Yet another Lebowski + Honda IMA inverter & step-by-step

[amberwolf]

While I don't know for sure, I would guess that the inverter, as long as it is properly cooled, could do whatever power level it is built for, continuously. The original motor might not handle it continously (usually that is only if it doesn't have proper cooling), but the controller probably could.

 

[Pyclanmot]

В основном со спецификациями Honda IMA....
Из того, что я могу найти (не все включено), инвертор рассчитан на «вспомогательный» двигатель мощностью 10 кВт ... для меня это означает прерывистый режим работы; в то время как на велосипеде MH602 будет работать в непрерывном режиме (10k/30k пик)... ?? IDK... кажется, что IMA недооценен для двигателя Enertrac??

Да I squeezed more from it and everything is fine, the inverter works up to 250В And it is capable of delivering up to 30 kW without problems, but with a nominal value of 20 kW for sure for a long time

 

[marcexec]

Controller is re-installed on the bike after sealing and additional connectors (see 7).
I did an extended test run (emulated commute on Sunday evening, up to 90km/h and 20°C->~40°C in the uphill leg) and deemed it safe to commute for real today. See my build thread as well.

However I'm still struggling with the launch performance. Once it reaches 15 km/h and FOC kicks in it's smooth and seems to deliver full power up to the configured limits.

Post initial configuration with Roger the journey was as follows:

• transition to FOC was rough & unreliable at 1500 eRPM
• dropped phase current from 300 to 150A, re-autocompleted at 200A (for higher error currents) and items further down then just reduce phase for this tweak. Note battery current is limited at 150A due to fuse and contactor ratings.
• consecutitvely raised i\R from 53 to 90 mOhm, which got me going
• I have it now at 175A and 100 mOhm, which is still reliable:

version:2B0_
0x019A    0xE000    0xFFFF    0xFFFF    0x02C8    0x0003    0x4F80    0x000F
0x66B0    0x0000    0x0000    0x0000    0x0000    0x0000    0x017E    0x0000
0x00E9    0x0705    0x01C1    0x0400    0x0400    0x000E    0xFFDA    0x41EC
0xFFE2    0x5517    0xFFFF    0x08F2    0x00BA    0x0396    0x044E    0x0AA4
0x00E9    0x0150    0x0348    0x0026    0x01A4    0xFF8F    0xFFB0    0xE4D8
0xE86C    0x08F3    0x16F7    0x7FBC    0x7D2E    0x0010    0x795E    0x0000
0x4CCC    0x0018    0x0000    0x01E0    0xFFFF    0xB334    0xFFE8    0x0000
0xFE20    0x0000    0x4CCD    0x0018    0x0000    0x01E0    0xFFFF    0xB333
0xFFE8    0x0000    0xFE20    0x05B8    0x02C4    0x0064    0x0017    0x36CF
0xFF00    0x3A37    0x9033    0x6A37    0xEA37    0x1033    0xBA39    0xFF00
0x1122    0x0026    0x000B    0x0005    0x00C2    0x0000    0x0000    0x051E
0x028F    0x0400    0x0400    0x0400    0x085D    0x0357    0x0093    0x042D
0x0008    0x0D1A    0xFFC0    0x08F2    0x1000    0x0000    0x0000    0x0000
0x0000    0x0000    0x00BB    0x02AF    0xFFFF    0xFFFF    0x147B    0x0106
0x118E    0x05DA    0x0578    0x0258    0x0064    0x0004    0x0004    0x000E
0x0298    0x0003    0x0000    0x0078    0x0000    0x0000    0xFFFD    0x0000
0xFF88    0x0000    0x0000    0x0005    0x0000    0x0064    0xFFFB    0x0000
0xFF9C    0x000C    0x0000    0x00F0    0xFFF4    0x0000    0xFF10    0x0619
0x01D3    0x08C7    0x004D    0x014E    0x01D3    0x002E    0x0010    0x0E10
0x07D0    0x03B6    0x6000    0x0624    0x0074    0xBC10    0xA9A6    0x0803
0x8600    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000
0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000
0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000
0x0000    0x0000    0x0000    0x0000    0x0000    0x00A0    0xFFFF    0xFFFF
0xFFFF    0xFFFF    0xFFFF    0xFFFF    0xFFFF    0x0052    0xFFFF    0xFFFF
0xFFFF    0xFFFF    0xFFFF    0xFFFF    0xFFFF    0x0001    0x0000    0x0005
0xA000    0x3CA8    0x0038    0x0000    0x0000    0x0000    0x0000    0x0000
0x0000    0x0000    0x0000    0x0000    0xFFDA    0x41EC    0xFFE2    0x5517
0x6C1C    0x010E    0x2800    0x098A    0x0000    0x0CA8    0x005F    0x003E
0x0A80    0x1062    0x0285    0x0206    0x0000    0xFFF8    0xFFDA    0x41EC
0xFFE2    0x5517    0xDB91    0x0005    0x8000    0x059D    0x0000    0x0000
0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000
0xFFDA    0x41EC    0xFFE2    0x5517    0x059D    0x0000    0x0000    0x0000
0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000
0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000
0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000
0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000    0x0000
0x0000    0x0000    0x0000    0x0000    0x0000    0x0324    0x0648    0x096A
0x0C8C    0x0FAB    0x12C8    0x15E2    0x18F9    0x1C0B    0x1F1A    0x2223
0x2528    0x2826    0x2B1F    0x2E11    0x30FB    0x33DF    0x36BA    0x398C
0x3C56    0x3F17    0x41CE    0x447A    0x471C    0x49B4    0x4C3F    0x4EBF
0x5133    0x539B    0x55F5    0x5842    *

