Any driveability quirks/complaints?

Topics concerning the Tesla front and rear drive unit drop-in board
jon volk
Posts: 572
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Any driveability quirks/complaints?

Post by jon volk »

Since more or less completing my quest for max power tuning, I've spent a bunch of time improving low speed driveability of the LDU. Im currently controlling boost, fslipmax, thortramp, idlethrotlim, brakenompedal, and idlemode over CAN. These are all being mapped in relation with rpm, pot1, and pot2 or some combination of 2 parameters in order to get what is approaching a pretty OEM driving experience.

Most of my grievances are pretty minor, but I tend to have a bit of OCD chasing the "perfect" tune from my standalone EFI days. Some examples would be some bucking from a little too much light throttle application at low speed, or some jerkiness of sitting behind slow car getting on and off the brake pedal and idle mode in turn switching on and off, or getting a consistent brake off idle startup as loads or SOC vary.

Id like to hear if anyone else has encountered minor or major annoyances in driveability and if you could do your best to describe in detail and post your parameter json, that would be awesome. I only have a data point of one with my car and if it seems others have a similar experience, I can see about sharing teensy based CAN program to help alleviate some of these.

This is mainly geared toward LDU users, but SDU input is welcome. The base parameters do an admirable job of control, but certain static values can how their limitations when operating over a broad range.
Formerly 92 E30 BMW Cabrio with Tesla power
User avatar
Jack Bauer
Posts: 3563
Joined: Wed Dec 12, 2018 5:24 pm
Location: Ireland
Has thanked: 1 time
Been thanked: 87 times
Contact:

Re: Any driveability quirks/complaints?

Post by Jack Bauer »

I get the shudder and oscillation at low throttle % and low speed. Keep meaning to investigate.
I'm going to need a hacksaw
arber333
Posts: 3241
Joined: Mon Dec 24, 2018 1:37 pm
Location: Slovenia
Has thanked: 74 times
Been thanked: 223 times
Contact:

Re: Any driveability quirks/complaints?

Post by arber333 »

Jack Bauer wrote: Mon Sep 14, 2020 8:15 am I get the shudder and oscillation at low throttle % and low speed. Keep meaning to investigate.
I also get this in my "ordinary" ACIM motor since min slip is quite high and i get some play in 2nd gear. I still think this is P issue since this interface does not use PI or PID cars mass has to dampen the power pulses until they sound off.
User avatar
Boxster EV
Posts: 434
Joined: Fri Jul 26, 2019 9:32 pm
Location: UK
Has thanked: 35 times
Been thanked: 26 times

Re: Any driveability quirks/complaints?

Post by Boxster EV »

The below really is very much about fine tuning - overall the car is performing great:


1) Low speed oscillation of the motor when on/off brake - the issue is linked to running automation idlemode in 'no brake' (and idlespeed in a positive number) . This usually occurs when parking the car carefully into a space where low speed control is important. The best way I could describe the symptoms is that It almost feels as if the motor doesn’t know which way to rotate - it just oscillates until the brake is fully released allowing the motor to turn without resistance. I've learnt to live it.

2) I'm getting vibration when pulling away from a standing start at a moderate speed - especially evident when facing uphill. I've been experimenting with lots of settings over the past 6 months and I've not come across anything that makes a difference to improving this 'quirk'. Once the car is moving the motor is very smooth under any conditions and at any speed.

3)Throttle power delivery. It's reactive throughout the throttle range but could be smoother. I've played around ampmin and slipstart with varying success but it's not perfect.
Porsche 986 powered by a Tesla large drive unit. Backwards. Build documented here and Instagram @tesla_porsche here.
arber333
Posts: 3241
Joined: Mon Dec 24, 2018 1:37 pm
Location: Slovenia
Has thanked: 74 times
Been thanked: 223 times
Contact:

Re: Any driveability quirks/complaints?

Post by arber333 »

Boxster EV wrote: Mon Sep 14, 2020 9:18 am The below really is very much about fine tuning - overall the car is performing great:


1) Low speed oscillation of the motor when on/off brake - the issue is linked to running automation idlemode in 'no brake' (and idlespeed in a positive number) . This usually occurs when parking the car carefully into a space where low speed control is important. The best way I could describe the symptoms is that It almost feels as if the motor doesn’t know which way to rotate - it just oscillates until the brake is fully released allowing the motor to turn without resistance. I've learnt to live it.
This is usually connected to either too little boost or too high fslipmin. Could be that you used too much ampmin to offset lack of boost.

