openinverter forum

While the general parameters seem to be well documented on their own, I thought it would be a good conversation to talk specifically about what parameters you should and shouldn’t change on the LDU and how these changes affect actual driveability performance as well as power levels/delivery

I’m a newb so I’ll let the experts start.

great Idea ! . any input from other members ??

Ok I'll get things started.

Firstly, a few DO NOT CHANGE parameters.

POLEPAIRS
NUMIMP
ENCMODE
DIRMODE
PWMFRQ
PWMPOL (change this and kiss your inverter bye bye)
DEADTIME

These are the main do not touch. There are a few others but these are the main ones who can ruin your day in a heatbeat. Please don't ask me how I know.

Now, I know you guys want moar powwa but before we go there a cautionary tale from my younger self :

Good stuff!
Damien, what was your boost value set to when you had the car on the dyno? Aside from breaking driveline, when might the boost value start to have unfavorable consequences from a motor/inverter standpoint?

So first we need to understand the meaning of Boost in the context of the inverter.
All motor stators have a DC resistance. Until that resistance is overcame, no current can flow. It is current in the stator that creates the magnetic fields that allow the motor to tun electricity into mechanical effort.
The throttle input to the inverter commands voltage and slip. Without Boost it may take a large percentage of the throttle travel to generate enough voltage to overcome the DC winding resistance and start the motor motoring. So ideally we tune Boost to give us enough voltage at the beginning of throttle application to get the car moving. As Magnum PI used to say " I know what you're thinking and you're right" More Boost = more starting torque. BUT we need to be careful. At very low RPM the motor is not making much back emf so if we kick lots of volts into the stator we are going to get lots of amps and lots of those lovely magnetic fields that make torque. Spare a thought for the poor power electronics. When we talk about Tesla drive units pulling 1000 amps, 1200 amps etc we are talking about DC battery amps for short acceleration periods. Now the AC current generated by the inverter and fed to the stator can be much higher for much longer periods. This strains the igbts and makes them sad. So moral of the story? Tune boost in small amounts and watch your IL1 and IL2 parameters using high speed graphing.

Great explanation. Interesting to note that I had a couple recurring over current faults from a dead stop when the boost was at the pre-loaded 2000 value. That seems to have gone away after backing that value down to 1500. I plan on doing some datalogging over the weekend to get a better understanding of it all as its new territory for me.

Makes sense as you have a much stronger battery than me.

Good Morning (us est)
It looks like we'll get the #P85DPorsche to the track in the next couple weeks. I will add this for now. So far we've only had one bad experience with inverter settings, encoder mode. We were testing with 40V and vaporized the test leads when I accidentally moved the field because I hadn't clicked off the field before attempting to scroll off the page, didn't notice it had run to the bottom of the enc mode in the process. The leads burnt up as soon as the start button was pushed. I was very glad it hadnt happened under full voltage (403V pack).

I reckon the "bad" parameters should be flagged hidden That will prevent them to show up at all. Don't forget to "save" after flagging

johu wrote: ↑Thu May 09, 2019 7:08 pm I reckon the "bad" parameters should be flagged hidden

Code: Select all

flag encmode hidden

That will prevent them to show up at all. Don't forget to "save" after flagging

Is this active on 4.12 or just on the latest push?

Yes I added it in 4.10.R

So we've had some discussion regarding go paramaters, so Ill sway it toward stopping.

Just to clarify the brake parameters slightly.
brknompedal : Preset level of regen when 12v is applied to pin 2 on the AMP connector.
brkpedalramp: Ramp up time from 0% to brknompedal % setting. From my experience, setting of 1 is more than adequate.

brknom: This one gets a little fuzzy for me. Does this become POT2 max regen % setting when utilizing a 0-5v pressure transducer on AMP pin 14? Is this a completely independent brake signal that will function without 12v on pin 2? Assuming you have both brknompedal and brknom set to max 20, do they combine for a total of 40%, or does the higher value always prevail? Does the firmware recognize POT2 values as brake input for cruise and idle settings?

brkmax: Seems pretty straightforward, pedal off regen %

brkout: Not relevant to the LDU board, correct?

Yes parameter naming could be improved
brknom: percent of throttle travel used to control regen
brkmax: regen when completely off throttle
brkout: brake light output threshold

If potmode=0 AND pot2min/pot2max are setup AND a secondary pot is actually connected this can be used to limit regen to a value below brkmax or brknompedal, respectively

Ahhhhh I think I've got it. So pin 2 is required to enable pedal braking and the POT2 does not provide any discreet trigger, only acts a trim value to brknompedal/brkmax.

I can also confirm brake light out pin is not supported on current rev Tesla logic boards.

OK, so I’ve got a 1500psi transducer all set up for pot2. Min (no pedal) at 750 and max (standing on the pedal) at 4095. Now, will this increase or decrease effective brakenom regen with pedal input. I’d like to set it up that it brakes harder the more the pedal is pressed. I have reduced hydraulic pressure to the rear with a proportioning valve. I’d rather ask the question here than experiment on the road.

Perfect. With pot2min=750, pot2max=4095, potmode=0 you will have no off-throttle regen and regen will start as soon as you touch the brake pedal. If you want some off-throttle regen reduce pot2min towards 0 and if that isn't enough you can also reduce pot2max.

It might be worthwhile to *not* connect the brake light switch so that there is no step between off-throttle and touching the brake pedal. In that case brkmax is the only parameter you need to touch for max regen strength.

You should set the brake ramp low, just like the throttle. Unless you want your rear wheels to lock up on a tesla motor.

tom91 wrote: ↑Fri May 17, 2019 6:39 pm You should set the brake ramp low, just like the throttle. Unless you want your rear wheels to lock up on a tesla motor.

I’d argue even a setting of 1 is too fast for aggressive regen. I almost feel like we would benefit from greater resolution by increasing the time period of the ramp per 100ms instead of 10ms in the settings.

I’m all set up with the pressure sensor and will be testing as soon as the roads dry out.

The pressure sensor works brilliantly. This has been the smoothest and most usable transition from throttle to braking I’ve experienced with this motor.

jon volk wrote: ↑Fri May 17, 2019 6:48 pm
I’d argue even a setting of 1 is too fast for aggressive regen. I almost feel like we would benefit from greater resolution by increasing the time period of the ramp per 100ms instead of 10ms in the settings.

Yes I would second that, or have it take 0.1 increments. A bit more control over the ramp would be nice. You cannot run a Tesla Large DU over ramp of 3-4.

Where do I find a list of the parameter description?
Thanks!
Peter

Have you seen the wiki? https://openinverter.org/wiki/Parameters

Well I have seen it but feel it is very limited...would like a complete list

I bought a fully build and tested Tesla LDU card and am trying to get it up and running.
So far I connected Ignition 12V, ground, start button.
I don't get it to respond to the start signal, I can only activate it manually in the Web interface.
Do I have to add any more cables at this point?

etapp01@hotmail.com wrote: ↑Fri Jun 14, 2019 11:50 am Well I have seen it but feel it is very limited...would like a complete list

I bought a fully build and tested Tesla LDU card and am trying to get it up and running.
So far I connected Ignition 12V, ground, start button.
I don't get it to respond to the start signal, I can only activate it manually in the Web interface.
Do I have to add any more cables at this point?

I would suggest reading the FAQs that Damien posted recently in this thread, as well as ask such questions there. I am trying to keep this thread soley for discussion of how the tuning parameters affect motor operation in the LDU. There is also a Parameter documentation thread elsewhere as well for discussing parameter definitions in general.
viewtopic.php?f=10&t=76