Megasquirt Support Forum (MSEXTRA)

KeithBoden wrote:I think a good algorithm would be:

Look up throttle % vs VSS to get base pressure%, call this X
Handbrake is on? -> Is X > <hand brake limit>% ? If so, reduce X to the <hand brake limit>%
Launch mode on? -> Set X to the <launch mode pressure>% (or during the burnout in 3-step, set X to 0 or the <hand brake limit>)
X is above <pump threshold>%? -> Turn the pump output on, otherwise turn it off
Output the steps or PWM or whatever to set the pressure to X

Later add things only if needed like the temperature cutoff, vehicle brake input (although that is handled by the table lookup if you drop the throttle), solenoid, accelerometers, etc.

I think this handles everything in the CRC Haldex Performance link you posted and is actually more configurable.
Here is the hardware I'd use to make it quick:
https://www.adafruit.com/product/2590 $12.50
https://www.adafruit.com/product/2448 $5.00
https://www.amazon.com/CHENBO-Electroni ... 015W4D9WY/ $9

The code for the Arduino would be about 30 lines, which I already have done for another project; I'm using some of the same stuff to feed EGTs into my MS3-Pro and to do CAN to USB to listen to MS output data.

Times like this I wish I pay attention in class growing up... as best I can read the algorithm I think that would work the only thing I want to make clear is that I believe the handbrake and brake pedal send its own signal to the haldex controller so it opens a bypass because when people use the mechanical controller the 50-50 doesn't dissipate for a few miles after the pump is shut off that's why I'm led to believe that they have a quick bleed off solenoid connected to the brake pedal and handbrake so the pressure of bleeds off immediately.. those parts are reasonable and I have no problem purchasing them I have access to wiring and everything else I think I would need I would just need some help with coding like you said and maybe some circuitry but that I should be able to figure that out for the most part... I think the crc is a very good product mimic as a base point and then add too like you said if needs be! The only other thing that would be nice but I don't know if it will be possible is to have the three different modes... street race and 1/4. At the same point tho it wouldn't be hard to change it via laptop or better yet tune it or setup the tuning for all three in one...


I haven't looked yet but are we going pwm to stepper or do you think it would be better to use the iac stepper? Never Mind can bus

Here's that soleniod! Looking at those boards you linked and since its running can bus it would be easy to boardcast the diff temps and split % to a small read out too if I understand correctly...

???? METRO Mini is the culmination of years of playing with AVRs: we wanted to make a tiny, breadboard-friendly development board that is easy to use and is hacker friendly. Metro Mini can be programmed with the Arduino IDE (select 'UNO' in the boards dropdown)

Do I need to buy it with UNO?

No problem, I can handle the MS code and help with the Arduino code and electronics. (Hopefully others will comment on the electronics too.) Shouldn't be hard to incorporate the solenoid too, we'll count on that as a must-have for the first try. Any other info you find on how it works would be helpful but we could assume that the solenoid needs 'hit' any time we need to drop the pressure.
As far as modes, I'd lean towards the MS theme which is to try to have 1 'tune' with enough intelligence to handle all conditions. Although, it wouldn't be hard to have multiple settings though for street/track/drag with a button to toggle between them.
We could feed the diff temp back into the MS via CAN so you can log it or display it on a gauge and of course take actions like cutting off the pump/pressure if it got too hot. I bet most people are just computing split% based on how much pressure is being applied: full pressure, assume 50/50, no pressure, assume 100/0. That approximation could be calculated in either the Arduino code or the MS code.
The UNO reference is just for programming the board, it is a setting you select when you want to send your program to it. How far along are you on the AWD swap itself? Is it already done and you are rolling around FWD or is there some mechanical work left to be done?