2) I'm getting vibration when pulling away from a standing start at a moderate speed - especially evident when facing uphill. I've been experimenting with lots of settings over the past 6 months and I've not come across anything that makes a difference to improving this 'quirk'. Once the car is moving the motor is very smooth under any conditions and at any speed.
This is also connected to either too little boost or too high fslipmin. This one feels like stabilizing a rubber cord no?

3)Throttle power delivery. It's reactive throughout the throttle range but could be smoother. I've played around ampmin and slipstart with varying success but it's not perfect.
Again try to experiment with more boost. I went as high as possible untill i got a desat event.
That said try to vary boost and fslipmin values with regard to ampmin. Well maybe you should select one fslipmin and stick to it. Easier that way.
BTW motor on the road at speed is not drawing too much power vs other users by any chance? Thst should indicate you have enough fslipmin and ampmin and should only vary ampmin in small ammount.
User avatar
Boxster EV
Posts: 434
Joined: Fri Jul 26, 2019 9:32 pm
Location: UK
Has thanked: 35 times
Been thanked: 26 times

Re: Any driveability quirks/complaints?

Post by Boxster EV »

arber333 wrote: Mon Sep 14, 2020 10:46 am
Boxster EV wrote: Mon Sep 14, 2020 9:18 am The below really is very much about fine tuning - overall the car is performing great:


1) Low speed oscillation of the motor when on/off brake - the issue is linked to running automation idlemode in 'no brake' (and idlespeed in a positive number) . This usually occurs when parking the car carefully into a space where low speed control is important. The best way I could describe the symptoms is that It almost feels as if the motor doesn’t know which way to rotate - it just oscillates until the brake is fully released allowing the motor to turn without resistance. I've learnt to live it.
This is usually connected to either too little boost or too high fslipmin. Could be that you used too much ampmin to offset lack of boost.

2) I'm getting vibration when pulling away from a standing start at a moderate speed - especially evident when facing uphill. I've been experimenting with lots of settings over the past 6 months and I've not come across anything that makes a difference to improving this 'quirk'. Once the car is moving the motor is very smooth under any conditions and at any speed.
This is also connected to either too little boost or too high fslipmin. This one feels like stabilizing a rubber cord no?

3)Throttle power delivery. It's reactive throughout the throttle range but could be smoother. I've played around ampmin and slipstart with varying success but it's not perfect.
Again try to experiment with more boost. I went as high as possible untill i got a desat event.
That said try to vary boost and fslipmin values with regard to ampmin. Well maybe you should select one fslipmin and stick to it. Easier that way.
BTW motor on the road at speed is not drawing too much power vs other users by any chance? Thst should indicate you have enough fslipmin and ampmin and should only vary ampmin in small ammount.

I'm not aware of the motor drawing too much power compared with others but I could be wrong. Incidentally, the range of the car is decent for a 16kwh pack.

My boost is at 1400 which is the highest I can go before regularly hitting OC. :(

Stabilizing a rubber cord is about right with many of these parameters which is why Jon's leaning towards dynamically adjusting them over CAN. :D
Porsche 986 powered by a Tesla large drive unit. Backwards. Build documented here and Instagram @tesla_porsche here.
arber333
Posts: 3241
Joined: Mon Dec 24, 2018 1:37 pm
Location: Slovenia
Has thanked: 74 times
Been thanked: 223 times
Contact:

Re: Any driveability quirks/complaints?

Post by arber333 »

Boxster EV wrote: Mon Sep 14, 2020 11:18 am My boost is at 1400 which is the highest I can go before regularly hitting OC. :(

Stabilizing a rubber cord is about right with many of these parameters which is why Jon's leaning towards dynamically adjusting them over CAN. :D
Hm... 1400 is really small margin. That is like 15V generated on start! Probably you have to help the motor with some more throttle which then quickly gets redundant.
My motor is wound for 180RMS and it needs at least 7000 to move without rubber cord effect. Maybe you should rebalanse this with ampmin and fslipmin parameters.
What is your minimal slip? Mine is 3Hz and you can tell motor likes this better than 1.8Hz or 2.4Hz i tried according to calculation. Also to be honest the larger (diameter) motor the smaller the minimal slip should be... Except if you vary rotor length, than this increases somewhat.

