Controller mods or build for E-assist altermotor

[HaroldinCR]

This thread is being started to take all the non-relevant off topic posts off MPaulHolmes 3 phase controller build thread.

Would/could a moderator move ALL the BAS E-assist information from the 3 phase controller thread, into this thread ? Thank you.

Further discussion input on my part will proceed once the moderators have cleaned up the other thread. Anyone that wants to participate on this topic or relevant topics, feel free to do so.

[cajunfj40]

FYI, if you do a search for this part on car-part.com, it is listed as the alternator for the 2012-2014 Buick Lacrosse with eAssist 2.4l engine. Prices are stupid-low, though I'd be wary of unlisted core charges. There's one for $100 29 miles from me, and $50 225 miles from me. Since the BAS motor/generator replaces the standard alternator completely, there's no other alternator on the car that could be confused with it. One search result has a photo, and compared to some cutaways online of the eAssist system, it is clearly the correct part.

Looks like it has a lot of potential.

[thingstodo]

Ebay lists a used one at $159 canadian, with free shipping

2012 2013 2014 2015 2016 Buick LaCrosse Regal 2.4L Alternator 40K OEM | eBay

The one on the left seems to be it?

$159 for 27 HP seems like a steal!

[cajunfj40]

Hello thingstodo,

I think all of the ones pictured in that ad are plain ordinary alternators. The eAssist one has a smooth outer housing rather than a vented-looking one. A lot of the Ebay listings have something like "...3 6L..." in the description string, which looks suspiciously like 3.6L, which is the larger non-eAssist engine option in the LaCrosse. Searching for a 2013 LaCrosse leather sedan with 2.4L engine gave me a result showing a brand new eAssist generator/starter for $300.

Beware poor cross-listings. Human and/or bot error does crop up often.

[thingstodo]

Originally Posted by cajunfj40

Beware poor cross-listings. Human and/or bot error does crop up often.

Thanks for the pointer!

I drive by a wrecker on my way home. They are closed for the day by the time I drive by, but I could maybe drop by on Monday and see what they have, or what they can get.

These are the vehicles that Wikipedia lists under BAS hybrid. Any others that I could cross-reference ... or would I rely on the wrecker to do that cross reference?

2012-2014 Buick LaCrosse w/ eAssist
2012-2014 Buick Regal w/ eAssist
2013 Chevrolet Malibu w/ eAssist
2014 Chevrolet Impala w/ eAssist

[foxgrockefeller]

Found one from 2013 malibu for $100, did all 2013 malibu's have this system?
I would be super excited to get this thing working, would be very cheap EV drivetrain for anything light.

[HaroldinCR]

YEEeeeepppp, Y'all done found out where they are. Mostly these prices are a little negotiable, especially if not through LKQ or Fenix corporate yards.

I believe the 8" X 8" would be very close to dimensions. Shaft is NOT tapered.

I'm dodging rain showers trying to get projects out of the way, but, will try to add much more information maybe late today.

That core charge is just double dipping of money by salvage yards. Ain't nobody gonna rebuild one of these.

[HaroldinCR]

Looks like one of the moderators has brought all the info from Paul's 3 phase controller thread, to this thread. Thank You to whoever did this.

Now for some more information: From my contact in Spain. We had discussed using the Lebowski chip, but, Lebowski said he had no time to get into the code writing, researching, so, here we are.

( I know someone that could design any inverter IGBT or MOSFET (paying him of course) if we had the controller, but controlling an AC induction motor is the most complicated/costly part, and using a commercial VFD is not safe neither efficient)

(e-assist motor can work at 65V (500A) with the delta modification. I think it is a good voltage for a 15-20 HP motor
It is a delicate modification but anyone with welding or plumbing knowledge can do it. It is necessary to use silver wires to weld the copper wires (to increase conductivity of the union)

(I would never pay an external company for such a project, it is extremely complex and the know-how would remain outside. Any further modification would require their additional work (and charges). And also you would never have a SW free of errors, because any error detected after the project is considered closed would have to be discused to decide if it enters in the warranty or not, which causes the project to be ethernal in time.
I worked on SW companies for years, and I know how they work)...