As of right now I'm still collecting parts I have everything for the swap but a clutch, front cv axles and a bushing set.. I'm working with someone as of this week to build me a jig since the car was built one way in Germany as what they call the syncro but then you cut that dif carrier off to weld on brackets for the haldex... there are some added torsion differentials I'd like to buy before I do the swap so that I don't have to pull the transmission and diff out again but in all honesty it can and should probably be done latter anyway since there 1k a piece ... so if we could make this work for relatively inexpensive price I'm more than willing to compensate you for your time as well I should be ready to start testing it in car I'd say within a year faster if I decide to start this winter... oh and I honestly think the solenoid I'm talking about for the pressure relief is operating off of a 12 V seeing it when the brake light/handbrake switch is pushed so I'm not sure we would have to adopt that necessarily into the process unless you see a need beyond oem. Exactly correct I think they use a 6 position step 50/50 60/40 70/30 80/20 90/10 100/0 I bet the CrC control even seems to have finer controls either way I don't think that'll be much of a challenge I think the challenge is gonna be getting one tune control page to work for all aspects which is also a huge plus since no one will have this ability on the market. Thanks again I'm gonna order the parts I can start bench testing at anytime

I have some circuits worked out and some code that is looking good for this project. I'll be testing most of it over the next month on a different MS3-to-hardware integration effort I'm working on; PM me your email so we can stay synced up on your AWD build and we'll cut back over to this thread when it is Haldex time!

Awesome. I have some r32s to convert this plus a dbw solution and I am set!

Cool I'll get those parts ordered, I'm waiting on a jig to swap the mount brackets on the rear beam but I'm not scared to do bench testing I could even drive the diff with an electric motor

Alright parts are ordered!

Good to hear! My other MS integration project is coming along nicely and I should be able to use a bunch of that code for this as well. PM me your email address so we can work out some details with the circuits and so we can ship code back and forth.

Sorry I wasn't subscribed

Hello, guys, allow me to chime in;

I am in a similar position where I converted my 99 Golf from 1.8t to complete V6 engine and powertrain ( Gen 2 HALDEX) which I turbo'ed obviously!!
My car is still on jacks for 5 month until I figure out what to with haldex...

While researching I came to conclusion that the haldex can be controlled with CAN bus, the reason for this is that many aftermarket 4wd controllers are able to act as a standalone by only sending messages through the CAN bus.

I managed, by using ARDUINO and CANbus shield to send various data to instruments cluster and succeeded. I never found any source for haldex CAN message ids, so tomorrow I will try to use VCDS's and output test the haldex unit and sniff some data with the help of Chinese CAN to USB, with some luck I will get some info.

I believe this way, safe measures on the haldex controller will remain intact..
Will post back my results..

doubas wrote:While researching I came to conclusion that the haldex can be controlled with CAN bus, the reason for this is that many aftermarket 4wd controllers are able to act as a standalone by only sending messages through the CAN bus.
...
I believe this way, safe measures on the haldex controller will remain intact..
Will post back my results..

If you can discover the required CAN messages, then adding them to the firmware is not usually difficult. However, that initial reverse engineering step could be very difficult indeed.

James

It was difficult indeed!!! for some reason I couldn't manage to even read or identify the HALDEX with VCDS... Investigating..

Also from what I understand is the stock control is only good for roughly 80/20 split that is why the modded blue controller box came out and then the orange one. The good news is its code able from what I've just learned and I do know someone that claims to have the Audi Motorsport code which was only given to the racing teams.. still not quit sure how your going to communicate fully with the stock can bus controller when we don't have steering angles and a few other sensors.

Also I have not heard from Keith so there's a possibility we might've lost one of our leads on this

I've read once about some standalone Haldex controller that it could function even without the ABS is present..

Well after a full day of research and experimenting, I found out that HALDEX controller only work with the presence of certain set of messages from the ECU/ABS or any other device mimicking those such as wheel slip, accel pedal, steering angle...
Now since this is out of my league for now, I will order some stepper motor controller for arduino and try it out.

I confirm it is working with a potentiometer, lcd screen connected to Arduino to the Haldex stepper!

Next step is to get CAN commands set by MS.

doubas wrote:Next step is to get CAN commands set by MS.

So have you had to modify the internals of the controller or are you using 100% OEM stock controller if so I'm gonna talk to my buddy that has the Audi motorsport commands which I believe are can code signals if I understand correctly