what do you need measured

[aff-man]

Right project time guys. Now i don't know exactley how this project is gonna work but here is my idea. I wish to build an onboard system that monitors a bike while racing. i.e. lap times top speed ect. But in order to put this idea forward to my lecturers i am gonna need to have a design brief sorta worked out. So here is the hum dinger, i need to know what to measure. I have lap times and top speed, but what else is there????? any idea's welcome maybe i can find out how to make cheap ones for all the budding racers out there. :niceone:

[Two Smoker]

Definately average speed, as well as top speed and lap time.... hmmmm dont think i would need much else, oh storage of lap times would be good, and a number next to the time representing the lap.....

Count me in for one :niceone:

[merv]

The GP guys measure just about everything don't they - speed at all times and then the actual bike operational stuff like gear your in, braking, temperatures and all that stuff - so really take your pick how far you think you can go.

[dangerous]

How about the amount of time that the frount wheel is off the ground and for how far :whistle: :rolleyes:

[wkid_one]

Shit you could go absolutely mad.....when braking applied, brake force, suspension load, engine revs, temp, gears, front v rear brake useage, throttle application, oil pressure, fuel consumption, max top speed, av speed, tyre temp, tyre pressure, weight distribution etc etc etd

[Two Smoker]

Shit you could go absolutely mad.....when braking applied, brake force, suspension load, engine revs, temp, gears, front v rear brake useage, throttle application, oil pressure, fuel consumption, max top speed, av speed, tyre temp, tyre pressure, weight distribution etc etc etd
Whilst he could do that, i think for cost and simplicity the ones i stated would be good..... Ofcourse i would love to know section times and braking points etc.... but $$$$

[scroter]

try performance bikes magazine[uk] they often datalog there bikes for all sorts of things,dont tell you how its done but!!! :done:

[Ghost Lemur]

It all depends on how indepth you want to get. As dangerous said the GP guys measure everything realtime. So you can see speeds throughout the circuit, temps, pressures, rev count, etc.

The more indepth it is the more useful it will be, but on the downside the more work you'll have to do.

[Kickaha]

My priority for a basic system would be a dual display

1 Lap time
2 store lap times so you can recal and scroll through
3 display what lap is fastest time
4 Average speed
5 split times,using more than one beacon
6 Temp,with max temp recall
7 max speed,it's a bit of a wank,but good for seeing if your speed onto the straight is getting better(lap time is far more important)

check here to see what other systems offer
http://www.aim-sportline.com/
http://www.piresearch.com/

[geoffm]

Right project time guys. Now i don't know exactley how this project is gonna work but here is my idea. I wish to build an onboard system that monitors a bike while racing. i.e. lap times top speed ect. But in order to put this idea forward to my lecturers i am gonna need to have a design brief sorta worked out. So here is the hum dinger, i need to know what to measure. I have lap times and top speed, but what else is there????? any idea's welcome maybe i can find out how to make cheap ones for all the budding racers out there. :niceone:

My Datamite (www.performancetrends.com) will pretty much measure whatever you want, depending on the sensors and the volume of cubic dollars you want to put into it. Generally, engine rpm, rear wheel rpm, front wheel rpm, distance (from those), temperatures, fork and shock travel with a linear potentiometer, etc.
Laptimes via Hotlap or similar IR sender and reciever unit.
Download the software (demo) from their site for ideas
Get a book called "data power - racecar data aquisition" - has a lot of this stuff.
There used to be a data aquisition mailing list, but it seems to have fizzed.
Check the gokart guys - they do a lot of this stuff.
Geoff

[geoffm]

You also want to measure speed at all points around the track and overlay laps. Lets you know if you really are going faster or not, and where.

You may also want to hook up exhaust gas temperature probes (k type thermocouples) for jetting, although this is usually done on with live readouts.
Geoff

[aff-man]

Cheers guys that's heaps of info. The unit i was thinking of making would be a plug in system that you plug in the bike and measures stuff i.e. lap times(from various sensors to get times at different possitions), top speed and i'll probably add engine temp, gearing and some other fancy stuff. And then gonna write a comp program that basically you detach the unit from the bike and via a serial link you can download data and plot graphs ect of stuff. Just an idea but i think a cheapish device could be benificial as well as interesting to work on. Now i just gotta convince the lecturers oh well wish me luck.