How about increasing ampmin at existing boost? Up to which value you can go without OC being triggered?

Did you look at your deadtime setting? This should be set optimal before you can really use your best boost. I had lot of desat events before i setup delay time to help me with supress false desat reporting on drivers.
Since you use Tesla stuff which is OEM, try to increase this to 160, save and reset. This should be safe to experiment with radical boost settings.

EDIT: Anyone saw Tesla LDU phase transistors scope shot? It would be very helpfull to determine best deadtime.

A
jon volk
Posts: 572
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: Any driveability quirks/complaints?

Post by jon volk »

So the low speed moderate throttle bucking seems to be the most prevalent issue. I believe I have generally solved this one.

You may notice that in my static parameters, I run a fairly high slipstart parameter. This is to reduce bucking by getting some speed built up, but also creates a somewhat soggy pedal low end.

The short version is the motor appears to not like much slip in low to moderate load situations. What I am doing is basically setting fslip max equal to fslipmin at 0 rpm and map that over rpm for say the first 4,000 rpm if pot is below a certain point, then if pot is over x, it will then map fslip max to pot instead. This has resulted in being able to lower slipstart to 35% and not get into any bucking while getting a strong throttle feel over the full range.

Ill get the CAN control code separated from my VCU and draw up a simple pcb for a Teensy3.2 with power supply and CAN transceiver from JLC if there is interest in a plug and play peripheral. Ill post the code regardless.
Formerly 92 E30 BMW Cabrio with Tesla power
jon volk
Posts: 572
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: Any driveability quirks/complaints?

Post by jon volk »

With idle, if the motor is at 0rpm, it sets idle to brakeoff mode. Then upon brake release, it will ramp idlethrot from 8 to 27% until 250 rpm at which point idlemode switches to always on. This ensures a smooth transition to movement even with varying load conditions or SOC. When in this mode, idlethrot is controlled by pot2 with brake pressure between 0 and 27%. This way once the car is in motion, you can very finely modulate the slow forward movement with your foot on the brake pedal. Coming to a complete stop resets the idle mode. It appears the brake off idle does not have any built in throttle ramp so on/off transitions have an jerky oscillating effect. This eliminates that with one exception that a rapid off brake can give a little shudder. I need to build in a ramp rate to my code.

I may be able to do away with the idle mode switching, but the inner OCD likes the inverter effectively being off when stopped.
Formerly 92 E30 BMW Cabrio with Tesla power
User avatar
Jack Bauer
Posts: 3563
Joined: Wed Dec 12, 2018 5:24 pm
Location: Ireland
Has thanked: 1 time
Been thanked: 87 times
Contact:

Re: Any driveability quirks/complaints?

Post by Jack Bauer »

Thanks Jon:)
I'm going to need a hacksaw
jon volk
Posts: 572
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: Any driveability quirks/complaints?

Post by jon volk »

If anyone wants to give where Im at now a try. I pulled this code out of my VCU project and havent tested it, but Visual Studio didnt yell at me when I compiled it. Still work to be done, but feedback from others would be helpful beyond my data point of one.

From memory, the only real difference from my latest parameter set was changing ampmin to 1 and slipstart to 35

Code expects a Teensy3.2 connected to a transceiver on CAN0.

https://github.com/jonvolk/LDU_CAN

From the readme

Code: Select all

Mapping of inverter parameters over CAN

Must use brake pressure input on POT2! Parameter index values based on SINE Firmware 4.90.R!

Openinvertercustom CAN mapping.

rpm,309,0,16,1,TX
pot,275,0,16,1,TX
pot2,275,16,16,1,TX
din_brake,79,8,1,TX

All other parameters are set via CAN SDO standard.
Formerly 92 E30 BMW Cabrio with Tesla power
jon volk
Posts: 572
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: Any driveability quirks/complaints?

