Honda ima and his abilities An old beetle

[Romale]

Thank you
I thought something went wrong because of the noises.

no, in this regard (stability), the ima inverter is cooler than the sheet inverter! no matter how much I mocked, I could not drive it into an error even once (for example, an overcurrent), an inverter from a Nissan can sometimes allow this. IMA never. always super stable.

[Jacobsmess]

250 volts (60s li-ion battery). 40+ kW of power! the car is 1500 kg! 130 km / h.
as a motor, a generator with a Nissan Serena. as a ima inverter.

Is this a Nissan Serena motor and IMA inverter?

[Romale]

Is this a Nissan Serena motor and IMA inverter?

Yes.

[marcexec]

250 volts (60s li-ion battery). 40+ kW of power! the car is 1500 kg! 130 km / h.
as a motor, a generator with a Nissan Serena. as a ima inverter.

Impressive, that gives me a good bit of headroom with the Lebowski! - currently 155A battery (contactor limited) and 230A phase at 140V:
https://endless-sphere.com/sphere/threa ... st-1777461

[Romale]

Impressive, that gives me a good bit of headroom with the Lebowski! - currently 155A battery (contactor limited) and 230A phase at 140V:
https://endless-sphere.com/sphere/threa ... st-1777461

I tested the ima inverter at a voltage of 530 volts!! he survived it perfectly. so with sufficient cooling, it will be a real little miracle

[marcexec]

I tested the ima inverter at a voltage of 530 volts!! he survived it perfectly. so with sufficient cooling, it will be a real little miracle

Crazy, thanks for taking one for the team!
What's the max temperature at the heatsink you had it at? I'm keeping it below 50℃ (drilled hole at a corner for NTC) and am not actively cooling it yet.

[arber333]

I tested the ima inverter at a voltage of 530 volts!! he survived it perfectly. so with sufficient cooling, it will be a real little miracle

Hm... i was seeing a lot of EMI while using HSG motor with IMA and Lebowski over 220Vdc.
I would love to use it with 320Vdc where HSG is made for. But current sensors wouldnt cooperate over 200Vdc and there was a lot of rocking at low rpm. Can you explain how did you solve this? Did you use specific frequency with IMA inverter? I was using 6kHz for lots of EMI.

[Romale]

Hm... i was seeing a lot of EMI while using HSG inverter with IMA and Lebowski over 220Vdc.
I would love to use it with 320Vdc where HSG is made for. But current sensors wouldnt cooperate over 200Vdc and there was a lot of rocking at low rpm. Can you explain how did you solve this? Did you use specific frequency with IMA inverter? I was using 6kHz for lots of EMI.

I have not encountered these problems. I just use the v3 O.I. board in the ima inverter, the current sensors and the driver power board are regular, the ima processor did not solder off and did not close any legs to the mass. from the ima, I take two wires with current sensor signals relative to the total minus of the 12 volt power supply. +5 volts current sensors are received from the standard board. also, relative to the general minus, I supply 6 pwm wires to control the transistors.
and everything.. it works by itself.

Previously, I tried to use my own isolated 6 pcs dc/dc (12/15) to power the drivers, and I took +5 volts for current sensors from the OI board, then I also got a shutdown of the inverter at a voltage above 78 volts.

[Romale]

8 kHz, 160 deadtime.
do you remember that in all circuits for the ima inverter, the upper lower control points are confused in places?
I showed it somewhere. the TOP of the dome is closer to the transformer. the input BOT is closer to the current sensors. and in the existing pdf files with diagrams, the opposite is indicated.

[arber333]

8 kHz, 160 deadtime.
do you remember that in all circuits for the ima inverter, the upper lower control points are confused in places?
I showed it somewhere. the TOP of the dome is closer to the transformer. the input BOT is closer to the current sensors. and in the existing pdf files with diagrams, the opposite is indicated.

Hah! That figures!
I used Lebowski in sensorless and i guess it is way more sensitive. thenk you for that info.
I will test this with HSG and 360Vdc sometime when i assemble Lebowski board again.

[marcexec]

Hah! That figures!
I used Lebowski in sensorless and i guess it is way more sensitive. thenk you for that info.
I will test this with HSG and 360Vdc sometime when i assemble Lebowski board again.

