Any microcontroller success stories?

[Enki]

I'll be leveraging multiple Raspberry Pis and Arduino/clones (4 Teensy 2.0) in my project car.

So far, I've been able to do basic testing for air ride control and have successfully tested my Arduino Uno and air ride logic separately, but I've not yet built them together and done a full test. I'll probably make a new thread for that specifically.

For the project car, I'll have four RasPis for realtime video recording/compression and viewing (as they will replace the side mirrors; yes it's legal where I live) and minimum 5 Arduino/clones for air ride and misc control; four Teensy 2.0; one for each wheel for ride height control, plus maybe an IOIO Mint or equivalent to control the four Teensy and misc relays and whatnot (since it can hook up to the phone, I won't need a key to start it).

Anyone else doing anything interesting with their cars electronics wise?

[timjs]

I just found this thread by searching for "arduino" on google. ha!

Sounds like a fun project!


I'm wanting to build a methanol injection controller using Arduino, because for one I hate using a Hobbs switch because it is a mechanical component, and two because I've never touched an arduino board and I'm dying to.

I have no experience with coding or programming, but I tend to pick up on things like this quickly.

[Enki]

I'm actually considering doing just that using my Arduino Uno as a test bed, and finalizing the build using a Teensy because it's tiny like the name suggests, and low power.

I'm wary of tying into the MAF sensor, but I think that's the best way to do it aside from maybe a boost sensor. I'm not sure what way I'll go but I'll figure it out eventually.

[timjs]

I was thinking just using the MAP sensor's output as an input for the controller.
Maybe ramp up meth with boost via a PWM output. Idk how much amperage the pump will pull though. Probably less than 10 amps.

Basically, everything you can get in any of the meth controllers on the market, except 100x more awesome because DIY.

[CosmicArkie]

AEM pumps have a "protect with 15 amp slow blow fuse" label FWIW.

[sidekick]

I just looked up some arduino methanol controller builds and that would be awesome. If only someone would make a kit... @Lex;

[timjs]

I looked them up too, couldn't find a project that someone actually finished, though.

A kit would be nice, but Arduino is more of a prototyping platform. A company would make a working product with Arduino, then get some custom pcbs and a fancy enclosure and sell it for quite a bit.

[Roddiy]

Originally Posted by Enki I'm actually considering doing just that using my Arduino Uno as a test bed, and finalizing the build using a Teensy because it's tiny like the name suggests, and low power. I'm wary of tying into the MAF sensor, but I think that's the best way to do it aside from maybe a boost sensor. I'm not sure what way I'll go but I'll figure it out eventually.

I'd definitely hook to the MAP sensor, the logic on the arduino code is pretty simple:

while (Voltage > X) //Where X is the threshold and voltage is the variable sensor input
{
digitalWrite(transistorPin, HIGH)

}


where the transistor pin is a basic circuit that will allow you to switch on the pump by grounding the GND from it via a transistor.



Of course, you should write a function for this, where you call it via an If statement, and run it for as long as the while command is true.


This is very basic, but I have extensive knowledge with the Arduino and overall programming, if you ever get stuck don't hesitate to give me a shout, I have a few lying around myself and wouldn't mind dedicating some time to helping out.

[Enki]

Oh, I already have very basic air ride control code for my Uno. I've got that much of it down, and probably won't run into any coding headaches unless there's some foible I don't know about.

I'm more interested in how to set up a cascading MOSFET setup to allow variable speed pump control or introducing PWM to a couple solenoids for multiple nozzle staging.

[Roddiy]

Originally Posted by Enki Oh, I already have very basic air ride control code for my Uno. I've got that much of it down, and probably won't run into any coding headaches unless there's some foible I don't know about. I'm more interested in how to set up a cascading MOSFET setup to allow variable speed pump control or introducing PWM to a couple solenoids for multiple nozzle staging.

Fookin' read my mind, I just realized it and came in here to add that PWM control should be possible and not too hard.


example:

Arduino PWM control of EGR solenoid valve - YouTube

Arduino PWM control of EGR solenoid valve - YouTube

Oh and don't forget an awesome carbon fiber ricer panel with a basic form of Linux in a Raspberry PI, with dedicated GUI, allowing you to control everything...

[Ziggo]

Originally Posted by Enki Oh, I already have very basic air ride control code for my Uno. I've got that much of it down, and probably won't run into any coding headaches unless there's some foible I don't know about. I'm more interested in how to set up a cascading MOSFET setup to allow variable speed pump control or introducing PWM to a couple solenoids for multiple nozzle staging.

You need the Fets to run these types of thing with the audrino boards, but it works really well. A friend of mine converted his lawnmower to fuel injection using one.

