daihatsu terios

[chentron]

This project started in 2016 encouraged by finding out that the electronics needed for the inverter are being open sourced by johannes.

Objectives: - Recycle as much as possible. low cost without sacrificing security.
-homemade
-open source
- daily use, 4x4, light weight

Initially, the chosen model was a Suzuki Vitara, but after researching the used car market, I discovered the Daihatsu Terios, a lightweight, more modern in its aesthetics than the Suzuki, with ABS and an airbag, which for me is the minimum essential in terms of safety without the excess of electronics in actual models.

In 2017 I took the step forward and started by buying a v2 kit from johannes, the AZD AC24LS engine bought from an American user on the DIY-EV forum. Assembling the kit and starting to learn little by little was stopped by the arrival of a baby in my life and it was not until 2021 that I resumed the project.
spinning the engine was an incentive that led me to buy the rest of the parts for the project, inspired by the experiences in this forum:
- MES-DEA 70/6E2 vacuum pump for brakes and differential system.
- TRW pump for the power steering of an opel astra series G
- CATL NMC lithium pouch type battery
- BMS ZEVA open source edition

Let's see some photos.

[chentron]

as you can see in the photos, finishing is not professional, all aluminium for cases and frames are made just with a hammer, from old windows, HV cable is RVK cable with aluminium and orange tape wrapping for EMC.

Now i am tuning the motor. I can just start and run some meters in 1st or 2 gear but i am not able to rev more than 500rpm. I can rev perfectly with no gear selected. I followed the guide and formulas from the wiki/parameters but not success what to change.

[chentron]

help is welcome to tune this motor. now absolutely lack of power
motor is ACIM 1700rpm , 60hz, 69v, 2 polepairs, battery is 335v DC

"boost": 2000,
"fweak": 206,
"fweakstrt": 400,
"fconst": 133,
"udcnom": 0,
"fslipmin": 3,
"fslipmax": 6.59,
"fslipconstmax": 10,
"polepairs": 2,
"respolepairs": 1,
"sincosofs": 2048,
"encmode": 1,
"fmax": 200,
"numimp": 64,
"dirchrpm": 100,
"dirmode": 1,
"snsm": 15,
"pwmfrq": 1,
"pwmpol": 0,
"deadtime": 63,
"ocurlim": 180,
"il1gain": 3.9,
"il2gain": 3.9,
"udcgain": 7.21,
"udcofs": 0,
"udclim": 399,
"snshs": 0,
"potmin": 980,
"potmax": 4095,
"pot2min": 0,
"pot2max": 4095,
"potmode": 0,
"throtramp": 100,
"throtramprpm": 20000,
"ampmin": 10,
"slipstart": 50,
"sinecurve": 0,
"throtfilter": 4,
"brakeregen": 0,
"regenramp": 100,
"regentravel": 30,
"offthrotregen": -30,
"cruiseregen": -30,
"regenrampstr": 10,
"brklightout": -50,
"bmslimhigh": 50,
"bmslimlow": -1,
"udcmin": 285,
"udcmax": 450,
"idcmax": 5000,
"idcmin": -5000,
"idckp": 2,
"idcflt": 9,
"tmphsmax": 85,
"tmpmmax": 80,
"throtmax": 100,
"throtmin": -100,
"iacmax": 5000,
"ifltrise": 10,
"ifltfall": 3,
"chargemode": 3,
"chargecur": 5,
"chargekp": 80,
"chargeki": 10,
"chargeflt": 8,
"chargepwmin": 0,
"chargepwmax": 90,
"idlespeed": -100,
"idlethrotlim": 50,
"idlemode": 3,
"holdkp": -0.25,
"speedkp": 0.25,
"speedflt": 5,
"cruisemode": 0,
"udcsw": 268,
"udcswbuck": 540,
"tripmode": 2,
"bootprec": 0,
"pwmfunc": 2,
"pwmgain": 100,
"pwmofs": 0,
"canspeed": 1,
"canperiod": 0,
"nodeid": 1,
"manualstart": 0,
"fslipspnt": 0.18,
"ampnom": 0.62
}

[johu]

Are all 3 phases working?

Otherwise parameters don't look too suspicious. Boost seems a bit low. And you can try lowering fweak slightly.

[arber333]

help is welcome to tune this motor. now absolutely lack of power
motor is ACIM 1700rpm , 60hz, 69v, 2 polepairs, battery is 335v DC
....
}

Normally with ACIM i used rather more boos that is required. Later on you can taper it off to improve effectiveness.
I use boost = 7000 for my Mazda. I suggest you try by 5000 first. You can calculate how much voltage your motor needs to overcome internal resistance.
https://openinverter.org/wiki/Parameters
However beware! Too much boost can cause desat event and if there is no IGBT protection in drivers...toast!

I see you use "fweakstrt": 400 Your motor may think there is not enough voltage. Try to decrease this to 220Hz.
There is a thing... Did you calculate your minimal slip? For small motors this can be quite high. Try increasing your fslipmin by steps of 0.2 and see if there is some change.

Can you post your motor plate? I may get more ideas from hard data...

[chentron]

Are all 3 phases working?

Otherwise parameters don't look too suspicious. Boost seems a bit low. And you can try lowering fweak slightly.

