Super MPGuino?

[gtkid2002]

it should be on pin 14 but i'm wondering if it isnt just frozen. the buttons still dont work though. Any ideas on what could mess with that?

If Pin 14 is Analog 0, then it yep. Should I try moving pins?

[nickdigger]

Try probing the voltage at different points in the vss circuit. I.E., roll the car in the driveway until vss hits 5v, then park it. Then measure V at both sides of your guino's vss resistor.

If you're using 100k ohm as per the instructions, your vss input may be getting reduced too low to trigger the input. It needs to flip above+below 2.5v or whatever, for the duino to detect movement. My first run in my honda got no vss, and i had to replace the 100k with a smaller one. It's currently 3.3k, which i think is a tad too low (i meant to use 33k, but grabbed the wrong one in the dark). 1000ohms was really too low, as it interfered with the real speedo. I think 100k is fine for cars with 12v vss (keep 100k on your injector, btw), but our 5v hondas have different needs.

Any signal on your injectors? I.e., any gallons-per-hour readings?

[gtkid2002]

Well, I lowered the resistances, but I can't tell if it works or not. As soon as I do, it puts a load on the vss, and drops the voltage. So I can't seem to get it above 1.8v. The VSS reads 4.75v, so I'm tempted to make an amplifier to boost it. I can't get the Mega to read a single screen, because even though the buttons are disconnected, it keeps switching screens on me. It likes to go to the scratch screen (odometer or whatever) or it will go to the cpu usage screen. Occasionally it'll just start spewing loads of coded text.

I'm just gonna go out and say without code revision, I don't think that the standard MPGuino code is compatible with a Seeeduino Mega. Unless I just got a crappy 'duino or screen or something, I don't think it works. I might just buy an mpguino and see if that'll work, then upping to a 328 or something and try and go from there.

I haven't been able to hook up my injectors but I know that they work. I've got some DPDT toggles on cylinders 2 and 3 that use resistors to bypass the injectors, so I know that those work.

Although on the GPH readings on the duino, when -nothing- was hooked up aside from USB power, occasionally the GPH reading would just tweak out. It would just flash random numbers and stop when I poke the thing, which makes me wonder if it had a short or something.

[bobski]

Loose connections probably. The coded text thing is common when there's an intermittent connection between the 'guino and LCD module. The data from the 'guino gets corrupted, but the display is relatively dumb device... It just does what it's told whether it makes sense or not, thus the garbage on the screen.

Floating inputs (no pull-up or pull-down resistor) will often cause erratic behavior since it's not clear whether the input is high or low. Unrelated electrical activity can trigger floating inputs. It could be that the program's favorite screen is just the one that produces the least electrical noise on the button inputs, so it's less likely to get switched away from.
I've noticed people like to use "internal pull-up resistors" by switching on a digital I/O pin before using it as an input. I'm a bit suspicious of that technique simply because it's not clear what exactly that's doing inside the chip. It could be the architecture of the '1280 doesn't behave the same way as the '168/'368 (maybe due to the greater number of I/O pins?) which breaks (or requires modification of) that functionality.
I would try external pull-ups (a ~5K resistor T'd off each switch signal wire and connected to 5V) and see if that helps.

[gtkid2002]

I was thinking the architecture was different. I tried looking up schematics, but I can't find anything in a .pdf for mac usage. There really isn't much on the Seeeduino Mega, which makes me wonder if I should have just stuck with the Arduino mega.


Could you explain more as to what the T'd means?

What I think I got was I should run a resistor from 5v to analog 3,4, and 5 inputs, then have the switches go from the resistor lead to ground where they normally go. That about right?

I tried it, and I still can't get it to work. I'm using the contrast button as my guinea pig, and it still isn't working.

http://opengauge.googlecode.com/svn/...no/mpguino.pde Is that the current stock code?

[nickdigger]

What resistances did you try for your vss input?

As for your random GPH readings when not hooked up, that happens to me too. Readings would vary, depending on how i was holding it. The only way to stop that is to ground your INJ input, or hook it up to your car (effectively, grounding it).

[gtkid2002]

I ended up with 100k, 12k, and 1.2k. I'll have to find a slightly higher resistor for testing tomorrow, it's too late for it now.

Any suggestions on what to do about the buttons?

[bobski]

Originally Posted by gtkid2002

Could you explain more as to what the T'd means?

What I mean is, leave the switch connected the way it is, but strip some insulation off of the wire that connects the switch and its 'guino pin. Solder one lead of a resistor to that stripped off area. If you soldered the resistor on at a right angle to the wire, it should have a T shape, thus the term. Really it just means connecting to a wire without breaking the existing connection, also known as "tapping into" a wire.
Connect the remaining resistor lead to 5V+.

This will make the 'guino see 5V until the switch is pressed, at which point any power coming from the resistor or the 'guino itself will be shorted to ground through the switch. The resistor prevents this power flow from from becoming a true short by limiting the current. A 5K resistor limits current to 1 mA (see ohms law). You could use a higher value resistor to limit the current further, so long as the 'guino isn't sinking any significant current... Anything up to 100K would probably work, though I'm guessing really. Going to a lower value means more current would pass through the resistor and be dumped to ground (wasted) when the switch is pressed. There's no practical benefit to doing so, but you could go as low as 300 ohms before you start pushing the limits of a typical 1/8th watt resistor.

[nickdigger]

What were the resulting voltages for each resistance (100k,12k,1.2k)?

I just went back & read your earlier posts, and you say your speedo is dead? You have a mechanical speedo, which spits out a 0v/5v vss signal, right?

It seems weird that 1.2k would drop your signal down to 1.8v. I dont remember offhand what mine was at 1.0k, but it was well over 3v, i'm guessing 3.3 - 3.5v.

[gtkid2002]

Bobski- I put a 5.1k resistor (1.4W) into the same header pin as the switch, and still nothing. I also tried it the way you suggested, and nothing either. I know the switches work fine as I've tested them with my continuity setting on my multimeter, and it beeps when I press them. I also tried it with a 10k, and still nothing. I even did it with a 1k and I didn't get anything.


Nickdigger - My VSS seems to not want to put out 5v flat anyways. Closest I can get is 4.76v, and the low is 4.73v. It is a mechanical speedo, with an electronic VSS right behind the cluster. The part that moves the needle is mechanical, and that's what is broken, which is why I don't want to fix it, although it is getting more and more tempting. When I turn the wheel, it spits out a 0v/5v signal though.

100k- 1.73v
12k- 2.1v
1.2k- 2.193v or something similar.

It really is close to tripping the 'duino, but apparently not close enough.