+1 on this - I have highlighted that in my how-to: https://endless-sphere.com/sphere/threa ... st-1761565

[Leo M]

В нашей версии это выглядит так.Конечно хорошо бы , чтоб тема работала еще и с датчиками холла. Весьма хороший вариант для байков с китайскими моторами типа QS. Ну и универсальность применения опять же. Многие просили Еханеса сотворить схему под датчики холла, но он к сожалению считает это нерентабельным (((

[Romale]

В нашей версии это выглядит так.

Could you add a diagram of the part that is responsible for measuring high voltage? And what are its voltage limits?
Thanks

[Leo M]

пишите нам в телегу решим вопрос. по напряжению.. да любые пределы, в зависимости от нужды. в данном варианте до 300в.

[Romale]

пишите нам в телегу решим вопрос. по напряжению.. да любые пределы, в зависимости от нужды. в данном варианте до 300в.

what prevents you from drawing this diagram here? she would be useful to many.

[arber333]

пишите нам в телегу решим вопрос. по напряжению.. да любые пределы, в зависимости от нужды. в данном варианте до 300в.

Have you considered RLS converter. A single polarized magnet mounted on axle with sensor plate behind it. They make several flavours; simulated UVW, resolver, ABZ...
https://www.rls.si/eng/rmb28-rmf44-rmf58

I use this same sensor with Lebowski controler and output is UVW commutation signals. Works really well in my 1.7T Peugeot 406Coupe.

[Leo M]

ACPL-C87at схема в даташите.

[Romale]

ACPL-C87at схема в даташите.

Thank you

[slociviccoupe]

Since some of you seem to know honda inverters. What about the "dumb" inverters from the insight, civic hch1, accord, and the crz. They have logic level inputs for each phase high/low, built in power supply and gate drivers, serial data out for voltage and igbt temp.
They have external phase current sensors.

I have a v3 main bord and would like to run the honda igbt but have few questions. Seems the pwm output of the phase signals are 3vdc, my igbt takes 5vdc logic level inputs. Just use a logic level shift breakout board?

The circuit for the contactor. Pre charge , ect, can they drive those 12vdc directly or does it need a circuit to drive those?

I am trying to replace the mcm that is the limiting factor on the crz with an open inverter main board and keeping the car hybrid.

[arber333]

I just use the v3 O.I. board in the ima inverter, the current sensors and the driver power board are regular, the ima processor did not solder off and did not close any legs to the mass. from the ima, I take two wires with current sensor signals relative to the total minus of the 12 volt power supply. +5 volts current sensors are received from the standard board. also, relative to the general minus, I supply 6 pwm wires to control the transistors.
and everything.. it works by itself.

I went back to this theme because i want to test IMA inverter more and i cant process the part in bold. Can you explain please...

1. Do you desolder main chip, or do you somehow disable it? How does it work?
2. You simply provide the same 12V to terminals at the IMA side AND for the OI board?
3. What is the current sensors sensitivity mV per Amps? I remember somewhere the value 5.8mV/amp, is that ok?

Tnx

[Beggernator]

Hi, I thought I would answer where I can. I have been posting my work on a drop in mainboard for the second gen civic hybrid IMA inverter in the YA-HI (Yet Another Honda Inverter) thread.

1. Do you desolder main chip, or do you somehow disable it? How does it work?

on mine I completely cut away half the original PCB. however it is my understanding that the Honda inverter sends no commands with nothing connected it is possible to add your own control board on top without issue. honestly though I would rather not risk it, better to remove the standard MCU then have some odd fault blow your igbts up.

2. You simply provide the same 12V to terminals at the IMA side AND for the OI board?

yep, I use one 12v on everything. in theory each of the gate drives are isolated so could have separate grounds if you wanted but its not worth it. as is common in other conversions that use this inverter I also supply the 12v through an un populated capacitor footprint just above the main connector.

3. What is the current sensors sensitivity mV per Amps? I remember somewhere the value 5.8mV/amp, is that ok?

you are probably correct with that value. but you have to remember the ADC inside the STM32 chip is maximum of 3.3V while the current sensor outputs 5v so the open inverter boards use resistors to scale the input to a safe level for the chip, thus changing the actual reading per amp.