Post by jon volk »

You might be able to just comment out the idle throttle function if not using a brake pressure input.
Formerly 92 E30 BMW Cabrio with Tesla power
User avatar
johu
Site Admin
Posts: 5682
Joined: Thu Nov 08, 2018 10:52 pm
Location: Kassel/Germany
Has thanked: 153 times
Been thanked: 960 times
Contact:

Re: Any driveability quirks/complaints?

Post by johu »

This is pretty awesome. I don't understand much of it to be honest but I wonder if it could be integrated into the firmware for people who don't have a VCU?
Support R/D and forum on Patreon: https://patreon.com/openinverter - Subscribe on odysee: https://odysee.com/@openinverter:9
jon volk
Posts: 572
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: Any driveability quirks/complaints?

Post by jon volk »

At the core it's nothing overly complex so I would think it could be added to the firmware, although I'm having a little trouble visualizing how to pare it down to a couple additional parameters or drop downs that make sense at the moment. I am also unsure how well these concepts translate to other motor/inverter setups.

In the future Id like to try running a Volt dual igbt control inverter on a Leaf motor behind a BMW DCT in something pretty light to have another application set, but the budget aint there at the moment.

*just re-uploaded the .ino. I accidentally removed the CAN read from loop.
Formerly 92 E30 BMW Cabrio with Tesla power
User avatar
Boxster EV
Posts: 434
Joined: Fri Jul 26, 2019 9:32 pm
Location: UK
Has thanked: 35 times
Been thanked: 26 times

Re: Any driveability quirks/complaints?

Post by Boxster EV »

I’ll give the code a whirl - thanks. Teensy ordered.
Porsche 986 powered by a Tesla large drive unit. Backwards. Build documented here and Instagram @tesla_porsche here.
jon volk
Posts: 572
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: Any driveability quirks/complaints?

Post by jon volk »

Great, let me know if you run into any issues. If I make any updates, which Im sure I will, they will be pushed to the github file.
Formerly 92 E30 BMW Cabrio with Tesla power
jon volk
Posts: 572
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: Any driveability quirks/complaints?

Post by jon volk »

Made an update in idlemode handling.

In runmode off, idle is set to brakeoff. Upon transition to run, idlemode goes to always on with idlethrot off pot2. (Firmware will not close the contactor if idle is in always on., even with brake pressed, this switching method works fine )

I found the delay in completely removing my foot from the brake to be a little too long. This also starts the motor up with some residual loading by the brakes, which should help smooth things out. Low speed navigation is extremely precise.

One additional CAN TX message is needed, which I added to the readme.
Formerly 92 E30 BMW Cabrio with Tesla power
jon volk
Posts: 572
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: Any driveability quirks/complaints?

Post by jon volk »

Added a parameter JSON to keep a baseline reference for anyone trying the CAN parameters.
Formerly 92 E30 BMW Cabrio with Tesla power
jon volk
Posts: 572
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: Any driveability quirks/complaints?

Post by jon volk »

arber333 wrote: Mon Sep 14, 2020 12:15 pm
Did you look at your deadtime setting? This should be set optimal before you can really use your best boost. I had lot of desat events before i setup delay time to help me with supress false desat reporting on drivers.
Since you use Tesla stuff which is OEM, try to increase this to 160, save and reset. This should be safe to experiment with radical boost settings.

EDIT: Anyone saw Tesla LDU phase transistors scope shot? It would be very helpfull to determine best deadtime.

A

I glossed over this comment about deadtime before. I dont know enough about this to comment, but here is the datasheet for the igbts used in the LDU inverter.

https://www.infineon.com/dgdl/Infineon- ... 2890113e25
Formerly 92 E30 BMW Cabrio with Tesla power
arber333
Posts: 3241
Joined: Mon Dec 24, 2018 1:37 pm
Location: Slovenia
Has thanked: 74 times
Been thanked: 223 times
Contact:

Re: Any driveability quirks/complaints?

Post by arber333 »

jon volk wrote: Mon Sep 21, 2020 12:51 pm I glossed over this comment about deadtime before. I dont know enough about this to comment, but here is the datasheet for the igbts used in the LDU inverter.

https://www.infineon.com/dgdl/Infineon- ... 2890113e25
See this application note:
https://www.infineon.com/dgdl/Infineon- ... aefc41005b
OK i would use 500ns dead time for the transistor then, just to be on the safe side. To that you need to add a delay from drivers. You get that on driver datasheet. Also beware if your drivers use any opto couplers you have to count them in also.
jon volk
Posts: 572
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: Any driveability quirks/complaints?

