t vago's MPGuino workspace thread

[t vago]

Originally Posted by FlashStopFall

Alright, I'm a bit lost. I built an mpguino using the legacy schematic with a Nano. I don't have the LCD yet, it comes tomorrow, but it appears to be working. Only thing is I tried to enable serial debugging so I could see some output on the arduino IDE serial monitor, but it never outputs anything. Is there a #define that I''m missing?

What version of MPGuino are you using? This is the thread of an older version of my own fork of MPGuino. I have since extensively refactored the code and am currently working on adding native USB support for AtMega32u4-based boards.

[FlashStopFall]

Originally Posted by t vago

What version of MPGuino are you using? This is the thread of an older version of my own fork of MPGuino. I have since extensively refactored the code and am currently working on adding native USB support for AtMega32u4-based boards.

Oh nice! Is there a new thread for the newer version?

I'm testing 1.94 and 1.95. I got the LCD so I'm not as worried about serial, but I have another problem. My LCD displays garbage data with every refresh. Sometimes I'll catch a glimpses of an "mpg" or a "0.00" but that's it. I'm using an arduino Nano clone with the atmega328p, set up to use 4 bit parallel for the 16x2 LCD. The LCD works with other sketch examples so I know the wiring is fine, but I'm not sure what I'm missing. I had everything wired but I stripped it back to just the LCD stuff for testing. Do you have any thoughts about this?

I just went back and tested 1.93, and it works for the first two frames, but then becomes garbage also

[FlashStopFall]

Originally Posted by FlashStopFall

Oh nice! Is there a new thread for the newer version?

I'm testing 1.94 and 1.95. I got the LCD so I'm not as worried about serial, but I have another problem. My LCD displays garbage data with every refresh. Sometimes I'll catch a glimpses of an "mpg" or a "0.00" but that's it. I'm using an arduino Nano clone with the atmega328p, set up to use 4 bit parallel for the 16x2 LCD. The LCD works with other sketch examples so I know the wiring is fine, but I'm not sure what I'm missing. I had everything wired but I stripped it back to just the LCD stuff for testing. Do you have any thoughts about this?

I just went back and tested 1.93, and it works for the first two frames, but then becomes garbage also

Okay I found it!! Flaky power supply to the LCD caused the garbled data!

[FlashStopFall]

t vago, I just wanted to say thanks again for all your work on mpguino code, you have probably single-handedly kept mpguinos alive and accessible for newbies.

I've got the hardware and software figured out now I think, and I have the arduino just set up on my bench. I've got the VSS lead and switched ground injector leads just tied to ground right now, and the data fluctuates all over the place. Is this normal behavior? Aka, do I need to hook it up to the car to see zeroes (or ---) in g/h and mph?

[FlashStopFall]

Once again, a little time solved my problems. I didn't realize that the code would simulate sensors by default - after finding and turning it off in settings, everything works.

[t vago]

Originally Posted by FlashStopFall

Once again, a little time solved my problems. I didn't realize that the code would simulate sensors by default - after finding and turning it off in settings, everything works.

That's more due to my not properly vetting the code before posting it.

Sensor simulation is a development feature, as you've guessed. This must be turned off before MPGuino can read VSS or injector signals, as sensor simulation overrides these inputs. Either leaving out this feature altogether at compile time, or manually turning it off via the sensor page in MPGuino (if it was compiled in), will enable MPGuino to read VSS and injector signals.

I think that Jeeps have slightly different fuel pressure specifications than the Dodge I used for development. 49 psig for fuel pressure, IIRC. My Dodge used 58 psig.

[FlashStopFall]

Originally Posted by t vago

That's more due to my not properly vetting the code before posting it.

Sensor simulation is a development feature, as you've guessed. This must be turned off before MPGuino can read VSS or injector signals, as sensor simulation overrides these inputs. Either leaving out this feature altogether at compile time, or manually turning it off via the sensor page in MPGuino (if it was compiled in), will enable MPGuino to read VSS and injector signals.

I think that Jeeps have slightly different fuel pressure specifications than the Dodge I used for development. 49 psig for fuel pressure, IIRC. My Dodge used 58 psig.

I think that’s correct. I’m excited to try out the barometric pressure adjustment once I get everything dialed in. I did have a thought, my jeep has a fuel rail pressure port I could hook a pressure sensor up to. My thought was that this could provide more accurate injector g/h data because of the Chrysler fuel pressure setup. Do you think this would work the way that I’m thinking, and would it be any more accurate than MAP/barometric pressure correction?

Also, just found the newer thread for your mpguino code, I suppose I’ll post there from now on as I’m using version 1.95

[t vago]

Originally Posted by FlashStopFall

I think that’s correct. I’m excited to try out the barometric pressure adjustment once I get everything dialed in. I did have a thought, my jeep has a fuel rail pressure port I could hook a pressure sensor up to. My thought was that this could provide more accurate injector g/h data because of the Chrysler fuel pressure setup. Do you think this would work the way that I’m thinking, and would it be any more accurate than MAP/barometric pressure correction?

Eh... I'm not sure that will work.

The problem is, that at the interface between the fuel injector nozzle and the intake manifold, there is a varying pressure differential. Most cars handled this by varying the fuel pressure with a modulated fuel pressure regulator operated by intake manifold vacuum so as to maintain a more-or-less constant pressure differential.

Chrysler fuel systems were different in that they forced fuel pressure to be constant (either 49 or 58 psig) and used manifold vacuum to figure out the differential and therefore the amount of adjustment needed to deliver the precise amount of fuel.

I'm thinking that if you do install your fuel pressure sensor to the fuel rail port of your Jeep, you will find it will have a fuel pressure that will vary a little bit due to absolute fuel consumption (ie, engine loading), but you'll still have to compensate somehow for the pressure differential between the fuel system and the intake manifold vacuum. Such a differential will cause anywhere between 3% to 9% variation between the rated flow rate of the fuel injectors.

You can still pursue it, and I'll help as much as I can.

Originally Posted by FlashStopFall

Also, just found the newer thread for your mpguino code, I suppose I’ll post there from now on as I’m using version 1.95

Good to hear! It is of course still a work in progress, so do let me know if you run into any problems with it.

For instance, I just discovered a nasty little bug related to the refactoring I did to the Chrysler fuel injection correction. Yay... The bugfix should be committed to the git repo presently.

[FlashStopFall]

Originally Posted by t vago

Eh... I'm not sure that will work.

The problem is, that at the interface between the fuel injector nozzle and the intake manifold, there is a varying pressure differential. Most cars handled this by varying the fuel pressure with a modulated fuel pressure regulator operated by intake manifold vacuum so as to maintain a more-or-less constant pressure differential.

Chrysler fuel systems were different in that they forced fuel pressure to be constant (either 49 or 58 psig) and used manifold vacuum to figure out the differential and therefore the amount of adjustment needed to deliver the precise amount of fuel.

I'm thinking that if you do install your fuel pressure sensor to the fuel rail port of your Jeep, you will find it will have a fuel pressure that will vary a little bit due to absolute fuel consumption (ie, engine loading), but you'll still have to compensate somehow for the pressure differential between the fuel system and the intake manifold vacuum. Such a differential will cause anywhere between 3% to 9% variation between the rated flow rate of the fuel injectors.

You can still pursue it, and I'll help as much as I can.



Good to hear! It is of course still a work in progress, so do let me know if you run into any problems with it.

For instance, I just discovered a nasty little bug related to the refactoring I did to the Chrysler fuel injection correction. Yay... The bugfix should be committed to the git repo presently.

Ah that makes sense! I didn’t consider the varying pressure in the manifold relative to the rail.