Cousin of OSW (Open Sim Wheel)

Why did it fail ?

Not sure. I used XLoader and put my Arduino in bootloader mode and it just shows "Upload failed" in the bottom left corner. I tried another hex file and it worked so idk why.

Hi mate, i know i am bumping an old thread, but its still a widely used method and it seems that the solution you give here is really important,

Can you help me and others for future how to cool the fet down, cheers.

I am using ibt4 for driving my rc buggy but any source on the ibt 4 turns up a blank apart from here.

Thank you

Although very off topic, i'll add this:
The IBT4 is much different from the IBT2 which most of us use. With an IBT4, the cooling options are not great since you can't attach a heatsink to the other side of the board like you can in IBT2. The black part of the chips/fets themselves is not a good heat conductor. Your options are pretty limited.

If I recall, it was determined here that the IBT2 is still the better option for our application but most run two or more in parallel, possibly look at running two of the IBT4 in parallel too?

Hi mate thank you for reply, even with a thermal pad on the ibt-4 it doesn't pass anywhere near enough heat dissipation to be of use, so I am heading down dual ibt-2 route, if anybody here can help with a schematic of linking 2 or more that would be great fully appreciated,

I can offer a solution for a motor that is designed for low amp stall by nature using gearing but yet still moves fast enough for the steering wheel, in your builds

It's an electric power steering rack from a car, mine is a Honda jazz rack, i chose this rack because the motor is to one side of the rack and that means I could cut the rack down and add tie rods to the rack directly driving the wheels without any loss of torque from servo arms or linkage!
I did look for ages for a solution but nobody had ever used this method that i had found before so i was glad to share it with others for large scale steering.
You can also use an electric power steering column, naturally these designs are made to stall when driving. NSK make newer brushless EPS units that are used on nearly all cars sold in the UK, but sadly I don't have the skills to speak to the modules (believe me I have asked NSK technical for a pinout, but they just ignored my request) they retail about £30 on eBay (look up Nissan Juke electric power steering) you will be able to buy the full column module motor for that price which I expect will be a solid contender for your designs.

I hope to have given you all something back that may help with your builds.

There is lots and lots of retro power steering on YouTube as it's a simple mod to do and pinouts for brushed motors as it's a drop in replacement Columb for pretty much most cars.

If the ibt2 ends up being a no-go then I'm aiming to buy a programmable controller (i have attached photo) which will be a 60 AMP unit continuous with 120a burst (adjustable by programming).

But first I want to try the ibt2 units as I feel there is value to be found,

All the best


Cheers

Also, i forgot to add that you were able to add additional cooling to the ibt-2 would you be able to offer any advice on a fan based cooling solution for the unit's.

Finally, on my ibt-2 units I see the top of the fets show a few holes of the board exposed, should i attempt to fill them all with solder will it aid with cooling.

Cheers

Interesting, I know at least some others have tried the power steering motors but they were not impressed with the results. I use 3 X IBT_2 on my 60+A setup and they barely get warm but I do have fan assistance. https://www.thingiverse.com/thing:3208235 however I have redesigned it and now running a 120mm fan for noise reduction.

Key points/mods:
Remove the heatsinks and chips
Clean the pads and fill all of the vias with solder - Yes, like you state
Replace chips
Install a heatsink pad between the board and the heatsink - The board is often warped, creating a void and it ensures it is insulated electrically

As for the wiring, just loop them all to the same inputs across all modules. Very simple. However, if you are running from a 3.3v logic controller i.e STM32F4, ensure you power the IBT Modules at the same voltage and NOT 5v or you will have issues. I highly recommend a dedicated 3.3v source such as the cheap AMS1117 modules available too as the onboard vreg might not have enough juice.

** Earlier back in this thread there is a connection diagram to use a buffer - This is NOT needed, just run the modules at 3.3v and not 5v. The IBT modules already have a signal buffer installed on them!

Thank you once again for a detailed reply on the correct method of driving the motors with multiple ibt2 units.

We are usually driving the ibt units using rc servos as the controller unit.

Whereby we would dissect a servo and remove the motor. We would then take the output of the servos h bridge ( the motors wires) and send the current directly to the ibt unit. Of course it is not the optimal way of doing it but it does work.
As when the ibt sees input 1 low or - x volt it shuts and input 2 goes high it opens. I'm guessing the servo h bridge pulses pwm and that pwm signal is seen by the ibt copying the pulse.

The latter is a guess as I am no means skilled enough to fully grasp how it works using the method. So if anybody out there can help teach me that would be great.


In linking the h bridge outputs are we running serial or palleral for the motor output?

I ask because I have not found the wiring diagram in this thread as of yet. If you know what page it's on would you kindly let me know.


Wondering if it's possible to use thermal potting compound or paste.

Cheers for your help as always.

I am currently in the process of building my first motion sim rig and have been reading through this thread for ideas and do's and dont's about making my own direct drive wheel and i have a couple of questions on what parts can be used:

Would this work correctly together:

Motor: Dayton MY1020 1000W 3000 RPM 48V DC Motor $99aud

Driver: I was wondering if a cytron MDDS60 would work? $222aud
These are a 60a dual channel with the ability to mix
the two channels and they are reliable and keep cool.