I test 3 cables AC with RMS multimeter and 3 are consuming about the same amperes(12.8 ) when throtling without gear 1800 rpm, so I asume are ok.

Normally with ACIM i used rather more boos that is required. Later on you can taper it off to improve effectiveness.
I use boost = 7000 for my Mazda. I suggest you try by 5000 first. You can calculate how much voltage your motor needs to overcome internal resistance.
https://openinverter.org/wiki/Parameters
However beware! Too much boost can cause desat event and if there is no IGBT protection in drivers...toast!

I see you use "fweakstrt": 400 Your motor may think there is not enough voltage. Try to decrease this to 220Hz.
There is a thing... Did you calculate your minimal slip? For small motors this can be quite high. Try increasing your fslipmin by steps of 0.2 and see if there is some change.

Can you post your motor plate? I may get more ideas from hard data...

I calculate with the formula of johannes, 3,33

I attached the photo of my motor plate. My motor is wired in wye, so 69v , parameters photos I got from DiY-EV forum, I dont understand the 90% of them, and not sure the reliability of the source. (https://www.diyelectriccar.com/threads/ ... 66/page-11 )

[chentron]

specs of the motor

[arber333]

Uh oh!

No wonder you get sluggish response. Motor is wired for 69Vrms. This means that best suited inverter voltage would be about 144Vdc and lots of amps.
In your case you have 320Vdc you need to extend your FW curve and as such you drag the torque response down.
I see your OEM settings to weaken the field are at 90Hz. That is native curve of the motor. You have fweak set at 206Hz so it should be very sluggish.
I had the same thing with my 28kW motor at 75Vrms and battery of 320Vdc. It was drivable but just barely. If i went back with fweak it would trip OC limit.
For permanenet solution I would suggest rewinding the motor to 140Vrms which is about double. You will get much better response with higher DC voltage and much less current too. Probably too much work... but i see motor should spin to 11000rpm so it should be ok as it is.

As it is you can reduce fweak to 160 or 120 and observe the response. Sooner or later you will arrive at OC limit. Then you can play with boost to mitigate start response somewhat.

[chentron]

thank you arber, i am sure this motor is not for speedtracking , but other users reported in the past be enough good in high voltages like mine (they was using the original inverter mate DMOC445). So i think I might get similar results once the tuning is made correctly.
I tried with fweak 160, first I feel is i can take off on flat with 3 and 4 gear (before was imposible). When in a hill still imposible to take off.
I tied with fweak on 120 but ocurlim happen inmediatly.
In that moment start to rain and must stop because engine bay is not covered. So now have in mind change ocurlim to 250, as i understand this parameter is the maximum motor phase current and the specs from AZD said 250 is the maximum (before it was to 180). Also I have 2 cells under 3,3v and i am charging to 3,65v.

[chentron]

After a few days of testing I managed to get the engine to have acceptable performance with fweak 120, boost 3000, ocurlim 250.
I was able to start in third gear, which at 6000 rpm would give me a speed of 113 km/h.
I just needed to put a new external thermostat so that a new fan would allow me to go out on the open road to test it, but...
I disassembled the inverter box, put in the thermostat and fan (external) and organized the internal cables of the inverter box and now I am not able to turn the motor in the slightest. When accelerating the engine it tries to start turning, but it makes one attempt/hit and it doesn't quite start turning.
I don't have an oscilloscope but I have measured and I have one phase that consumes twice as much as the other two. Any idea what could be happening to me?

[johu]

Classic - looks pretty now but no longer works

Sounds like one of the encoder cables came loose.

Or are you saying one gate driver pulls twice as much as the others?

[chentron]

not the gate driver.
I am measuring the cables that go from the inverter to the motor, with a multimeter with trueRMS, and twice as much current passes through one of them as through the other two.

Classic - looks pretty now but no longer works

Sounds like one of the encoder cables came loose.

Or are you saying one gate driver pulls twice as much as the others?

[chentron]

After more time than desired I have made some progress and with the help of a electronic technician We have found that the problem is in the gate driver board, the DC converter was KO, after changing it I also found that the Si8261 gate is also KO.
Once changed, the motor rotates correctly as before, but now the problem I have is that if the motor is in forward motion and car suffers a small oscillation backwards, such as when setting up on a slope, the Si8261 chip immediately is KO. and here I am again blocked on how to solve this. 3 chips burn at this moment !!
Same happens if car is trying to go rear in a slope and go a bit to the front.

[arber333]

...
Si8261 gate is also KO.
Once changed, the motor rotates correctly as before, but now the problem I have is that if the motor is in forward motion and car suffers a small oscillation backwards, such as when setting up on a slope, the Si8261 chip immediately is KO. and here I am again blocked on how to solve this. 3 chips burn at this moment !!
Same happens if car is trying to go rear in a slope and go a bit to the front.

Ya! I am telling from a long time ago, use desat PROTECTED drivers for your inverter. Or use OEM power section. You may have half dead igbt and it shows only when it has to push more current.

Or you use too small deadtime. Top and bottom igbt signals may have crashed which would seem like shaking to you. I would suggest you set deadtime to 3ms (default 0.8) and then observe adjacent igbt signals with diff probe to set it to best value.