Which Electric Vehicle would make the best 3 phase grid power emulator?

[galstaf]

I am looking to take the battery, charger, inverter etc off a crashed electric car/truck and have it power a 240/480 VAC 3 phase motor. Nominal power consumption is 7.5kW. The voltage seems to be able to be anything from 208VAC to 480VAC, so pretty flexible.
It runs currently off utility grid 3 phase power, but I would like to be able to move the setup and run this "offgrid" for a few hours at a time.

The inverters for electric vehicles also make 3 phase if I understand correctly. What would it take to hack the EV system to produce a steady 20 to 50amps of current at 200 to 500VAC 3phase to power an regular 3 phase industrial motor?
I am assuming by now someone has a hacked inverter controller to have this be possible.
Could anyone recommend a make and model of EV and a controller hack to make this work without too much messing around? I was thinking a Nissan Leaf would be a good size of battery to do this. Something around 30 or 40 kWh.

Recommendations or thoughts anyone?

 

[galstaf]

BTW this is not a constant load. For a sawmill so it runs for about 60 seconds at a time with at least a minute or two break, so about 50% duty cycle tops.

 

[JackFlorey]

I am looking to take the battery, charger, inverter etc off a crashed electric car/truck and have it power a 240/480 VAC 3 phase motor. Nominal power consumption is 7.5kW.

Problem is that you are going to need to match the motor constant. Most EVs are designed with current fed inverters and are run from a 360 volt battery, so they are going to be designed to output a drive waveform that peaks at less than that. Might work at 208 volts. Also note that most EVs are torque mode, not constant speed, so make sure that works for you.

 

[galstaf]

Problem is that you are going to need to match the motor constant. Most EVs are designed with current fed inverters and are run from a 360 volt battery, so they are going to be designed to output a drive waveform that peaks at less than that. Might work at 208 volts. Also note that most EVs are torque mode, not constant speed, so make sure that works for you.

Thanks for the response. Not entirely sure I understand that. Don't most EVs support "cruise control" which would give a constant rotational speed despite the load?
Wouldn't the inverter not supply the motor more current on demand as the motor is loaded?

I have tried to find a good write up of how these automotive inverters work. Do you know of one?

 

[galstaf]

so they are going to be designed to output a drive waveform that peaks at less than that. Might work at 208 volts.

Could you clarify with what it peaks at exactly? Why did you mention 208V versus 240 for example?

 

[JackFlorey]

Don't most EVs support "cruise control" which would give a constant rotational speed despite the load?
Wouldn't the inverter not supply the motor more current on demand as the motor is loaded?

I have tried to find a good write up of how these automotive inverters work. Do you know of one?

1) Yes. But cruise control does not set the rotational speed of the field. It measures vehicle speed then increases or decreases torque to maintain that speed. That's why, if the cruise control cannot maintain your speed up a hill, the car simply slows down, rather than "losing lock" on the motor.

2) It depends on the design of the motor and inverter. Again, EVs use a torque command - they give you the amount of torque you tell it to by pressing on the accelerator. If you are going up a hill and don't move your foot you will not get more power.

3) I had a good rotating machines book in college but I can't find it now. There's plenty of stuff on the web.

 

[JackFlorey]

Could you clarify with what it peaks at exactly? Why did you mention 208V versus 240 for example?

A current fed inverter always outputs less voltage than it inputs. So a 360 volt supply can give you 360 volts peak. Since peak is 1.4x the RMS value, that means you could generate a 257 volt sine wave, so 240 volts would be no problem. At 300 volts (min voltage for many EV packs) you could get 214 volts RMS, which means if the motor is designed for 208 volts 3 phase it would work.

You can also boost inside the inverter, or prior to the inverter. The old Priuses used to do this, and pretty much every 48 volt inverter does this. The Prius boosted from ~200 volts to ~500 volts.

 

[Chalo]

I am looking to take the battery, charger, inverter etc off a crashed electric car/truck and have it power a 240/480 VAC 3 phase motor. Nominal power consumption is 7.5kW.

Why wouldn't you use the EV motor as well, and skip the whole AC motor nonsense?

 

[galstaf]

Why wouldn't you use the EV motor as well, and skip the whole AC motor nonsense?

Lol.. that was my original intention... to convert an existing mill and use the EV motor and entire drivetrain from battery to wheel hubs.

However isn't a 3 phase industrial motor pretty much the same animal? The EV motor gets 3 phases of power also, right? Plus wouldn't the EV motor be a lot bigger and heavier?
The 7.5kW motor that is currently on the machine is about 20 inches long and about 14 inches in diameter. It is definitely not tiny either.

 

[galstaf]

A current fed inverter always outputs less voltage than it inputs. So a 360 volt supply can give you 360 volts peak. Since peak is 1.4x the RMS value, that means you could generate a 257 volt sine wave, so 240 volts would be no problem. At 300 volts (min voltage for many EV packs) you could get 214 volts RMS, which means if the motor is designed for 208 volts 3 phase it would work.