Control board: STM32F407G Discovery $30aud

Encoder:E6B2-CWZ1X 2000P/R 0.5M $550aud

Power supply: I plan to have a couple of batteries on the rig in paralell with a power supply to keep them charged for 2dof with Tl, Can i take power from the rig?

Misc items needed to complete: Approx $100aud

Total $1001aud
These prices dont include shipping

Now my last question is and don't flame me to hard lol,

I have had a look around at comparable options based on the above prices so please let me know if thats way out and i hope it is!

But if they are right then:

The Fanatec ClubSport Racing Wheel V2.5 BMW GT2 is $1299Aud (not direct drive though)

Theres also the Accuforce complete wheel for $1,403.99Aud

Then theres the AccuForce Wheel DIY Kit thats $841.83Aud

I am all for Diy projects and i was keen as until i started pricing all the parts up is it really that expensive to complete this project?


Thanks everyone

I have read this thread about 8 times to make my flyer that is in my profile photo.

we did 2 with a friend and we have equal parts and different parts but the same problem.

As I noted, in the 32 pages the only place where my problem is mentioned is here, that's why I highlighted it.

equal parts:
- Source, 12v 40A https://articulo.mercadolibre.com.a..._JM#position=3&type=item&tracking_id=71eab250 -c442-4bdd-a384-82a7c3428aa2
- my 1020 48v 1000w motor
- 3 H bridges bts7060

Different parts
- Software:
* mmos
* EMC
- encoder:
* 5v AB 1000ppr directly connected to stm32f103
* the other 12v with 1000ppr ABZ optocouplers
- motherboard:
* stm32f103 for EMC
* stm32f407vet6 for MMOS

We have the problem of it being off-center with the passage of laps during the race in Assetto Corsa Competizione.
Sometimes if we do a spin without crashing the steering wheel makes a greater offset and is at 30 degrees.
If we have no incidents, the steering wheel starts from 0 degrees and ends at about 5 degrees at the end of the 1 hour race. it is always off-center to the same side.

things we have tried to fix the error.
- we move the electronics away from the motor to avoid EMI from the my1020 motor
- we wrapped the electronics and encoder in aluminum foil
- we change cable diameters
- we put 2 sources in parallel to get 80A
- We used the https://www.thingiverse.com/thing:2974162 modifier design to have more space inside and thus be able to put all the cables and we do not know if when everything is there it generates a problem.

We are going crazy! We can't fix it.

help please .... We are available to try everything that is necessary.

Thank you very much and sorry for the google translation.

use an encoder with less pulses or a faster arduino board

I do not understand anything....

EMC stopped supporting arduino to use STM32F103c8t6 because the processor is much better and can apply more functions to the FFBB. For this reason we are using STM and we stopped using Arduino leonardo.

Regarding MMOS we use STM32f407vet6 https://articulo.mercadolibre.com.a...osition=1&type=item&tracking_id=ecb943ca7-7-7 9e00-e24dca4215c6. The stm32f4 discovery board as far as I can read it supports 2500ppr encoder which in the configuration ends up being 10000ppv. I am using 4000ppv. Okay, it is not the same board but it is almost the same processor. I don't think it doesn't support less than 1000ppr.

I can not find the technical explanation why it would not support a 1000ppr encoder both AB and ABZ. If you would be so kind as to explain it to me technically please. Thank you

it is not just the speed of processing, but how the program counts pulses - normally a counter counts the pulses using an interrupt - an interrupt normally breaks the main loop

there will be times where the processor is busy running the main loop and a pulse may get missed
there may be times where the main loop is broken and the logic does not count the new pulse

it is better to have 2 chips one that only counts pulses and sends this information to a second main processor

the code for this project is not available and may not have been written to take advantage of a faster/split processor

try running an encoder with less pulses

-----------------------------------------------------------

do you have a pullup / down resistor on your encoder signal ?

you may also be reading interference/noise that is being mistaken for an encoder pulse !

The stm32f103 family do have hardware encoder counter. No need for a wonky software solution. The developer of course do have to implement this feature.
Arduino code is normally extremely bad written and does not perform well.

its time for somebody to write new code to take advantage of the later faster hardware

a good place to start is sniffing the ffb data transmistted by games and understanding what exactly is transmitted

it may be a lot simpler than we think !

Have you checked the voltage levels on the encoder lines? They should go from 0 v to full logic level voltage. (3.3 or 5 v depending on which I/O ports they are connected to) If it doesn't fully reach the full logic level voltage, you can try a lower value on the pullup resistors.

By using an 12v 50a psu can a MY1016 350w be sufficient to get a strong ffb? How many bts are needed?

Two is enough. My experience is that they work ok up to 30 A before the start to cut.

Could I go with 1bts? The motor should be on avarage on 20-25A I think (?). So 1 bts can handle 43A max.
I saw a couple of DIY wheels where they used my1020 - my1016 and they drilled the backside of the motor to mount the rotary encoder.... Is there any tutorial or something that shows how to do it?

Hi mate I would say go with 2 or more, they are cheap to buy and simple to set up, the bts7960 can handle 43a max on the data sheet, but it depends on what the fet is mounted to which limits its ability to cool,

The ibt 2 is not the best design in cooling so i would link a few of them together to ensure they stay cool, you could of course test with one at a time, and see where the sweet spot is, or just go in with 3, in regard to mounting to the motor shaft i cannot help you on this as I am using a different method of calculating my position but im sure lots of other people will chime in here.