Sent from my SAMSUNG-SGH-I747 using Tapatalk

[Enki]

Originally Posted by Roddiy Fookin' read my mind, I just realized it and came in here to add that PWM control should be possible and not too hard.

I've already done PWM to dim an LED...not that hard really.

Originally Posted by Roddiy Oh and don't forget an awesome carbon fiber ricer panel with a basic form of Linux in a Raspberry PI, with dedicated GUI, allowing you to control everything...

So, basically the stock install for a RasPi (Raspbian)? It's basically a full version of linux for ARM processors and is the basic/tutorial OS. It's also what I'll be using to run the RasPi cameras with in the Falcon.

Originally Posted by Ziggo You need the Fets to run these types of thing with the audrino boards, but it works really well. A friend of mine converted his lawnmower to fuel injection using one.

I was under the impression that FETs and MOSFETs are largely interchangeable terms now... It's just that there are a huge number of functional variants.

[Roddiy]

Originally Posted by Enki I've already done PWM to dim an LED...not that hard really. So, basically the stock install for a RasPi (Raspbian)? It's basically a full version of linux for ARM processors and is the basic/tutorial OS. It's also what I'll be using to run the RasPi cameras with in the Falcon. I was under the impression that FETs and MOSFETs are largely interchangeable terms now... It's just that there are a huge number of functional variants.

Well you gotta take into account that PWM on a 1.5v LED =/= PWM on a 12v pump.

an LED can be hooked up directly to the Arduino, try hooking up a 12v 1A+ pump to the Arduino... yo shit's gonna blow.

I'm curious as to what amperage the pump is rated at under maximum load (ramping up the pressure) because that will affect what motor controller (IC) you'll have to use.


Have some fun doing PWM control on a DC motor first, the circuitry is largerly similar if not exactly the same as you would use to control the pump.


EDIT: I've used and have a couple of SN75441 ICs, but an L293d/L298 IC should work just as well, given the pump doesn't take more than 2A max




as for the RPI... Yes the raspberry stock OS should do fine, but the program still has to be written


Processing would be a very simple but quick way to get it done, since it's very similar to arduino language and very easy to communicate between the two.

[Enki]

Originally Posted by Roddiy Well you gotta take into account that PWM on a 1.5v LED =/= PWM on a 12v pump. an LED can be hooked up directly to the Arduino, try hooking up a 12v 1A+ pump to the Arduino... yo shit's gonna blow. I'm curious as to what amperage the pump is rated at under maximum load (ramping up the pressure) because that will affect what motor controller (IC) you'll have to use. Have some fun doing PWM control on a DC motor first, the circuitry is largerly similar if not exactly the same as you would use to control the pump. EDIT: I've used and have a couple of SN75441 ICs, but an L293d/L298 IC should work just as well, given the pump doesn't take more than 2A max

I'd be doing PWM through logic level MOSFETs or similar, like what I've already done with the air ride testing. I'm no dummy; the logic level MOSFETs I have (and I have a bunch of them) support upwards of 30 amps.

In fact, the PWM testing I did was not with a single LED, but with an LED matrix which required in excess of 9 volts to power *without* resistors, and was driven via one of these MOSFETs.

[Roddiy]

Originally Posted by Enki I'd be doing PWM through logic level MOSFETs or similar, like what I've already done with the air ride testing. I'm no dummy; the logic level MOSFETs I have (and I have a bunch of them) support upwards of 30 amps. In fact, the PWM testing I did was not with a single LED, but with an LED matrix which required in excess of 9 volts to power *without* resistors, and was driven via one of these MOSFETs.

Ahhh okay, sounded like you had done one LED straight from the Arduino hahah.



In that case you're good, but I gotta wonder, what kind of resolution do you get using PWM through a logic level MOSFET? the full 0-1023 offered by the Arduino's analog pins or do you have to use the map function to reduce that a bit?

[Enki]

Originally Posted by Roddiy Ahhh okay, sounded like you had done one LED straight from the Arduino hahah. In that case you're good, but I gotta wonder, what kind of resolution do you get using PWM through a logic level MOSFET? the full 0-1023 offered by the Arduino's analog pins or do you have to use the map function to reduce that a bit?

The resolution seemed quite good; I was able to make the entire array barely glow at the LED source. The datasheet for it is here: https://www.sparkfun.com/products/10213
The listed switching time is measured as less than 500 nanoseconds, so it should be viable for full resolution Arduino control.

Edit:
The threshold voltage is 2.5 volts so I think I might be overdriving it a bit; I should probably look into what the required full on voltage is vs threshold voltage.

