If I were to replicate Nissan ePower

[Scrappyjoe]

I’m putting this in General because it’s more of a chew-the-fat sort of question.

If you were to copy a Nissan ePower drivetrain in a conversion project, would you

A) Just get a salvage ePower vehicle, move across the components and reverse engineer how to control it?
OR
B) Sling together the components (a Nissan micra engine, an EM57 as a generator, an EM57 as a drive unit, and a 1.5kwh battery as ‘energy cache’) ?

I didn’t ask ‘should I do this’, but in case you wonder, I live in a place where batteries are very hard to come by, so it might be cheaper.

Further detail (here I can be boring):

It’s a hard thing to reason about, but I suppose I’m trying to make a judgement call around which (A or B) has a greater chance of success. That, in turn, probably depends on how ‘open to hacking’ and ‘similar to Leaf’ the ePower components are.

It’s actually quite a bit cheaper to ship myself an ePower from Japan than a Leaf (or two I suppose) for parting out (I wouldn’t be able to register it in my country though because of import laws). That’s why I’m considering this route.

Nissan ePower is basically a Nissan micra engine bolted to an EM57 motor and inverter that feeds electricity to an EM57 drive unit. It has a small battery to store energy - about 1.5kwh from what I can tell. The platform has been very successful in Japan for the last 5 years, and is now rolling out globally (with EM47 though).

Promo video from Nissan -

[arber333]

I’m putting this in General because it’s more of a chew-the-fat sort of question.

If you were to copy a Nissan ePower drivetrain in a conversion project, would you

A) Just get a salvage ePower vehicle, move across the components and reverse engineer how to control it?
OR
B) Sling together the components (a Nissan micra engine, an EM57 as a generator, an EM57 as a drive unit, and a 1.5kwh battery as ‘energy cache’) ?

I didn’t ask ‘should I do this’, but in case you wonder, I live in a place where batteries are very hard to come by, so it might be cheaper.

Further detail (here I can be boring):

It’s a hard thing to reason about, but I suppose I’m trying to make a judgement call around which (A or B) has a greater chance of success. That, in turn, probably depends on how ‘open to hacking’ and ‘similar to Leaf’ the ePower components are.

It’s actually quite a bit cheaper to ship myself an ePower from Japan than a Leaf (or two I suppose) for parting out (I wouldn’t be able to register it in my country though because of import laws). That’s why I’m considering this route.

Nissan ePower is basically a Nissan micra engine bolted to an EM57 motor and inverter that feeds electricity to an EM57 drive unit. It has a small battery to store energy - about 1.5kwh from what I can tell. The platform has been very successful in Japan for the last 5 years, and is now rolling out globally (with EM47 though).

Are you sure there are two EM57 motors? That means quite some power n the system.
Like we discussed before. You can go the "replace the brain board" route hich is more expensive, but can liberate more power. You have several systems to choose from, like OI or Lebowski or even Axiom .

Or you can get a complete car and try to reverse engineer its CAN commands. This route would be the most rewarding in the long run since you wouldnt have to take OEM stuff apart. However any OEM limits would be transferred to your system.

Petrol engine is juat a source of rotation for the generator. If you can get the engine to turn at optimal RPM than you can command generator with negative torque and it will produce power. If you load the gas engine too much you can make RPM feedback loop which would reduce torque at falling RPM and keep the engine happy.

Regarding the battery. Motors and inverters dont care where the electrons come from. You can have a pack of lead acid batteries or prius battery for testing.
Or maybe you can get a 3phase autotransformer and rectify its output for 400Vdc and some capacitors to feed your system for testing... no regen then .

[Scrappyjoe]

The Wikipedia article on the Nissan Note ePower lists the drive train as comprising 2 x EM57 motors under the ‘Second Generation’ specifications. I’m fairly sure they are both EM57 from eyeballing pictures. What I am much less sure about is the inverters; specifically, whether the generator inverter would respond in the same way as a leaf inverter if subjected to a Lebowski for instance.

The engine also seems to be a really common one - the HR12DE engine is in millions of vehicles.

Maybe one can get 2xEV conversions out of each ePower vehicle

Can you expand a bit on this part? I don’t quite understand -

you can command generator with negative torque and it will produce power. If you load the gas engine too much you can make RPM feedback loop which would reduce torque at falling RPM and keep the engine happy.

Do you mean that I can set a simple self stabilizing feedback loop, where the ‘requested draw’ from the drive motor results in higher draw from the generator motor which in turn simply puts the engine under higher load? This feedback loop is the one I am most unsure about. If you know of any web resources - a YouTube video or something -that demonstrates it I would be most appreciative.

[arber333]

Can you expand a bit on this part? I don’t quite understand -

you can command generator with negative torque and it will produce power. If you load the gas engine too much you can make RPM feedback loop which would reduce torque at falling RPM and keep the engine happy.

Do you mean that I can set a simple self stabilizing feedback loop, where the ‘requested draw’ from the drive motor results in higher draw from the generator motor which in turn simply puts the engine under higher load? This feedback loop is the one I am most unsure about. If you know of any web resources - a YouTube video or something -that demonstrates it I would be most appreciative.

If you consider petrol engine, its response time is a bit slow in respect to electric motor, by an order of magnitude. So best thing in my opinion would be to isolate petrol engine to its own control and adapt/hook generator control to some relevant output from the petrol engine control.
Best way in my opinion would be the RPM band. That information would then be used to directly reduce/increase torque demand on generator with a variation for lag or ramp.
There is no need to keep a strict control of engine RPM but due to efficiency i recommend to keep engine in its "best torque RPM area". Generator can be adapted.
How to do it? Well how do OEMs solved this in autogearboxes? If the generator has a direct connection inside gearbox casing why not use generator resolver report for carrier .
If not a simple encoder can be made, or it is already on the motor via inductive pickup and steel rotating ring with 3 missing holes to signify full rotation.

[Scrappyjoe]

What I’m taking from this is that the slickest route would be to buy a salvage ePower and attempt a transfer. If it fails, I’ve got the backup of having lots of components for a leaf-type conversion. But first I should do whatever I can to really understand how the ePower drive train really works - try get CAN dumps, part numbers, diagrams etc. Both to increase likelihood of success, but also to verify there isn’t some design detail that renders it an unsuitable platform.

This suits me, as it means I can keep ‘advancing’ along my learning without spending any money yet!