Man in the middle on inverter PWM?

My Nissan Leaf Forum

Help Support My Nissan Leaf Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

chaz

Member
Joined
Nov 15, 2016
Messages
14
Location
Salt Lake City, UT
Hi all, I have one leaf and love it to death but it is time I convert my old '91 miata over.
There are a few threads on here about trying to reverse engineer the can protocol to get more torque but I wanted to bounce a new Idea off you guys.

How would one think the leaf would react if it had 2 competing sets of PWM IGBT gate signals that could be swapped at an instant.
The idea is that you keep all systems intact and run everything as normal except when shifted into forward and the pedal pressed.
Then you switch the IGBT gate signals over to a ti instaspin or similar brain board that is also sensing the pedal but can drive the IGBT with free will. I figure you would need to split or have redundant current meters on the phase legs and pedal sensor as well as a 6 channel switch.

So thoughts? what does an IGBT do when all of a sudden there is no signal? what do we think the leaf will do when it is trying to keep the FOC correct but feels like the motor is running away? I would love to keep most of the leaf control system but not if i can't get awesome power.

thanks
 
Before giving any thoughts or advice, please accept this disclaimer: Any modification done on any car drivetrain is dangerous. I am not an automotive engineer and do not recommend any drivetrain modifications. What follows is just thoughts from an engineer who has designed low-power motor systems. When in doubt, don't do it. I take no responsibility for this advice or any outcome from this advice. And above all else, please don't sue me. :)

That said, it is bad practice to float the gate of an IGBT. You can do it and may get away with it, but the risk is very high. In the scenario you describe, it would be much safer to add a logic stage before the IGBT driver and engineer that stage with a pull-down circuit so that it is assured off when switching.

Even better is to use a processor to coordinate switch-over between control signals and run both control signals through the processor. Again, this requires an IGBT driver on the output of the processor. It also requires software to deal with different scenarios.

Whatever you do, it will require rigorous testing to make sure that every possible case is covered. What if someone simultaneously pushes brake and accelerator? What if they pulse the accelerator and brake pedal quickly? What if protection signals (ex: overheat) switch on or off at the worst possible instant? You will be really unhappy if one of these unexpected events damages the motor control (been there). You will be even unhappier if you crash or skid because of a malfunction.

Bob
 
Couldn't agree more on the safety aspect.

Theory is one thing. If i actually tried this I would make a test setup and try and work all the bugs out of the IGBT switching prior to driving the car through the garage wall with a code bug.

If all IGBT gates are pulled low there shouldn't be a load on any of the phase wires and the microcontroller would be confused when trying to calculate the FOC signals correct? when you are stopped and trying to start there is no voltage and it slowly starts applying that. however when you are moving cut all gates then try to brake and the leaf controller has no phase signals I would imagine it would freak out for a second before the phases match up. For sure not something you want to try while running.

Simplest way I can think to test this and may be very destructive would be to find a way to full all the IGBT gates low at will and then allow them to return again.

Start parked, brake on. Pull the gates low and see if there is a fault. If there is, then this wont work and I hopefully didn't break anything.
Next
In a large parking lot. Pull gates low when driving 10mph for a few seconds then reconnect. hopefully no blue smoke and no errors.
Then
In B mode, coast at 10mph disconnect and reconnect. Then same thing while braking.

IF all of those passed I would have some more confidence.

As far as the 85Kw power limit, I don't believe that is true.
on a regular leaf I have seen it peak at 95Kw and after about 25mph it starts to limit the current.
I want to eliminate the current limiting so i can still pull 300A at will above 25mph.
the batteries are technically capable of 900A if there was a way to get it without melting things.
 
Ok maybe 95kW output is possible out of 24kWh pack. But only at higher SOC due to amp limit.
15kW power is not something that should triple the cost of the project.
120kW is definitely not achievable . If Tesla can't do it you will definitely not be able to do it safely ;)
Stick to the OEM solution. Otherwise you will have gigantic plans with very little progress.


There have been numerous attempts to do grand-EV-drivetrain-retrofit-mods - almost all have failed.
 
I would agree that out of all the attempts I have seen, only two have succeeded in using a leaf motor at higher power.
One is this:
https://endless-sphere.com/forums/viewtopic.php?f=30&t=63982&start=100
the other MPaulholmes':
http://mynissanleaf.com/viewtopic.php?f=8&t=22884

Both of course ditch the existing controller.

Still might be fun to try. Maybe I'll get a junked 2012 after new year and try it out.

I guess another route would be to turn down the gain on the current sensors on the phase legs.
The system would think the current on the legs are low and allow higher rates potentially.
Also a lot less potential danger and complexity.
 
I really wish I knew what you guys were talking about.. I so want to convert my VW bus to EV and would love to get the storage and power on it as the Leaf... When I read the above posts.. all my little brain sees is QUACK QUACK QUACK, CHEP CHEP CHEP, woof woof woof, meow.
 
Back
Top