Since then I tried to tweak the transition:

1. D\f (erpm dr2 jump to dr3) from 1000 to 750, which is too low even on N\h (minimum # of cycles going from drive 2 to 3 = 5000), 900 is OK even on 2000 cycles.
2. 190A phase is too much for a neat transition even at D\f of 5000 (still based on 200A)
3. G\g (phase control loop, drive 2, 3rd order) autocompletes to 0.03 but as per Roger's recommendation is 0.3, tried doubling it to .6 to see if I get better acceleration under Drive_2 (Halls), but did not notice a difference, so reverted
4. G\n (immediate motor phase step) autocompletes to 64 and was at 16 per Roger's recommendation (for hub motors), tried 8 (more "jumpy") and now 24, which causes a bit of humming but feels more responsive. This value seems to affect both Drive_2 and Drive_3 (FOC)

TBH, I don't have a full understanding of the parameters but I can change a value at a time. I'm especially wary of increasing the error currents (I only autocompleted based on 200A instead of Roger's recommendation of 300A) until I'm a little more confident.
I might try calibration based on 250A next and a phase current of 190-200A.
Edit: tried the 250A, autocomplete then 190A phase current settings this morning. Worked OK, but the transition has become more "rough" and it conked out once (only) on an incline.

Further tuning of my setup starts on this post in my build thread

 

[mk77]

Да I squeezed more from it and everything is fine, the inverter works up to 250В And it is capable of delivering up to 30 kW without problems, but with a nominal value of 20 kW for sure for a long time

you sure about that? 30kW seems like a lot for that little heat sink.... (i.e. and NO fan!)
Do you know what FET's or BJT's are being used??
I would be interested in the Rsdon value.

 

[amberwolf]

They're IGBTs; I forget what p/n. Might be data in one of the links in my version of this thread.

 

[mk77]

.... Worked OK, but the transition has become more "rough" and it conked out once (only) on an incline.

what does "it conked out" mean???
what is "it" ? what happened?

 

[mk77]

Intro:

...
Note that the board is designed to be used with a resolver, mine was "downgraded" by Roger for my motor.