[Hoon]

Yeah I briefly entertained the idea of designing a bike data logger. I originally tried a microcontroller but it ran out of memory. I now reckon using something like an ipaq to drive the sensors and using its built in wireless to connect back to the pits is the way to go. You could even mount a webcam!

But yeah the stuff I'm interested in is:

engine rpm
rear wheel rpm
front wheel rpm
front suspension travel
rear suspension travel
throttle position
front brake
rear brake
handle bar position
temperature
o2 sensor

Also some sort of mapping would be cool too!

Hey geoffm, that datamite is seriously impressive looking! Do you get all the sensors with it too?

[geoffm]

Hey geoffm, that datamite is seriously impressive looking! Do you get all the sensors with it too?

I got the 4 channel Datamite 1 - the 32 channel Datamite 2 would have been nice, but an overkill for what I want.
I upgraded to the 512K memory when I got it, which should give 20-30 mins of data recording, depending on the sampling rate
I got the standard kit-o-bits, which includes a rotation sensor (a hall effect sensor and magnet) for my dyno drum or wheel, and it comes with a wrap a wire around the lead inductive rpm pickup. I also got the clamp on inductive pickup (as it is easier on the dyno) and optical isolator for the computer, so I can see real time RPM readouts.
Most of the other sensors should be pretty straightfoward if you have a solid electronics background. The manual covers it pretty well and is worth a read.
I did get a spare loom for on the bike, but haven't actually used it yet - no time, no money and no working bikes...
Geoff

[geoffm]

Cheers guys that's heaps of info. The unit i was thinking of making would be a plug in system that you plug in the bike and measures stuff i.e. lap times(from various sensors to get times at different possitions), top speed and i'll probably add engine temp, gearing and some other fancy stuff. And then gonna write a comp program that basically you detach the unit from the bike and via a serial link you can download data and plot graphs etc of stuff. Just an idea but i think a cheapish device could be benificial as well as interesting to work on. Now i just gotta convince the lecturers oh well wish me luck.
Go for it if you can get it to go. Have a look at the Datamite software for ideas. Do a version for running the bike on a dyno as well, and I will help you test it out.
Have a look at using a BasicStamp - a guy in Aussie (just out of Melbourne) has used one for his dyno, however I don't know what the limitations are for storage. Easy to program.
Most cheap aquisition systems are made to read directly into the PC and so only have a small storage buffer.
My Datamite has 512k, and they say it is good for 20-30mins, depending on the channels used and sampling rate.
Geoff

[aff-man]

That data stamp is kinda stooooopid. Not a lot of memory limited capabilities bad interface ect ect ect. Just use stuff at tech and use delphi for the program

[Warren]

If you want to overlap info from different laps you really need something like GPS.

[Hoon]

If you want to overlap info from different laps you really need something like GPS.

I looked into GPS as a means to track but it wasn't up to the task. Other than the costs, you need something that can sample 10 times a sec and it also has to be accurate to 10 cms to get any worth while data.

[Quasievil]

Hi this is made by Beltronics, does it help you with any ideas ???

http://www.beltronics.com/fx.html (http://)

[Motu]

Here's another one...

http://www.roadtune.co.uk/

[Redstar]

One real simple one, Distance travelled, not the lap length as I guess thats taken as the center line of the lap? but what the bike actually travelled,
that would have to be a measure of the quality of the line chosen? and easy to measure?to the mm

[geoffm]

If you want to overlap info from different laps you really need something like GPS.
no, you have the velocity at various distances. Knowing the
distance around the track, you can zero it. Use the slowest point (hairpin at Puke eg) as the zero, with an offset if you want to set it to the start finishe straight. You can pick up the slow point easily enough on the speed/ distance chart, and then break it into your laps. Good software may even be able to do it automatically.
Another, better way is to use an IR sender as per laptimers, and use this as a trigger on one channel.
Geoff

[aff-man]

yip plan on it will find out tomorrow if my idea has substance from the lecturer

[FzerozeroT]

hey you fief :Pokey:

I've been thinking about this for my third year project as well! I was thinking about gyros on swingarm and tail section then creating a data system to analyse changes in rebound/damping etc. The challenge is on! :D

[bgd]

How about rider performance as well - heart beat, blood pressure, etc ;)

[aff-man]

got the big OK from lecturer. But he said i would have to simulate bike movement using lab inputs because it would take to long to set up the sensors on a big. Well i will do as he says but i might spend some dosh and design/build/mount sensors on my bike anyway and use the programmed processor. To see if it works practically with no thought of personal gain :pinch: ;) . hahahahaha so let the planning begin

[Drunken Monkey]

Good luck!

