Building a he351 standalone controller

[Me78569]

I think the arduino has it's own regulator because it is a 5v system.

[Varrieablefirstgen]

Hey would you know why my EB isnt working correctly? I checked the values in the code and they are set to 900, 1000 which is right for 3cm but when i engage it it goes to 25cm just to try it i changed the values to 40, 140 and it did not make a difference? Any help would be appreciated everything else seems to be working as it should.

[Mopar1973Man]

Can you write a bit of code to show the variable for the vanes so you can see if they are actually be set right or is there another bit of code over riding your value. I sometime fight web code the same way and just plop in a more or less "print X" and look at what the software is doing. I've made coding mistakes where the set value was being over written by anothe command. 

[Varrieablefirstgen]

When it comes to code i really dont know what im doing ive done a little research on it but that only gets me so far. The code i have is from the building a standalone controller on this fourm, but the tests i have done the turbo speed sensor is not hooked up idk if that will affect it.  All i know i test fitted the controller on the truck and the pot sensor works good and the cm^2 numerals on the lcd screen are correct to the vein position. Do you know of a video or source online that i can use to troubleshoot this ?

[Mopar1973Man]

I remember @Me78569 making a calibration tool for the turbo or a way to reset the vane position. I wonder if that would help figure this out.

[Varrieablefirstgen]

i wouldn't think so but i have no idea honestly i have the "turbo reset code" and it positions the veins where they should be then when i upload the main code the veins will cycle the way they should be IE if theres over 10 psi of boost the veins will open the pot switch will work correctly with the lcd screen 25cm is opened all the way and 3 is closed maybe the code is getting thrown off since the tps isnt hooked up ill try editing that out just to see what happens.

[Me78569]

it might be.  I dont honestly remember if there is a stop limit for high and low values allowed to be used.  It might be trying to close to 3cm then wrapping around to 25 

 

 

 

 

[Varrieablefirstgen]

Interesting if that is the case how would i go about to fix this issue or atleast diagnose/troubleshoot?

[Great work!]

You need to limit the value of the variable: final_vane_position right before the CAN send function so that only valid values are sent. So that if something goes wrong while tinkering.

 

Insert this:

final_vane_position = constrain(final_vane_position, 40, 960); // prevent non valid position data

Just before line:

byte lo_byte = lowByte( blah blah);

Try that. You can hack up an  lcd.print(drive or whatever); to lcd.print(final_vane_position); to actually see the value.

Edited February 24, 2021 by Great work!
Spelling

[Varrieablefirstgen]

Ok thanks ive been pretty busy lately but when i get some time ill give it a shot!

[Eco77]

Hi Everyone,

I’m following this forum for long time and I wanted to thanks everyone for sharing their knowledge and experience, I learn a lot from you guys, so thank you very much.

 

Currently I’m working on refurbished Holset He300VG turbo actuator, and I’ll be glad if someone can assist me little with controlling the motor.

I’m using 12v\10A power supply.

I connected everything according the instruction.

(On holset main plug the connection are: (GND – CanL – CanH – 12v). (See pic)

I sent the calibration code: (CFFC600, extID , 8,FF FF 02 FF FF FF FF FF)

And I can see the motor rotating to one direction as it supposed to do, after that I must recycle the power.

I connected my Arduino to can bus shield and then connect it to 4 potentiometers + 2 on\off switches.

The sensor connection are:

Boost pot – A0

Drive pot – A1

Pot – A2

TPS  - A3.

“Switch” connect the GND to digital pin 7

“EBswitch” connect the GND to digital pin 8 (just to make sure pin 9 is not used by the SPI (See pic)

I upload the A_HE351Mainfreq_TPS 1.11 sketch (not the 1.12), the LCD has some problems so I comment his lines.

My can bus shield use pin 10 to communicate with the Arduino.

Next I checked the can bus messages using Ixxat (usb to can). (See pic)

And I can see the numbering changing when I move the pots and switches.

The problem is that the motor is not rotating left or right.

I saw on the sketch :

//const int RPMpin = 2;
//const int RPMpinInterrupt = 1;
//int RPM = 0;
//
//volatile int RPMcounts;
//const byte CrankPulsePerRev = 2; //How many crank pulses you get.. if your RPM readings are off, check this.. 

 

I connected everything in my lab so I don’t have RPM reading, can the actuator work properly without these readings?

Maybe you have an idea what else can I do?

I upload some pictures, may be it can help...

[Great work!]

