HE351 code for me78569

[me78569]

I do have to say with the current tune I am at

7000' altitude

33 psi boost, 35-40 psi drive
1295 EGT
max rpms 145,000

10psi by 1400 rpm, I am sure I could generate more, but I don't like to lug the motor that much.

I think I can drop some shaft speed, but keep boost up if I was to be able to read engine RPM. 

So overall pretty happy, there is some more to be gained, but overall it spools like nuts and pulls nicely, egt's and smoke is VERY low.  I can take off smoke free at 25% then hammer it after rpms come up above 1600, and be %100 smoke free.

[hakcenter]

love to see obd working, that'd be pretty nice.

I wonder if the TPS plug is in the manual truck harness. Anyone got a pic of location for a 2nd gen, and what it looks like ?

[Rx7man]

3 prong metripack flat, wire color for the TPS is Orange/bark blue (looks black)... you should have one since yours is auto anyhow... just tap into it (and sensor ground (black/green?)) and bring it to a analog input.. just sensing the TPS alone (even without RPM) would make a world of difference... I don't see it on my truck, but it's a standard

[me78569]

Yea tps is huge for getting things working well.

Rpms will be even bwtter.

[hakcenter]

My truck is a 5spd bruh hopefully eventually a 6spd.

Just wondering if the harness is in there, I'll look in the engine bay tomorrow. Have to get in there and fix my broken HID ballast again. Things always blow the 400v cap every 6mo or so. No idea why.

[Rx7man]

weren't you working on an auto tranny in your basement? 

I'll steer my buddy away from the HID's in that case.. he works on his truck enough.

[me78569]

Added a warmup mode to my code. 

Code Select Expand

void WarmupManage() {
   if (IAT == 12 || IAT == 13 ||IAT == 14){
      if ( EngineRPM > 1000) {
      constrain(vane_position, 918, 960);
      vane_position += .05;
      if (vane_position > 940) {vane_position = 800;}
    }
    else{vane_position = last_vane_position - 10;}
   }
   
    else if (IAT == 22 || IAT == 23 || IAT == 24){
      vane_position = 960;
     }
    else {vane_position = 1000;}
   // }
  }


I am also working on getting my code ready for the obd reader that is on it's way.  I am gonna pipe in TPS, ECT, IAT ( hopefully don't know if that is j1939 or j1979), engine RPM.

I trying to work on a good way to adjust the turbo map in regards to RPM mainly. 

right now I am mapping engine RPM from (0, 3200, 0, 80) so I can use the 0-80 to adjust the vanes larger as engine speed increases. 

Code Select Expand

int ReadEngineRPM(){
if (obd.read(PID_RPM, EngineRPM));
  return map(EngineRPM, 0, 3200, 0, 80);                    // read the value from the sensor I am mapping the rpms to give a reasonable vane adjustment without having to divide later.
}

then adjusting the position

Code Select Expand


        if (turbo_rpm > curve_rpm[3]) {vane_position -= TPS_range;}
        if (turbo_rpm > curve_rpm[4]) {vane_position -= RPM;}

TPS_range is used in the same manner it used to be, I might need to adjust it some as it was put in place to try and predict engine rpms and open the vanes.

[Rx7man]

I think the OBD inputs are going to greatly help drivability and responsiveness... doing only an hour of tuning on my truck I'm already pretty happy with the way it runs,.. I gotta track down that exhaust leak though, my engine brake pressure SUCKS, and no doubt spooling is affected by it too.

My code for a 2d map may not be working too great on a low memory device.. it requires about 120 bytes... depends on the other bells and whistles you got going on.  My sketch would still theoretically fit on an Uno, but as I tested it the serial port starts getting gibbled up.  My map is a map of "byte", so I had to do some scaling.. I have 2 maps, one gives me my 'base position', so it's 1/10th of the turbo position, and the other is desired pressure.. the two maps share the array for the axis labels.  The instance of the map class has a function called "Interpolate" (taking RPM and TPS as arguments), which gets a smoothed interpolation of the values between the discrete points... it returns a floating point number

[Rx7man]

The truck sounds AWESOME when I'm warming it up with the engine brake on!.. at initial startup, the oil is pretty thick right now, so it takes a while for the turbine to get going, and then you hear it come on slowly and steadily.. When I was warming it up after playing darts last week there were some girls out in the parking lot commenting about it sounding like it was going to take off.. They weren't my type unfortunately

[me78569]

haha,

You should see my truck is 3 cylinder high idle with the EB on haha.

I can go from 20*f coolant to 160*F in <5 minutes.  egt's sit at about 1000* and it smokes and smokes and smokes and thumps and thumps.  sounds and looks like an old WWII v12 fighter starting.

I am really quite happy with mine at this point, but it can always be better.  Even at 7000' altitude I now struggle to make it smoke when the tuner is off.  Even with the tuner set to max it cleans up quick.

Anyways looking forward to reading through your stuff. 

[Rx7man]

Mine only stumbles badly when it's under 0F.. the engine brake sure does help it warm up quicker though, especially if you fast idle it to about 1200-1400 RPM,.. but the governor setup on mine makes it so it *really* doesn't want to be there.. it's either 1000 or 1500 when it's cold and no in between.  the 24 valve trucks were always a little easier to warm up than the 12V's were.. they are just miserable.. when it's -10F out, I can plug it in for an hour, warm it up for 10 minutes at low idle, and it still take about 8-10 miles (half way to town) before I get any heat.. I can idle it for a half hour and if I'm lucky I might get a little defrost action.

[me78569]

I think I am going to remove the turbo mapping functions and do more a step for each turbo rpm range. 

I can use the rpm variable to take it's place and I think it will work better and have less "slingshot" action.

[Rx7man]

Did you get my PM?

So far I haven't had any problems in my code with overboost... except when I had the logic backward (doh) and I momentarily wrapped the 35PSI boost gauge in my truck.. I don't see how the logic was backward because I studied and studied it but in real life it was still backward.. I created a class of variable that keeps track of the history, thus making it easy to determine the slope of any given parameter... in this case I use Intake manifold pressure (MAP), I start to open the vanes based on slope alone at about 20PSI, and start limiting based on actual pressure as well at about 25 PSI, this makes it come up to 20PSI fast, but starts a little bit of opening action if it's really spiking.  Again my code isn't the prettiest, but it seems to be working nicely

[me78569]

I did, I haven't read through it yet though, is there supposed to be a .ino in the pm's?  I only see your libraries.


Oddly enough I don't have any overboost issues, strictly overspeed.  I am thinking that I am getting invalid speed readings above 100k haha.  I see 30psi at ~130,000, but I have to drive rpms up around 160,000 ( arrording to the arduino) to see 35psi. 

The map doesn't really support that.  I am hoping that with the OBD I am trim out some fat.    I am not sure where my delay in the rpm read is though,  I see the same result regardless of what code I use.  gotta be in the board install someplace. 

RPMS are VERY steady 0-60k then jumps to 70-90k VERY quick and then more stable 100,000-130,000


I did test for AC noise on the alternator and I was sitting at .01 / .02 volts

Still haven't looked at the solder joints, I have been too lazy to pull the controller out haha

[Rx7man]

The .ino is going to be all wrong.. in the .c file I gave an example of how to use it though