I also use an Enertrac MHM602 Hub motor.
Am considering the IMA/Lebowski route.... but what does this "downgrade" entail? Tks.

 

[kiwifiat]

I also use an Enertrac MHM602 Hub motor.
Am considering the IMA/Lebowski route.... but what does this "downgrade" entail? Tks.

The board was designed for driving a Hyundai HSG IPM motor that has a resolver not HALL sensors. It is a straight forward job to remove the resolver to digital components to suit HALL sensors which are the native Lebowski supported rotor position sensors.

 

[mk77]

@marcexec: sorry for all the questions --> I don't mean to hijack your thread....

@kiwifiat: so ... If I eventually decide to go with an IMA/Lebowski .... is this "modification" something I would be required to perform (( I have the skills to do a mod but would need all the pertinent info )) or can the board be delivered with the mod already installed? THANKS!

 

[Pyclanmot]

Ct300DJ0060

you sure about that? 30kW seems like a lot for that little heat sink.... (i.e. and NO fan!)
Do you know what FET's or BJT's are being used??
I would be interested in the Rsdon value.

 

[mk77]

Ct300DJ0060

please, you have a datasheet for that pn ???
All my favorite parts suppliers return "no data".... it's like it doesn't exist ... could be a "HONDA specfic" part number too....?

 

[amberwolf]

A google search on it turns up some info

Electric Power Module meets needs of hybrid, electric vehicles.

Electric Power Module meets needs of hybrid, electric vehicles. - Jun 06, 2011 - Powerex Inc.

news.thomasnet.com

MITSUBISHI ELECTRIC News Releases Mitsubishi Electric to Release Sample J-Series T-PM Extra Compact Type

Mitsubishi Electric Corporation announced today it has developed a new J-Series transfer molded power semiconductor module (T-PM) mainly for motor drive applications in electric and hybrid vehicles. Samples of the J-Series T-PM extra compact type will be released starting February 19.

www.mitsubishielectric.com

 

[Pyclanmot]

пожалуйста, у вас есть даташит на этот pn???
Все мои любимые поставщики запчастей возвращают "нет данных".... как будто их не существует ... может быть и номер детали "HONDA specfic"....?

 

[mk77]

@Pyclanmot you tease so well...

 

[marcexec]

what does "it conked out" mean???
what is "it" ? what happened?

Better experienced than described
It's basically when the controller detects overcurrent or another fault condition and falls back to Drive_1 / recovery trying to re-sync.
You abruptly lose power, often with a "clunk" sound/feeling. "It" is the motor controller setup, especially the configured parameters here.
I have braided sleeves coming that might clear up the sensors' signals, see my edits above about shielding and being able to change parameters easily. It's a slow learning curve plus you should be prepared to test methodically. Some sort of dyno setup would be ideal...
Edit: that being said it transforms the bike massively - the frustration comes with a decent reward, just delayed. Have a look at Eric's videos starting from this one:

 

[YTW200]

Насколько я могу найти (не все включено), инвертор рассчитан на «вспомогательный» двигатель мощностью 10 кВт…

с помощью мозга OpenInvertor я получаю от этого инвертора мощность 45+ кВт на аккумуляторе 250 вольт в своей машине массой 1500 кг. более того, я проверил этот инвертор от 500+ вольт и он работает нормально!

 

[YTW200]

Знаете ли вы, какие поляковые или биполярные транзисторы используются?

Igbt 600v 300a.

 

[YTW200]

 

[marcexec]

@marcexec: sorry for all the questions --> I don't mean to hijack your thread....

@kiwifiat: so ... If I eventually decide to go with an IMA/Lebowski .... is this "modification" something I would be required to perform (( I have the skills to do a mod but would need all the pertinent info )) or can the board be delivered with the mod already installed? THANKS!