Is rpm sensor 2 or 3 wire type?

Try toggling LED on uno and connect to scope for debugging. Toggle LED  in the CAN send funtion to test. 

Try short cutting the code by sending a pot reading to CAN.

Example final_vane_position = (analog.read(A0) / 2); where your pot connects. 

[Eco77]

Great Work!
thanks for your reply.

I didn't connect the rpm sensor, I saw in the code it was in comment. can the stand alone tester work without the rpm input?

Is there a way to simulate this sensor, lets say sample the throttle value that is corelate to the rpm value and insert it to pin 2 (as written in the code)?

 

5 hours ago, Great work! said:

Try toggling LED on uno and connect to scope for debugging. Toggle LED  in the CAN send funtion to test. 

I'm not sure what did you mean by that, you want me to send a canbus a command in order to toggle a led? I can try it, but I believe my communication is working well after I sent the calibration command : 0x0ff0600,extID,8, 0x00,0x00,0x02,0xff,0xff,0xff,0xff,0xff - and the motor rotate to one direction.

 

5 hours ago, Great work! said:

Try short cutting the code by sending a pot reading to CAN.

when I rotate the pot I can see the number changing. so I know the can bus its ok.

 

5 hours ago, Great work! said:

Example final_vane_position = (analog.read(A0) / 2); where your pot connects. 

so, instead of sampling the outside potentiometer (pot), you suggest to sample the boost pot(A0) divide by 2 and send it to the final_vane_position?

void set_turbo_position() {
    //final_vane_position = vane_position; //original code
    final_vane_position = (analogRead(A0)/2);

what we want to accomplish by that? I tried it without any change....

i dont have the max9924 chip so i tried to simulate the rpm by sampling the throttle pin(A3) and send it as PWM to pin 2 (RPMpin2)  - what do you think about that? can it work?

[Me78569]

The pot to set vanes manually will override the RPM calc section of the code, but hte pot I used was a push pull pot to enable or disable on another pin, I dont remember which anymore.  

 

The most common thing I seen was the wrong pin being used for the can control due to issues with board differences / spec and other conflicts on the can pin.   

 

You might try a simplier versioin of the code to only send can position based on pot position.   however if the turbo acts like it wants to find the limits when the calibrate command is sent, but does not want to do anyhting when the normal code is running then it suggests that the can communication is working, but something else isn't.   

 

another thing is if the code does not fire off a message every ~50-200 ms then the turbo should "reset" or jump back to a position over and over.   I dont remember how long though,  if I remmeber right I think it jumps to wide open.

 

Edited March 8, 2021 by Me78569

[Eco77]

28 minutes ago, Me78569 said:

The pot to set vanes manually will override the RPM calc section of the code, but hte pot I used was a push pull pot to enable or disable on another pin, I dont remember which anymore. 

thanks for your reply, according the code you use only pins A0, A1,A2,A3,7,9,2,1 so maybe you used the push pull pot on another code.

 

34 minutes ago, Me78569 said:

another thing is if the code does not fire off a message every ~50-200 ms then the turbo should "reset" or jump back to a position over and over.   I dont remember how long though,  if I remmeber right I think it jumps to wide open.

you right about that, i remember I read somewhere that can messages should be every 10ms, i capture my communication with my scope and the can messages are every 2.080ms so i believe its not the problem.

 can you or someone else can upload simple code for moving the motor with only one pot? in the past i took the calibration code and changed it to read the pot and send the command to the vans, but with no success...

 

[Me78569]

looks like pin 7 is what I used to enable or disable the pot.    pin 7 will have to be triggered to enable the pot to work.  

[Eco77]

you right, when pin 7 is grounded the pot is changing the can bus message dramatically. but still nothing happen

when pin 7 is grounded this is the can bus message

[Me78569]

the calibrate command should make the actuator find the limits in both directions.  Does your actuator rotate 1 way then back in the other direction?  or just one way?

[Eco77]

Hi,

It's a good question!!

when I'm sending the calibration command my motor rotate only to one direction, I believe its happen because he can't find the limits.. maybe he goes to "shutdown" state and 
that's why I need to recycle the power after every calibration command....

So what you actually saying is only after the calibration success (moving on both side and hitting the limits) I can effect the motors movement with the whole circuit..

its a good way of thinking.... thank you very much, I must try it !

[Me78569]

No sure honestly, I never calibrated with the actuator off.   I do know the power needs to be cycled between calibrate and trying to send a command to the turbo.