Not trying to be a nay-sayer or doom-sayer, but I find it hard to see how someone will be able to make a low volume of electronic metering equipment at lower cost than any existing production units out there. Granted you'll hopefully save some of the normal development costs as part of your project, but production costs for low volume will probably end up putting your unit at around the $500+ mark anyway...

If you do manage to sort it out for around $200 a unit, then count me in!

[Drunken Monkey]

While on the subject - I also run an Escort accelerometer unit. Measures acceleration, braking, turning force, etc...in realtime. Gives you a 10 run history on 1/4 mile, 1/8 mile, 60ft, 330ft, 660ft, 1000ft, top speed on run. All that can be had for $500 (probably less if you go for a different brand).
Doesn't work too well on the m/bike if you wheelie or stoppie, and the cornering data is naff from the tipping.

[Hitcher]

And, last but not least, a little arrow showing which way is up!

[jrandom]

simulate bike movement using lab inputs

... visions of students running around the room waving accelerometers going "brrrrm! brrrrrrrrm!"

[Drunken Monkey]

... visions of students running around the room waving accelerometers going "brrrrm! brrrrrrrrm!"

nah, they're 2 smokers..."nnniiiiiiiiiiaaarrrr, reninininin, nnnniiiiiiiaaarrrrr....."

[aff-man]

hahahahahahaha. As to the cost should be using a pic processor so they not to expensive. I have to have a cost plan made out by the end of next week. So i'll post the final cost of production then. As to the question of space i will see if i can interface it with a flash drive/ removable hard drive which will give me ooooooooooodles of space to mess with. hmmmmmmmmm. the devices used for measuring stuff can either be bought or made, the latter might be cheaper so i'll see.

[Drunken Monkey]

oooh! That's piqued my interest.

Removeable CF or SDIO would be better than a download via USB/Serial in my books!

[aff-man]

was having a chat about it today with a guy who has used pic's and a removable hard drive and it would be easier to handel but the coding to make sure all the info got stored right and retrieved might take some planning hmmmmmmm. Lecturer said i could use a flash drive. Would make comp program easier as well. oh well we'll see. Co-incidentally i am also doing a bit of research into making just a standardized lap timer so if there is anyone out there who is in the electronics industry who can recommend a chip that has +- 256 bits eeprom, can run at say hmmmmm 2+ Mhz has 2 ports and hmmm that's about it.

[jrandom]

anyone out there who is in the electronics industry who can recommend a chip that has +- 256 bits eeprom, can run at say hmmmmm 2+ Mhz has 2 ports and hmmm that's about it.

You want a Mitsubishi M16C with integrated flash and RAM, you do. Cheap as chips and you can clock them up to something over 10MHz. The bespoke embedded electronic designer's standby.

It has a 12-bit A2D module as well IIRC, I'm sure you could come up with something interesting to do with that in a bike datalogger.

CF and SD cards can easily be bitbashed out the CPU's data bus with a few extra I/O lines. I've done software-driven interfaces for both in the past.

[geoffm]

got the big OK from lecturer. But he said i would have to simulate bike movement using lab inputs because it would take to long to set up the sensors on a big. Well i will do as he says but i might spend some dosh and design/build/mount sensors on my bike anyway and use the programmed processor. To see if it works practically with no thought of personal gain :pinch: ;) . hahahahaha so let the planning begin

A tech student did a dyno project some years ago, before going to foreign parts as a missionary or something. HE gave me his setup, including a test rig and source code (?). No use to me (didn't have the skills to finish it off properly) but if you want it, its yours. To good to throw out...
It is in the attic at home, so either wait until i am home agian (mid term) or I will get the missus to dig it out. Let me know.

[aff-man]

hmmmmm maybe with a few brains in here we cam get it up and running and use it to the benifit of all KBr's. It would be cool to have a dyno day that didn't cost anything. I'll be willing to have a look but with current work load a little wait might be in order.