I think i remember seeing 4.65 il 1 and 2 gain in one of Romale's parameter files.

[Beggernator]

Since some of you seem to know honda inverters. What about the "dumb" inverters from the insight, civic hch1, accord, and the crz. They have logic level inputs for each phase high/low, built in power supply and gate drivers, serial data out for voltage and igbt temp.
They have external phase current sensors.

I have a v3 main bord and would like to run the honda igbt but have few questions. Seems the pwm output of the phase signals are 3vdc, my igbt takes 5vdc logic level inputs. Just use a logic level shift breakout board?

The circuit for the contactor. Pre charge , ect, can they drive those 12vdc directly or does it need a circuit to drive those?

I am trying to replace the mcm that is the limiting factor on the crz with an open inverter main board and keeping the car hybrid.

sounds like those inverters are good candidates for an open inverter board.

for voltage levels I would suggest checking the gate drive circuit of the inverter you plan to use. my Honda HCH2 IMA inverter has opto isolators for the gate signal connections. Because inside each opto is an LED anything over 2v will allow current to flow and count as a logic high and drive it correctly.

A simple way to check if this is true is to put a multimeter in diode test mode and put it between one of the logic lines and ground. should be a reading of around 0.6 to 1.8v this is the forward voltage of the diode in the opto isolator. next reverse the multimeter leads and you should see no connection on the meter, this is the reverse current on the diode. if you see a small voltage reading that doesn't differ much forward or backwards then you do not have an opto isolator and it will likely need the full 5v logic.

It should be stated this test should be done with no power applied to the inverter.

The open inverter v3 board will drive pre-charge and main contacts. its all listed on the wiki page max current ect. plus some helpful pinout pictures.

interesting you are doing it to a CRZ, I was looking at getting one as a daily for a while. you are probably better off doing a current spoofing hack and manual IMA control. people got close to 30kw out of the stock setup with minimal effort on the insight central forums. (plus you can have a 20kw home backup generator if you rev it up and command full regen)

[arber333]

on mine I completely cut away half the original PCB. however it is my understanding that the Honda inverter sends no commands with nothing connected it is possible to add your own control board on top without issue. honestly though I would rather not risk it, better to remove the standard MCU then have some odd fault blow your igbts up.

Yes I am considering if it is ok to just add Lebowski brain directly to inverter or would the main chip protest in some way. You have to remember current flow and desat sensing is still active when power supply is on...

[Beggernator]

Yes I am considering if it is ok to just add Lebowski brain directly to inverter or would the main chip protest in some way. You have to remember current flow and desat sensing is still active when power supply is on...

Yeah you probably could, but is it worth the risk? really isnt much effort to remove the main MCU. not like it needs to remain functional. could tear it off with a screwdriver if you wanted to. just make sure nothing is touching afterwards and you are good to go.

I will be interested to see how the Lebowski brain performs, I stayed away because it didn't seem to be in active development anymore, and I didn't like the fact it didn't run resolver position feedback.

[arber333]

Yeah you probably could, but is it worth the risk? really isnt much effort to remove the main MCU. not like it needs to remain functional. could tear it off with a screwdriver if you wanted to. just make sure nothing is touching afterwards and you are good to go.

I will be interested to see how the Lebowski brain performs, I stayed away because it didn't seem to be in active development anymore, and I didn't like the fact it didn't run resolver position feedback.

I wanted to ask Romale how is his view on the matter as is personal experience.
What i am concerned with is would there be any difference if i removed the main chip vs if i keep it? Since he said he kept the chip in i am wondering if that would somehow prevent the EMI from throwing signals off. I.e. maybe those pulldown paths to the chip are keeping the other brain from browning out....? Maybe some active resistor dividers work there...?
Who knows what will happen if i remove the chip and have lebowski conk out on me... I cant easily resolder the chip back on can i?

I had similar happening to me on two cars. My Pug would operate steady, but friends Mazda RX8 is stil randomly conking out on regen... i cant seem to find any remedy....