You can also boost inside the inverter, or prior to the inverter. The old Priuses used to do this, and pretty much every 48 volt inverter does this. The Prius boosted from ~200 volts to ~500 volts.

I am trying to keep this as simple as possible with as few modifications as possible. Not sure boosting is needed plus I have no idea how to do it

If a 360VDC pack can give me 257VAC RMS, then that is about perfect. The motor would start to slow as the voltage depletes, but definitely not the end of the world.

 

[galstaf]

1) Yes. But cruise control does not set the rotational speed of the field. It measures vehicle speed then increases or decreases torque to maintain that speed. That's why, if the cruise control cannot maintain your speed up a hill, the car simply slows down, rather than "losing lock" on the motor.

This is exactly what I am looking for. As the load on the electric motor increases, the inverter compensates by throwing it more current to keep it moving at the same rate.
Eventually I suppose it could bog down if I push too hard by trying to saw something ridiculously hard with a blunt blade but I don't think that is wise.
Keep the blades sharp and slow down according to the wood being cut and the load on the motor is never excessive.

 

[galstaf]

2) It depends on the design of the motor and inverter. Again, EVs use a torque command - they give you the amount of torque you tell it to by pressing on the accelerator. If you are going up a hill and don't move your foot you will not get more power.

One of the advantages of electric motors is their almost instant and steady torque curve in most circumstances. In this particular application the load or "steep hill" is modifiable by the user. If the load gets too big, the user simply slows down and or uses a sharper blade.
With a sharp blade on the saw currently, the 7500W motor will go through a 25 inch wide pine tree like butter. However big knots or branches do slow you down. You have to listen to the machine and act accordingly.

In terms of torque command and especially emulating "cruise control" to pick a particular target RPM. That can be done with a CAN bus hacker to fool the EV system into thinking it is still in it's original vehicle, right? That's what I am seeing is possible, but a lot of this stuff is still out of my current knowledge range.
Given my criteria, can anyone suggest a good (hopefully low cost) crashed source vehicle and control system? Preferably well documented and open source.

I originally thought of the Nissan Leaf as a good start but would love some other ideas if there is a better, easier to control source vehicle.

 

[JackFlorey]

However isn't a 3 phase industrial motor pretty much the same animal?

Depends on the motor. Most 3 phase industrial motors are induction motors, which spin at close to 3600 RPM (or some other multiple of 60Hz.) The "close to" part is important because induction motors need some slip to maintain their internal fields.

Synchronous motors spin at _exactly_ 3600 RPM (or again some other multiple) since they have salient poles that align with the rotating field.

As to the inverter, the power sections for both inverters will be very, very similar - but the method of driving them (i.e. the intelligence that drives the power stage) will be different.

As the load on the electric motor increases, the inverter compensates by throwing it more current to keep it moving at the same rate.

Generally an inverter won't give you that as-is - but you could certainly do an external control loop to sense the RPM and feed back an error signal to keep it at the same RPM.

Note that if you have an induction motor you will have to _increase_ the RPM of the field to keep the motor spinning at the same RPM. As you load down the motor, the slip will increase.

As to how to do that with a given EV equipment set - that's a long and involved discussion, and is going to be different for every EV.

 

[galstaf]

Generally an inverter won't give you that as-is - but you could certainly do an external control loop to sense the RPM and feed back an error signal to keep it at the same RPM.

Note that if you have an induction motor you will have to _increase_ the RPM of the field to keep the motor spinning at the same RPM. As you load down the motor, the slip will increase.

Thanks for bearing with me. I really don't know what I don't know at this point and appreciate your patience!

The motor I am dealing with is almost certainly a standard industrial 3phase induction motor. There is no electronic motor controller or Hall sensors or any of the fancy BLDC motor stuff.
So I take it most EVs are synchronous motors then? Are they generally issued with Hall sensors or no?

I was wondering if I do get say a Nissan Leaf if I could take the Hall sensors off the car motor and mount them on the industrial motor so the inverter knows what it is up to, and could increase the current if the motor starts to slow under load.
Is that was you mean by external loop control?


If you can recommend a good write up for a somewhat technical reader that explains some of this stuff, I would be extremely grateful. I hate to pollute the boards with dumb questions if I can do some background reading.
Then I can post slightly less dumb questions!

Many thanks!

 

[JackFlorey]

The motor I am dealing with is almost certainly a standard industrial 3phase induction motor. There is no electronic motor controller or Hall sensors or any of the fancy BLDC motor stuff.
So I take it most EVs are synchronous motors then?

EVs use a lot of different motor types. Leaf motors are sychronous PM (similar to BLDC.) Model S motors are pure induction. Some are hybrids. At least one is switched relutance.

You can't run an induction motor with BLDC sensors; the field has to spin faster than the rotor, and the sensors won't let it if it's really trying to drive a BLDC motor.

