rating?

[Woodfie]

Yes, The generator is water cooled, amass of pipes, and a water to water heat exchanger for some complication.

[jrbe]

How much of the rex controls do you have? ECU, harness, etc? I assume it has a fly by wire throttle body too?

If they're outputting 25kw from the combustion engine they likely are sending that minus losses to the traction motor then the balance to charge battery.

I was reading up a bit on it. Sounds like there's at least 2 power output modes from sourceless Internet info. A low vehicle speed mode claimed at 12kw then the higher speed / load 25kw.

Would be great if utilizing these was some can bus messages to figure out.

[Woodfie]

How much of the rex controls do you have? ECU, harness, etc? I assume it has a fly by wire throttle body too?

If they're outputting 25kw from the combustion engine they likely are sending that minus losses to the traction motor then the balance to charge battery.

I was reading up a bit on it. Sounds like there's at least 2 power output modes from sourceless Internet info. A low vehicle speed mode claimed at 12kw then the higher speed / load 25kw.

Would be great if utilizing these was some can bus messages to figure out.

Have only the REX motor and Generator assembly.

Unfortunately none of the control or loom available.
Am hunting for a plug for the 3 phase motor connections, which are difficult to connect without the original loom bit.
So no CAN information or control hardware,
The Generator has been removed, and aiming at using an inverter which can run the 5 Pole pair, and 5 resolver Lobe Valeo generator, as a motor to start and run the ICE.
Leaf inverter has 4 pole pair and 4 lobe resolver, so maybe usable.
Prius inverter has 4 pole pair, but 2 Lobe resolver, so is unfortunatly unsuited .
Much to learn..

[jrbe]

I haven't seen any harnesses / controls for sale. It's a lot of work either way, factory spoofed commands or custom controls.

If you have to go with standalone, at least to get the combustion engine running you could look at going with a speeduino. They have fly by wire throttle, closed loop, sequential injection, etc. They have Atmel, STM32, and teensy designs but you'll have to dig some.
https://speeduino.com/forum/
The fixed speed generator modes will likely have to be coded along with communications.

A power stage design might not be that bad for something like this along with a OI main board. 100a / 600v Mosfets should do along with some good cooling.

[jrbe]

Would you need a resolver? As long as the combustion engine could tolerate spinning backwards a small amount you could probably run without one.

[Woodfie]

Presently checking for resolver usage. If can spin with a Leaf inverter with known leaf CAN info, will be able to start the ICE

Stand alone at this stage. Speeduino looks promising, may also do a hall effect switch, and bicycle sensor magnets, for Ignition, and use Lpg as first run, to see what can achieve.

Picture of the separated Generator, with the torsional drive , anti backlash, possibly engine pulse dampening, and menace to extract and refit Splined Coupling
In the foreground.

[Woodfie]

A step further.
Attempts to locate a Hv plug for the REX generator have not been successful.
A dismantle to obtain a suitable UVW connection was undertaken
Some interesting finds.

Coils are 2.00mm wire wound... how many Amps ??

15 windings , as in 5 per phase. AL 5 on one phase appear to be in parallel.. My guess is maybe an eyeball at 30 A in the winding, X 5 , so 150 A per phase, so maybe 250A total at say 50v, So that would be 12.5 Kw ish.

Think there was mention of 12Kw mode, but would think this would do 80 to 90 kph at around 120wh/ km.

Can anyone confirm what an i3 can do in ReX run when the main pack is depleted.?

[jrbe]

Rex or reme starts on page 142.
I changed the name so it hopefully doesn't get hunted down and pulled.

Components, high voltage.pdf

(4.5 MiB) Downloaded 81 times

[Woodfie]

Thanks for that...
Still more curious, the 23.3 Kw stated sounds Nice.
That's much more than I feel the motor hardware looks capable of, But more to Learn.

My Big Baffle is why only around 6 or 7 kw are used in operation..
Is it some USA regulatory nobbling?
I understand that the REX was made to only function when battery at 5% SOC, a USA requirement.
The Original software allowed the Rex to operate on request. And heard of one that had been reprogrammed several years ago, to earlier status.

[jrbe]

I'd expect the answer has to do with efficiency and emissions. Thought of differently, lightly accelerating up a hill is more fuel efficient than flooring it. If the rex is making enough power to cover average cruise power and lightly charge the battery it's likely running at the ideal efficiency.

There may be some pzev or some other acronym requiring the rex only kick on at essentially an empty battery.

5% battery might really mean 25% battery the way OEMs state the charge level.