(At the moment I own a company that is focused on SW for dynamometers (sportdevices.com). My interest in electric motors is basically to create a tool for measuring all important features of these motors like the efficiency, with the motor + controller as a "black box" (not controller alone, or motor alone). For this reason I am doing some investment in several commercial parts as motors and controllers. At the moment I have two Curtis (80V) and one Sevcon Gen 4 (which I couldn't start by now). And 3 e-assist motors and one ME115 water cooled motor.
I am spending some money to not have to develop any hardware by now, but in the long term I have the aim of developing my own inverter and controller, but as this is very time costly and we have lots of pending tasks related to daly work it is being delayed and delayed.

(There is an electronic engineer in the company (for this reason I told you that the inverter can be done here), but the controller SW should be written by me, and this is not possible at the moment).

(If we could arrange a team to cooperate to build such controller, we could provide all inverter design for a high power application, say 150V and 500A for Mosfets or 400V 400A for IGBTs, but it should be very carefully arranged and coordinated to no end in a total failure project
Also we can provide some tools for measuring purposes, we build software and electronics for inertial and braked dynamometers, and recently have tested a simple extension to measure battery and current voltage (efficiency is motor power / battery power)

(The point with ACIM motors is that it is not enough it works like PM motors (there are lots of controllers for PM motors: sensored, sensorless, BLDC, vectored, etc), it also has to be "the best one" because control strategy and parametrization in ACIM motors is a whole, if it is wrong the motor will draw a lot of heat and current. For instance for the same torque you can have higher slip (frequency difference with actual speed) and less current, or low slip and high current. There is an infinite set of combinations current and slip that provide different efficiencies for same torque. And also there are several combinations depending if you need max absolute torque for a specific rpm or max efficiency (it could be simplified to two operating maps).
As you can see it is not enough the controller works, it has to be also optimal and provide several modes of operation (for PM motors this is almost fixed). Self-tune process is also important as most of us would not be able to do it manually, but for that a high level of knowledge about ACIMs has to be reached.

(It is not so simple as adapting a PM controller to ACIM. PM can even work with a simple "BLDC" control (something similar to activate the right switch as the motor turns) and including the V/F parameter (KV) for that specific motor, but ACIMs need vector control to work in an efficient way. V/F algorithms are totally not recommended for traction applications).

(My main concern is investing too much in ACIM motors (parameter measurement, efficiency maps, etc) and discover afterwards that people mostly use PM motors, then I would have wasted most time
But also, I can see that there are much more solutions for PM motors, so I am not sure if could provide some interesting solution to improve what is already available. Honestly, I am bit lost)

(using a voltage converter is not a good idea, in general it would be a bad design:
* it increases the losses (minimum 10 or 15%)
* it adds a lot of weight and volume to the vehicle (the transformer for 15KW is huge)
* it increases the complexity and probability of failure
* you only save batteries in series (and BMS), but you add the converter
* you still not simplify the motor inverter, which still needs to operate with the motor voltage
* currents are multiplied to values that could be simply ridiculus. For instance if inverter and motor handle 400A, the converter (for 1:2 conversion) would handle 800A or 900A, which starts to be simply too much
for me it makes much more sense simply put more cells in series which will also add range and keep the current values low).

(Keep in mind that a converter over 500W or 1000W starts to be very big. Something over 3 KW is very big, and something to reach 15 KW simply does not make sense (it will take more room and weight than the batteries, for sure)

(Regarding kelly, I never tested them, probably they are worse than Curtis controllers)

(With the Curtis 80v 350A I got about 16 HP at 55V. With the 550A version and 75V I expect to be about 30 HP or more)

ALL the info in parentheses is the info from my Contact in Spain. He has purchased the Curtis controller. Cost $2400.00 +/- USA dollars

He also gave me permission to post our discussion. My part was merely asking questions that he so graciously answered. One part about the voltage boost converter will pertain also to Paul's thread.

OK, lets get this thing working.

[thingstodo]

Paul's AC Controller already does a good deal of what is described - vector mode control of AC induction motors, control of BLDC.

The V/F or volts/herz controller is much simpler and can run without an encoder, resolver, etc. It is less efficient. But it will run multiple motors in parallel from a single controller. I have more to investigate about doing 4WD in V/F and how much the wheels will slip - the question being whether the front wheels can pull but still spin at different speeds for corners.