Based on a couple of weeks riding, I's say it's a very good option if you're happy to do some adjustments. Shield everything and bring the power up bit by bit. The responsiveness and smooth running is unreal compared to the Kelly I had before. I haven't even tried field weakening and regen yet...

 

[amberwolf]

Sorry to derail your thread further, but:

The board was designed for driving a Hyundai HSG IPM motor that has a resolver not HALL sensors. It is a straight forward job to remove the resolver to digital components to suit HALL sensors which are the native Lebowski supported rotor position sensors.

I got lost following the chain of posts for this part....

Do you mean it is possible to use a non-UVW/ABC-hallsensored motor with the Lebowksi, by adapting the sensor system the motor has (assuming the motor itself is still UVW 3-phase)? If so, does this include being able to adapt a SIN/COS position sensor?

(I feel like I may have asked this or something similar somewhere else, but I couldn't find it; maybe I just thought about it?)

I ask because I still have this powerchair motor:

Powerchair Direct Drive Brushless Motor (High-Torque)

A friend dropped by tonight with an interesting motor. It is from a powerchair, and the wheel bolts directly to it. Normally it would run at less than 200RPM, on 24V SLA, for a few MPH on a small-diameter wheel, but the manufacturers have apparently tested it up to 65MPH on the same size...

endless-sphere.com

that I'd love to drive with this system (once I finish building mine ); if it works I would get a second one (and possibly a third) to use for a future version of the SB Cruiser trike, since they are easy to mount on the trike frame, and I can make "normal" wheels to bolt to them (or even use as a kind of middrive) rather than having the motor as part of the wheel.

 

[marcexec]

Sorry to derail your thread further, but:


I got lost following the chain of posts for this part....

Do you mean it is possible to use a non-UVW/ABC-hallsensored motor with the Lebowksi, by adapting the sensor system the motor has (assuming the motor itself is still UVW 3-phase)? If so, does this include being able to adapt a SIN/COS position sensor?

(I feel like I may have asked this or something similar somewhere else, but I couldn't find it; maybe I just thought about it?)

I ask because I still have this powerchair motor:

Powerchair Direct Drive Brushless Motor (High-Torque)

A friend dropped by tonight with an interesting motor. It is from a powerchair, and the wheel bolts directly to it. Normally it would run at less than 200RPM, on 24V SLA, for a few MPH on a small-diameter wheel, but the manufacturers have apparently tested it up to 65MPH on the same size...

endless-sphere.com

View attachment 341972
View attachment 341973
that I'd love to drive with this system (once I finish building mine ); if it works I would get a second one (and possibly a third) to use for a future version of the SB Cruiser trike, since they are easy to mount on the trike frame, and I can make "normal" wheels to bolt to them (or even use as a kind of middrive) rather than having the motor as part of the wheel.

@kiwifiat designed the board I have, but I'm pretty sure it can take a resolver input as @eric1565 uses a Hyundai/Kia HSG starter generator: Hyundai/Kia HSG - openinverter.org wiki. The circuit is just jumpered out for me as I only have (misaligned) Halls.

 

[kiwifiat]

Sorry to derail your thread further, but:


I got lost following the chain of posts for this part....

Do you mean it is possible to use a non-UVW/ABC-hallsensored motor with the Lebowksi, by adapting the sensor system the motor has (assuming the motor itself is still UVW 3-phase)? If so, does this include being able to adapt a SIN/COS position sensor?

The IMA board has a resolver to HALL converter that is designed to support the type of resolvers found on automotive ev motors like the Hyundai HSG, Nissan Leaf, Fiat 500E/SMART ForTwo E drive that look like this:


The type of resolver your motor has looks more like a RLS type that won't work out of the box but could be made to work with a change to the firmware. The easiest option would be to change the encoder on your motor to the type that outputs HALL type signals that would work with the Lebowski controller.

