loudhvx, look inside

PS - I should have been an electrical engineer, I love this stuff so much.

I didn't realize you meant the Arduino is the entire datalogger.

T_Dub wrote: The arduino has a 20MHz clock, so it should be fine, as long as by short you don't mean less than one twenty millionth of a second.

I measured a digital read to take about 4 usec but occasionally up to 12 usec due to background interrupts. But the time between polling will even be considerably longer than that since you will have the arduino carrying out many other instructions. They add up quickly and really slow things down. At 10,000 RPM, you'll likely have to poll for tach signal at least once every 80 usec or so to avoid missing any.

An external D-to-A might reduce the processor load (pulse interrupts, calculating RPM from elapsed time, etc) since you really don't care about the precision of the ignition pulses to just datalog the RPM.

You're right of course. Maybe I should have gotten something with a bit more oomph.

I wonder if I can get around it. Hmm. Once the RFID is read, that instruction set is over, and it only polls one bit after that to see if I shut off the bike. So the only code after that will be to poll that, read AFR, MAP, and RPM. Maybe I can alternate, like read AFR, read RPM, read MAP, read RPM, poll shutdown bit, read RPM, repeat. There's also a couple interrupts built in I can use.

I'll see how it works when I get it set up, you never get this shit right the first time, not in my experience anyways. Always requires some tweaking.

T_Dub wrote: I'll see how it works when I get it set up, you never get this shit right the first time, not in my experience anyways. Always requires some tweaking.

Heh, don't I know it.

I think you'll want throttle position in there too. With those CV's I think you're going to have to find some correlation between MAP and TP to sort of figure how the slide is moving.

Here's a cool CV video on a later GPZ550. I hope it shows up. It really shows what's going on with the slide while actually running. You can hear the motor and see the throttle linkage to get an idea of what the rider is doing. Kudos to Glbishop.

s1219.photobucket.com/albums/dd427/glbis...rrent=0820111114.mp4

Thats a cool video! Really shows what's going on well. What I'd really like to have is an ultrasonic sensor mounted inside the carb, measuring it rising and falling. But that would be expensive. I'll probably mount the MAP at first, see if I can use it to tune alone, I think with the raw data and a first order differential I might be able to get some good idea of what to do. I once tuned an EFI engine to run on just MAP and TPS, that was fun, no AFR is a bitch. I know I'll probably want TPS, but its gonna be a pain to mount so I'll try first without it.

So we've changed our minds on the board to use now, going to switch to a .NET based system. About the same cost and you can get a 72MHz processor, still open source too.

Only problem is this board likes 3.3V digital inputs. Think there'd be any problem just making a voltage divider to step down the signal after that circuit you made for my tach?

Allow me to clarify my question. What I should have asked is where on the intake tract are you going to take the vacuum signal from? I had planned on connecting a vacuum hose to each intake boot and then connecting them together through a vacuum tee to get a conglomerate vacuum signal. I'm wondering how this would have worked out. I'm still not sure if the signal would have been smooth enough to take meaningful data while looking at a log file. If you look at an individual cylinder, it pulses wildly. I guess that you could dampen the pulses with a restriction in the vacuum line, but that would tend to slow the response time, no? I still might try and hook mine up, but at some point I became interested in riding the bike, not gathering data. I think that yours would be very informative if you could get both a MAP reading and some sort of linear sensor to determine actual slide position.

Problem with any linear sensor is it'll interfere with slide motion. The slides on these are pretty sensitive since they're vacuum operated. I don't want to make one lift less because I stuck a linear pot on it.

All the sportbikes use something like that. I think they connect 1&4, 2&3, tee them off and then into a third tee or something.

Yeah, I usually design as much as I can in hardware since I'm not much of a programmer at all. For me, the Arduino is good because I can do the hard stuff in hardware, then use the Arduino to just read time-insensitive information and store it.

I'm sure that if you can get a reliable measurement of vacuum, you can correlate that to slide height anyway.

Did I mention that I'm glad you are back posting? When I did my thread about logging AFR data, I think everyone except loudhvx thought I was a quack so I gave up on it. I tell you one thing, I learned a lot about my bike. Even some stuff that I didn't want to know.

Yeah I understand that. I'd rather program away my problems haha. These boards have some really neat features. I'm gonna set up mine so the kill switch flips a digital input to toggle the controller in and out of hibernation. When it wakes up it'll look for the RFID, and when it finds it it'll power a relay to send power to the bike. I'm gonna use a MOSFET to power the relay. Also after it finds the RFID it'll come out of a loop and start logging sensor data. And when I flip the kill switch again the board will finish its log file and hibernate. It'll run direct off battery power, thats why the hibernation. When its hibernating it uses very very little power.

I just had a look at the ninja parts diagrams, looks like they just tee all the cylinders into a common hose and read the pressure form there. Just have to keep the hoses as short as possible to get the best response.

testarossa wrote: I'm sure that if you can get a reliable measurement of vacuum, you can correlate that to slide height anyway.

Did I mention that I'm glad you are back posting? When I did my thread about logging AFR data, I think everyone except loudhvx thought I was a quack so I gave up on it. I tell you one thing, I learned a lot about my bike. Even some stuff that I didn't want to know.

Yeah thats what I figured, I should be able to correlate MAP to slide height fairly well.

I'm glad to be back haha, happy there are people around who don't think I'm a quack too. In my mind, the only way to tune any engine is to get hard data. Spark plug reading is way too mystical for me. I'd much rather look at a spreadsheet and say "Hey, I'm low on fuel between 3000 and 3400RPM while accelerating, better fiddle with my jets."