I think this BAS alternator/generator is a good fit. I have one of Paul's AC controllers (serial #2 I think) and I'd like to try it. The firmware may need to be changed if I find something that won't work.

As always, the connection between the motor and the wheel is my problem. I'm not good at mechanical things. You'd think I'd be able to get a belt drive working .. perhaps with a very simple tensioner that let the belt slip if there was too much load?

I will call my local wrecker this afternoon. Perhaps they have one of these E-assist motors available nearby. If I could pick one up Monday that would be AWESOME.

[cajunfj40]

Hello thingstodo,

Thanks for the pointer!

I drive by a wrecker on my way home. They are closed for the day by the time I drive by, but I could maybe drop by on Monday and see what they have, or what they can get.

These are the vehicles that Wikipedia lists under BAS hybrid. Any others that I could cross-reference ... or would I rely on the wrecker to do that cross reference?

2012-2014 Buick LaCrosse w/ eAssist
2012-2014 Buick Regal w/ eAssist
2013 Chevrolet Malibu w/ eAssist
2014 Chevrolet Impala w/ eAssist

It looks like they go up to 2016, possibly 2017. Not sure. Best get the part number off a GM parts website and use for comparison if you go look at it. The wrecker cross-reference may or may not adequately dis-ambiguate between eAssist and regular vehicles.


Hello foxgrockefeller,

Found one from 2013 malibu for $100, did all 2013 malibu's have this system?
I would be super excited to get this thing working, would be very cheap EV drivetrain for anything light.

Not all had it, just the ones with eAssist. Only 2.4L engines had eAssist, but not all 2.4L engines do. I have not exhaustively researched it.

General:

I did some note-taking:

All numbers in *engine crankshaft RPM*:

Engine has horsepower peak at 6,700RPM, so redline is probably 7,000-7,500 or so?

Generator output is 15kw(20.1hp) from 1570-3180RPM
Max cranking torque is 110lb-ft(150N-m), so call that from 0RPM or so.
Max motor torque during assist is 79lb-ft(107N-m) at 1,000RPM
Max motor power output is 11.2kW(15hp) from 1,000-2,200 rpm

Poking around I found an OK picture, so I stuck a ruler up by the screen to try and figure out the pulley ratio. It looks like a 2:1 reduction from motor to crank. Could be more, but that's an easy number to math with. Hmm. Need to divide peak torque by 2 as well, I think, to make the math work. Yes, that checks out.

So I did some quick math.

All numbers below here in *motor/generator rpm*:

So, we have a motor that has a stall or pull-up torque of 55ft-lbs, falling to 39.5ft-lbs that it can carry up to 2,000rpm where it goes to constant-power of 15hp, steady up to 4,400rpm (torque has fallen to 17.9ft-lb), and falling off in an unknown manner after that.

As a generator, it can absorb 15kw/20.1hp from 3140-6360rpm, falling off in an unknown manner above and below that range.

Stock, this thing would be a stump-puller with top speed potential roughly that of a 200-250cc gas 4-stroke. Great city/dirt bike.

If I recall correctly, the transition from roughly constant torque output to roughly constant power output happens when the back EMF equals the battery voltage, so in this case 2,000rpm. Doubling the voltage (230VDC) will roughly double the RPM (4,000rpm) to which the motor can make roughly constant torque. This will double the output power to 30HP, which will then be roughly constant to ~8,800RPM. If the winding insulation can handle it, going to triple voltage (345VDC) would net 45hp from 6,000-13,200. All at the stock max currents, so no additional I^2R losses, and still under the assumed motor redline of ~14,000-15,000rpm.

Of course, if you want to shove more current in, you can increase the power even more, or with less voltage, at the expense of more heating/losses.

If one could nab the voltage buck/boost circuit out of a later Toyota HSD system and control it, that'd allow more drive/regen voltage out of a smaller battery made from Leaf or Volt modules, or one could hack said modules to make them all series, rather than series/parallel, but I think that's more work.

I'm interested in an inexpensive controller circuit that can take single-phase 240VAC and make 3-phase power to run this motor. All I need is soft-start - variable speed is a bonus. I have a snowblower that wants converting...