As to the mitsubishi well it might be over kill but i'll see.

[Eddieb]

Have a look at this, 190 pounds sterling (~$550NZ or so)

http://www.veypor.co.uk/product.html

[aff-man]

Hmm interesting. I have run into a snag though. The low level code that is needed for file sorting on the usb drive will take ages to work out. So if anyone has an idea how to do it or a site that explains it then whoo-hoo we are ready to go.

BTW i can probably get a lap timer with LCD module for under a hundy from tech will have to see though.

[geoffm]

hmmmmm maybe with a few brains in here we cam get it up and running and use it to the benifit of all KBr's. It would be cool to have a dyno day that didn't cost anything. I'll be willing to have a look but with current work load a little wait might be in order.

As to the mitsubishi well it might be over kill but i'll see.

Well, I have a (homemade) dyno, and when I am back in Auckland for longer than a single weekend, and have sorted out the power supply problem and tweaked the calibration, will put out a general invite. I will have to come up with a bulltproof disclaimer - having lots of bikes through the machine would be good for my testing, but no way am I going to wear the cost if something grenades.
Watch this space - probably towards December.
Geoff

[Motoracer]

BTW i can probably get a lap timer with LCD module for under a hundy from tech will have to see though.

One that is set off by IR or manually?

[wkid_one]

Hey - talk to Jimbo - it appears he could need a hand measuring things - he seems to have a liberal interpretation of length and height

[jrandom]

Hmm interesting. I have run into a snag though. The low level code that is needed for file sorting on the usb drive will take ages to work out. So if anyone has an idea how to do it or a site that explains it then whoo-hoo we are ready to go.

Eh? What exactly are you talking about? A FAT implementation, or something? Or do you want to implement a USB master controller so that you can use generic USB storage devices for output?

Be more specific, I can probably get you what you need.

[aff-man]

One that is set off by IR or manually?

IR

As for the code i need to basiclly get hold of the c code that does the file sorting in the usb (basically the driver) i can get more info but the problem is, i need something that is effective but quite small so it can fit on the pic controller. If i can't then i will use a flash ram extension through one of the ports

[jrandom]

i need to basiclly get hold of the c code that does the file sorting in the usb (basically the driver)

I still have no idea what you're talking about.

What is the usb of which you speak?

[aff-man]

Ahhhh what i am talking about is a usb flash drive. The code i am looking for is the one that is on the micro that when getting the info stores it in a nice "filing system" so when plugged into a comp it can easily be read and recognised.

[jrandom]

Ahhhh what i am talking about is a usb flash drive. The code i am looking for is the one that is on the micro that when getting the info stores it in a nice "filing system" so when plugged into a comp it can easily be read and recognised.

Aha, right. Like one of those key-ring drives, you mean.

Do you have the USB host/master protocol stack implementation sorted, then? That's a pretty heavyweight thing to put on a PIC! In fact I don't know of *any* products that use software USB hosts. You'd really need an external USB host controller slaved off your micro.

And after you've sorted *that* out you'll need a FAT16/32 implementation, which is your 'filing system'. The USB storage devices present a sector-level interface, similar to ATA hard disks and the like, which you need to plug in at the device-driver layer of your filesystem. You're talking a *lot* of code there, DEFINITELY more than you could do yourself as a project, so you'd need to find a GPL'd implementation that you could plug in, etc.

I should really make the point that implementing a USB host to run devices that present the standard USB mass storage interface seems like far too much work for this project of yours. In fact you would do very well to avoid USB altogether, it's a minefield of design-by-committee protocols and nigh on impossible to get working well in an embedded system from scratch.

I strongly suggest you do an SD/MMC card interface instead, that way you still have interoperability with desktop computers, etc, but it'll be an order of magnitude easier to implement. You'll still have to find your FAT16/32 source code though.

[Ghost Lemur]

IIRC - didn't one of the OSS groups release an open source verison of FAT?

I know MS's paitent over FAT is currently being reviewed and should hopefully be voided. Will make things like this a hell of a lot easier to impliment.

[jrandom]

Google for FreeDOS.

Or you could rip the FAT management code out of the relevant modules in the Linux kernel.

Keep in mind the requirements of the relevant open-source/free-as-in-speech licenses - redistribution, copyleft etc.