 

[YTW200]

На плате IMA имеется преобразователь резольвера в HALL, предназначенный для поддержки типов резольверов, используемых в автомобильных электродвигателях, таких как Hyundai HSG, Nissan Leaf, Fiat 500E/SMART ForTwo E, которые выглядят следующим образом:
View attachment 342571

Тип резольвера вашего двигателя больше похож на тип RLS, который не будет работать «из коробки», но его можно заставить работать с изменением прошивки. Самый простой вариант — заменить энкодер на вашем двигателе на тип, который выводит сигналы типа ХОЛЛА, которые будут работать с контроллером Лебовски.

Я сделал эти универсальные платы на одном чипе. достаточно одного вращающегося магнита, и эта плата выдает от 1 до 7 пар полюсов сигналов Холла или сигнал A/B/Z. очень легко монтируется на различные двигатели с резольвером.

 

[beggernator]

The IMA board has a resolver to HALL converter that is designed to support the type of resolvers found on automotive ev motors like the Hyundai HSG, Nissan Leaf, Fiat 500E/SMART ForTwo E drive that look like this:
View attachment 342571

The type of resolver your motor has looks more like a RLS type that won't work out of the box but could be made to work with a change to the firmware. The easiest option would be to change the encoder on your motor to the type that outputs HALL type signals that would work with the Lebowski controller.

Hi.
Firstly, awesome thread, thanks for making the steps so clear.

I have a Honda ima inverter identical to this and am planning on running a hyundai HSG I have with a 20s battery initially. upping to a 48s battery once operating well. (essentially the same as erics build.)

@kiwifiat Sorry if this a question with an obvious answer.
In this thread I see a newer version of the Lebowski board you designed that looks to be designed to drop directly onto the standoffs of the IMA inverter.
looking through other threads the best I have been able to find is the small Bobc board style. Not the one like is used in this thread.

As you mentioned the HSG has a resolver exactly like pictured. Ideally I would like to use this as directly as possible but I could do similar to eric with a resolver swap if needed.

Basically I am looking for where best to find more information, i had a look on your YouTube channel and your post history but I failed to track down what I was after.

Thanks
Tom.

 

[mxlemming]

9) FAQ, tuning tips, snippets and educational face palms lessons learned

Input welcome!

FAQ

• f_sample as mention in threads and documentation is h) loop sample frequency: in menu A, and should be at least 24kHz for good operation.
• Drive modes are as follows:
  • Drive_0 - startup/enabled
  • Drive_1 - Recovery (after errors)
  • Drive_2 - Sensored (e.g. Halls)
  • Drive_3 - FOC operation

Tuning tips

• hall sensor calibration is done without load, but the subsequent tuning best done under load (i.e. driving). To save the collected data (Hall sensor stats as displayed in L\#) you must enter Setup mode after a (test) run! The data is lost if you just turn the controller off. Thus the recommendation for a button mounted on the handlebar/dash.
• if the jump from Drive_2 (Hall) to Drive_3 (FOC) fails or is rough, the first step is to increase the transition eRPM (D\f) temporarily to give you a bigger window to tune Halls first. The transition can then be smoothed further by giving the controller a bigger "window" through an increased (H\b) (dr2: speed filter 50% step response time [ms]). I went from 50 (Hall default) to 225, note 200 is the default under sensorless start.
• "Conking out" (i.e. fallback to Drive_1 due to e.g. an overcurrent error) in Drive_2 after increasing phase amps can be partially addressed by autocompleting on higher phase amps, which calculates a higher B\G value and then dropping down again.
  This reduces your margins of error! See my build tread as well.

This is quite interesting. Over in VESC world there are endemic problems with hall sensor to sensorless transition as well. In my code I experienced this and tried to sidestep the issue by using the sensors to preload the sensorless... It's very interesting to see Lebowski has a very similar sounding issue with transition.

Makes me think it's some fundamental property of motors rather than an actual controller issue and that the problem is just masked rather than fixed by "tuning". Possibly at the expense of performance in each region.

I wonder if @Lebowski has any insight into this.