openinverter forum

Hi Johannes. I tested the inverter today, the purpose of the experiment was to understand what kind of torque I can get when starte. I compared it with another factory inverter. On it, with the same phase current, the torque is obtained higher. It turns out that in the factory inverter, when the position of the gas pedal increases, the frequency of sliding increases. At the initial position, it supplies the engine with a frequency of 4 Hz, as the pedal increases, the sliding increases to 12 Hz. The sliding frequency can be adjusted. In our inverter, the sliding frequency does not increase with an increase in the position of the gas pedal, and we lose torque at the same time, besides, we can supply a maximum of 10 Hz, this is not enough. Can we add this algorithm to the firmware?

Same for me: torque in the low rpm is less then on the stock inverter software by the manufacturer.

Two different topics here. For asynchronous motors slip should also increase once you increase potnom past slipstart. Of course limited to the 10 Hz slip limit which can be changed.

Starting torque with synchronous motors is probably down to a mismatch in the MTPA curve.

johu wrote: ↑Sat Oct 30, 2021 3:11 pm Two different topics here. For asynchronous motors slip should also increase once you increase potnom past slipstart. Of course limited to the 10 Hz slip limit which can be changed.

Starting torque with synchronous motors is probably down to a mismatch in the MTPA curve.

For some reason, I do not increase more than the slipmin frequency is set, or it increases slightly. In any case, I can't get the value that is in slipmax. Perhaps some other parameter is affected? Or an error in the firmware algorithms?

I tested two inverters on an engine that is jammed with a dynamometer. When moving of course the sliding increases

johu wrote: ↑Sat Oct 30, 2021 3:11 pm Starting torque with synchronous motors is probably down to a mismatch in the MTPA curve.

Do you know what it would take to address this? Is it something we could club together and sponsor the development of?
I've tried making sense of https://in.mathworks.com/help/mcb/ref/m ... rence.html but not much does. Is this down to a difference between the physical motor characteristics and the hardcoded values for fluxLinkage etc? or the implementation of the algorithm?

I'd contribute to that project - though you probably want my money more than my code!

I'm afraid money won't help but if someone wants to come by and swap the shaft seal of my Prius transmission and integrate the Outlander charger into Touran and test my new BMS, then yes, I'd have the time to give this the attention it needs. Too many projects!

That aside people have started playing with the MTPA constants so it might not even have to be me who fixes this

I hear you. Does look like a few people are engaging with this now though so may get addressed fairly quickly. Be interesting to see what impact it can have.