[Roddiy]

Originally Posted by Enki The resolution seemed quite good; I was able to make the entire array barely glow at the LED source. The datasheet for it is here: https://www.sparkfun.com/products/10213 The listed switching time is measured as less than 500 nanoseconds, so it should be viable for full resolution Arduino control. Edit: The threshold voltage is 2.5 volts so I think I might be overdriving it a bit; I should probably look into what the required full on voltage is vs threshold voltage.

This is exactly the one that came up on my search hahah.


IIRC reading the comments the full on voltage is 5V+

[Enki]

Originally Posted by Roddiy IIRC reading the comments the full on voltage is 5V+

Looking at the datasheets, less than 3 volts is needed to supply 10 amps, with about 7 volts needed to flow 100 amps. 5v should be enough to supply full rated amperage with diminishing returns starting in below 4v.

[Lex]

Originally Posted by sidekick I just looked up some arduino methanol controller builds and that would be awesome. If only someone would make a kit... @Lex;

Working on it sir. But you have to remember that Arduino is not the best for reliability and longevity in the automotive environment. On top of that it has too many features for such a project.

[Enki]

It does have a couple things going for it though: availability, price, and ease of use.

Once you have working code, even if you burn up one of them, you can get another for under $20 (teensy) and re-load the code.

[sidekick]

Originally Posted by Lex Working on it sir. But you have to remember that Arduino is not the best for reliability and longevity in the automotive environment. On top of that it has too many features for such a project.

Hey, if you're making a legit meth controller, I don't even care what it's made with. Lol. I'm sure it will be much better than some of the current offerings. I like the idea of being able to control the meth with more than an on/off switch.

[Enki]

I can easily rock some open source community code, but I don't have the hardware to play with yet (meth parts) to actually make something.

[Chimmike]

curious what you're using camera wise to go with the raspi's for video. I picked up a raspi and no zero coding but I'd like to set up forward/rearward video capability.

[Enki]

I'm going to use Raspi cameras. I'll have at minimum four Raspis on board configured to automatically boot to console and start the camera app (that comes with Raspbian).

The command line for that looks like this:

To run the camera software: raspivid -t 86400000 -w 1280 -h 720 -b 2500000 -o - > /mnt/DriverSide/Driver_$( date '+%Y-%m-%d_%H-%M-%S' ).h264 This will run it on the primary montior as well as save the 720p output to a file on a network share at 2.5 MB/s, using a file name with the date and time of run (for easy management).

More info: Enkiworks Motorsports

[Chimmike]

well, I have a Pi with the camera, I just wasn't under the impression that the quality will be good enough for use in that manner.

[Enki]

Originally Posted by gingermike well, I have a Pi with the camera, I just wasn't under the impression that the quality will be good enough for use in that manner.

Watch this in full screen and let me know what you think:

Also, keep in mind this is compressed compared to what I actually saw on the LCD (which was beautiful and very much real time).

[Chimmike]

Video quality is great.......................but that scenery is OOOOGLY! haha.

[Enki]

Yeah, well, desert. The video doesn't do justice what I was able to make out on the screen, and the screen isn't even 720p.

[Chimmike]

Wish I was skilled enough to do some of this shit.

[Enki]

To be fair, I didn't know how to do ANY of this shit when I started.

All it really takes is determination and Google.

[Chimmike]

you have a programming background?

[Enki]

Nope.

I've learned what I know on my own. The only classes/certifications I have are MCP on Win NT4/2kpro and an A+ cert (which as we all know is a joke).

[ms34m3]

Here is the progressive water\meth controller i made for my devils own kit
- arduino pro mini 5V
- freescale semi MXP series pressure sensor
- logic level power mosfets (IRL540 with heatsink) for solenoid and pump, work great because of their relatively low duty cycle
- RGB led output (for system status) mounted in a-pillar
- float switch input (I have a coolingmist tank that had a float switch installed)
- Aux analog input. I have a labonte motorsports flow controller that outputs 0-5v according to flow. I used this as another alarm condition, indication of bad pump or clogged nozzle

my code is laughable, but it works. If other people here know how to program, I will strip out my progressive control part and just make a file of helper functions to read\write the ports, set LED color, etc etc.. that way it can be duplicated easily. I'll throw it on github if there is any interest in contributing

In the attached pictures, I use a sparkfun FTDI board for debug. Thats what that crazy header thing i soldered on was for. I used 7809 voltage regulator to the RAW pin on arduino, which uses its onboard 5v LDO reg. Peripherals are then run off the VCC pin. I find multiple regulators smooth out cruddy automotive DC power.

- to do: automatically pull timing when error condition is present with controller. Perhaps this is a relay output, or intercepting and spoofing crank position sensor.
- add bluetooth so it integrates easier with android
- i started laying out a custom board, possibly finish it and upload to OSHPark to make a standardized open source water meth controller

[timjs]

That is just fantastic!