Post by jon volk »

Happily Tesla uses plenty of well documented bits.
https://www.infineon.com/dgdl/Infineon- ... rId=121859

It looks like that adds around 200ns if I'm vaguely understanding. I'm not sure on the use of optocouplers as it's been a while since I had the cover off to look at the gate driver board. Perhaps Damien can chime in there.
Formerly 92 E30 BMW Cabrio with Tesla power
arber333
Posts: 3241
Joined: Mon Dec 24, 2018 1:37 pm
Location: Slovenia
Has thanked: 74 times
Been thanked: 223 times
Contact:

Re: Any driveability quirks/complaints?

Post by arber333 »

Well I see you need to add about 250ns for driver delay. So (500ns + 250ns)*1.2 = 900ns

Since gate resistors are not the same we leave some ns for safety and I would set 1us for. Which is still very sharp. Remember I use 1us only for blanking time.
I am still not sure how rest of transistors come into play in parallel.
jon volk
Posts: 572
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: Any driveability quirks/complaints?

Post by jon volk »

I think you may be on to something. I really dont fully grasp this level of detail on inverter function so I started looking for technical papers to begin digesting. An excerpt from this one really kind of struck me as relevent.
When an ac motor is supplied by a voltage source power inverter, the stator voltages of the motor will
contain harmonics generated by dead-time effect. The motor operation will be badly influenced by these
harmonics, particularly at no-load and low frequencies operations, making more losses of machine, reducing
efficiency and make torque pulsation of the motor [12], [13].
That seems like a pretty accurate description of what we're perceiving. Per the wiki for comparative figures, if value 23 is 800ns and 62 is 1.5us, then it would appear that we perhaps may be running significantly more dead time than needed.

That being said, I want to do some more research before attempting any changes since it can have severe consequences. If anyone else has wisdom to add, Id love to hear it.
Formerly 92 E30 BMW Cabrio with Tesla power
arber333
Posts: 3241
Joined: Mon Dec 24, 2018 1:37 pm
Location: Slovenia
Has thanked: 74 times
Been thanked: 223 times
Contact:

Re: Any driveability quirks/complaints?

Post by arber333 »

jon volk wrote: Mon Sep 21, 2020 6:29 pm I think you may be on to something. I really dont fully grasp this level of detail on inverter function so I started looking for technical papers to begin digesting. An excerpt from this one really kind of struck me as relevent.
When an ac motor is supplied by a voltage source power inverter, the stator voltages of the motor will
contain harmonics generated by dead-time effect. The motor operation will be badly influenced by these
harmonics, particularly at no-load and low frequencies operations, making more losses of machine, reducing
efficiency and make torque pulsation of the motor [12], [13].
That seems like a pretty accurate description of what we're perceiving. Per the wiki for comparative figures, if value 23 is 800ns and 62 is 1.5us, then it would appear that we perhaps may be running significantly more dead time than needed.

That being said, I want to do some more research before attempting any changes since it can have severe consequences. If anyone else has wisdom to add, Id love to hear it.
Well like i said i was experimenting with boost and deadtime in my car and everytime i tried to took off from stop desat event would happen. This happened because i didnt have enough "blanking time" preset so driver thought that the "miller plateu" spike was an actual critical spike and would stop the process. When i managed to hardcode more than 1us of blanking time i could drive my car normally. BUT for that to stay within normal PWM limits without distortion i had to increase deadtime on Olimex over 160 which is 2.8us!
Also with Volt inverter i use 3us for deadtime which is at 180.
When talking of dead time, you have to know it is just a blank space between signals. By itself it does no damage or causes no interference. The energy sum of your inverter however is somewhat smaller because of that zero. Inverter will feel more sluggish that way.
So i propose you increase dead time to 85 or even to 120 and see what happens. You cant go wrong with MORE deadtime. You can however damage your IGBTs with too short deadtime.
jon volk
Posts: 572
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: Any driveability quirks/complaints?

Post by jon volk »

Sounds good. I’ll throw adding more dead time into the mix since it’s a safe experiment.
Formerly 92 E30 BMW Cabrio with Tesla power
Post Reply