You can use the power stage of almost any motor inverter to run almost any motor. It's the control scheme for the power stage where things get different. An inverter that "expects" an induction motor will be easiest to drive.

 

[Chalo]

Lol.. that was my original intention... to convert an existing mill and use the EV motor and entire drivetrain from battery to wheel hubs.

However isn't a 3 phase industrial motor pretty much the same animal? The EV motor gets 3 phases of power also, right? Plus wouldn't the EV motor be a lot bigger and heavier?

EV motor is sure to be lighter and smaller per maximum HP, but maybe heavier or bigger if it's much more powerful than the industrial motor.

My point is that the motor controller and interface is all worked out, if you use the motor and controller and battery together the way they were designed. Then your challenge is just packaging and structure and human interface, not all kinds of back end electronics and motion control and tuning and troubleshooting fault conditions.

 

[galstaf]

EVs use a lot of different motor types. Leaf motors are sychronous PM (similar to BLDC.) Model S motors are pure induction. Some are hybrids. At least one is switched relutance.

You can't run an induction motor with BLDC sensors; the field has to spin faster than the rotor, and the sensors won't let it if it's really trying to drive a BLDC motor.

You can use the power stage of almost any motor inverter to run almost any motor. It's the control scheme for the power stage where things get different. An inverter that "expects" an induction motor will be easiest to drive.

Thank you again, Jack!
Interesting point about the Tesla S. So that would likely be the best choice to drive a non Tesla induction?
Do you have any idea what commands the computer sends to the inverter to have it drive an induction motor?

I was hoping to use a Nissan Leaf as they seem to be extremely reasonably priced for a crashed model, and my power requirements for this application are pretty modest so only really need a smaller 30kWh battery. Leaf's also seem to have a pretty robust hacking community although I have no idea which way to turn right now for a modestly priced open source solution to make the inverter just do something relatively simple.

How difficult is it to just have these inverters emulate standard US 240VAC 60Hz power?
One would think that would be fairly straight forward as it is a very simple one speed deal, right? Or am I again misunderstanding?


Good point on the stator field on the induction motors. I did know that at one point but I appreciate the reminder!
Much obliged for the schooling!

 

[galstaf]

EV motor is sure to be lighter and smaller per maximum HP, but maybe heavier or bigger if it's much more powerful than the industrial motor.

My point is that the motor controller and interface is all worked out, if you use the motor and controller and battery together the way they were designed. Then your challenge is just packaging and structure and human interface, not all kinds of back end electronics and motion control and tuning and troubleshooting fault conditions.

Excellent point, Chalo!
The sawmill in question runs off a large smelly ICE diesel in other versions of mine.. so it is designed to lug around a heavy motor on the saw head.
This video shows the diesel motor about 3 minutes in if you don't know what this kind of sawmill looks like. As you can see.. it has all the usual ICE things... battery, fuel, radiator etc etc

Back to the EVs tho... once the drive train is out of the car, does it not need to be "hacked" regardless with a new controller to force to components to work outside of the native environment of being a car?


It should be a fairly simple control if I have any grasp of this at all (no guarantee of that!)... The load curve is pretty predictable (cutting a tree or not cutting a tree).. no change in speed, rapid acceleration etc etc.
The motor is ancient tech at this point.... I am pretty sure induction motors have been around for almost a century.
Is it that hard to get an inverter to send a constant speed sine wave pulse of current on each phase 60 times a second? If so, could someone explain why?

Or direct me to a good write up explaining this. I have googled as many ways as I can and cannot find something decent that can explain this situation to me.

Thanks for your insight!

 

[amberwolf]

It's not that simple, unfortunately.

If you know the correct commands to send to an inverter, or it is simply operated by an analog throttle input, then whether it will work depends on what the throttle modulates. On most EVs that would be torque, or phase current. What you need it to modulate is speed, to "set" it to a single speed that equates to the 60hz you're after. If the inverter you want to use can be set to a fixed speed, and you know the commands to do that, then it's "that easy".

BUT:

For voltage regulation, you'd need a separate feedback circuit that monitors phase voltage, and modulates the phase current output to maintain that. The controllers aren't normally required to do this, so it isn't a function any of them are likely to have.

There are open-source systems based on VESC that you could rewrite the code for it to do this, after adding the necessary hardware to monitor phase voltage (if it doesn't do this already). Some of them are here on ES, others are elsewhere. However, I'm not sure there's many that can do the voltage you want.

Or you could "decapitate" an EV's powerstage and create your own control software for it, running on the MCU of your choice. The control software would modulate the current at the fixed speed to create the required voltage at the load you have.

You could probably use the VESC core for the base of that. Or the Lebowski control board (which is documented for replacing the control portion of at least a few inverters, like the Honda IMA here on ES, and somewhere out there (linked in my version of that) for one of the Tesla inverters and also a Leaf, and i forget which other ones.