[Woodfie]

Yes the i3 was an "efficiency" designed vehicle in EVmode
So if a foot heavy Test Journalist was able to make it look like an ugly fuel performer , some sadistic glee would make it good reading for the 2012 Ev knock lot.
Instead they were only able to report how awful it was to be speed limited in its "get you home " mode, which fitted the the High Efficiency Vehicle design aim.

My script is to obtain useful onboard charging on the fly, as a get to a charge point, when one of the inevitable, out of the ordinary it shouldn't occurances happen.
A bit like our one in a hundred weather events, that happen bi annually !!

So monitoring the internal temperature and loading to achieve up to 70% of max output, and check fuel usage is acceptable and not abhorrent is the goal am hoping for.

[bundao]

Thanks for uploading the "Components, high voltage" file. But I wondered if anyone has the associated i3 Wiring Diagrams manual? If not does anyone know if "Service Manuals Online" is a safe website to purchase manuals from?
I'm looking for details on how the Park Actuator is driven. Mine looks ok upon inspection, but it wont activate so I want to check if something has failed in the EME.

[Woodfie]

I'm looking for details on how the Park Actuator is driven. Mine looks ok upon inspection, but it wont activate so I want to check if something has failed in the EME.

Have a friend ,with the same issue as you describe.
It had become worse, over a brief time, then failed altogether.. The details talk of solenoid in the EME,
Could be Bavarian for relay maybe, wondering if that would fail in a progressive manner?

[Phate]

Can anyone confirm what an i3 can do in ReX run when the main pack is depleted.?

Based on daily driving my 2017 i3 REx for the last 4ish years:

Between 65-70mph on relatively flat ground with mild climate control usage. I believe it runs at the higher power mode above 56mph.

As-delivered, US-market i3's would turn the REx on at 7-8% SoC. This along with a "less REx range than battery" requirement were done so BMW could classify the car as a BEVx/"range extended EV" rather than a plug in hybrid. Basically it runs for about an hour on a tank at highway speeds - they never increased the fuel capacity to match the increases in battery capacity.

It takes all of ten minutes with a phone app and bluetooth OBD dongle to code the car to re-enable the "Hold State of Charge" option so you can manually enable the range extender anywhere below 75% SoC, and it makes the car massively more useful.

All SoC references are based on what shows up on instrument panel - the 94Ah/33kWh i3 has about 4kWh/12% of the actual capacity reserved as a buffer.

Let me know if there's anything else you'd like to know. In general my biggest complaint with the i3's REx is that it needs just slightly more power, and it needs a larger fuel tank. Still, not bad for a detuned kymco scooter engine.

[Woodfie]

Great information, Thanks you.

Have just achieved the Generator spinning over oñ being patched to a Nissan Leaf gen 2 inverter,
Will see how can post this progress, am challenged with videos loading... Harder than making German electrics run on Japanese electronics..for me at this stage.

[Woodfie]

[uhi22]

The details talk of solenoid in the EME,
Could be Bavarian for relay maybe, wondering if that would fail in a progressive manner?

No, not really Bavarian, and not a relay, but an electromagnetic actuator. So it is directly the part which pulls/pushes the lock mechanism. A combination of a fixed coil and a movable permanent magnet.

[Woodfie]

""Between 65-70mph on relatively flat ground with mild climate control usage. I believe it runs at the higher power mode above 56mph.""

Thanks Phate for the assistance, A further question, if you have run the battery right down low.

Re the REX operation,
When down to the lowest level as in 6% SOC indicated, where there is effectively no draw in the battery,
Is the 65-70 mph available, OR is that only when both ghe REX and the battery are contributing to the drivetrain?

Just trying to determine what the REX'S power alone is by reference to it's performance when battery is depleted

[Woodfie]

With much premeditation , heaving levers and printing engineer's assistsnce, the Non serviceable Torsion spline has been re Torqued..
As the unit is extracted, the tension releases and the parts all fall apart... So there Was debate as to whether the tension should be forward and giving miniscule dampening, or reverse and the torsion being a backlash eliminator, with solid forward spline drive. Anyone have any idea..? Have gone gor the forward option..

With some guiding threaded rods and carefully assembling, the Torsion spline has been reassembled to couple the Motor and Generator ...
Best advise for this is the same as Hitchhiker's guide to the Universe about Bogon Spacecraft hitching...Don't do it. Where possible Don't separate the motor and generator.. Bmw manual requires you to replaced the Torsion spline ..
Good news is it's all back together and preparing to turn over with the Leaf inverter...

[Woodfie]

Reconnected to Leaf inverter, to test run engine turning as preparation for Ignition and fuel run.