Homegrown Digi... *ahem* sorry Nowhere Elysium!

[blade023]

I stole your project!

I got an Arduino ages ago for tinkering with, as I had a few things in mind that I could do with it, and it looked fun. I tried doing a few thing with it for a few weeks then promptly forgot about it, so I've decided it's high time I get on and make something interesting.

I had the idea to do something like a 'digital dash' for a bike ages ago, and Nowhere Elesium's thread on it got me really interested. I thought I'd start off fairly simple and work my way up from there.

So far I've hacked an 18 LED RPM gauge, and a crude 3 digit speedo with gear indicator. At present there's no input from the bike (a 2001 Yamaha fzs1000) but to simulate RPM I'm using a pot so I can test things out and get the code working how I want it. I'm pretty much done with the LED strip, it's now at a point where I'm happy with how it displays.

Anyway, enough of me rambling; here's a video showing what progress I've made. Let me know what y'all think!

https://youtu.be/ccAnfeS_dZA
____________________
Current: VFR800 2008 13.7k and rising rapidly!
Sorn: 2003 FZS1000 30k(ish) looking to sell.

[orac]

thats very nice, however you dont seem to have any resistors or driver transistors for the displays = i used a similar multiplexed design a while ago and needed transistor for the anodes as the IC was unable to sink the current from each display

also making a display work is easy, making it receive and interpret data get a bit more tricking, you will alos need the correct circuitry to tak the readng from the bike.

depending on what you already have on the bike you may have to use things like hal effect sensros and correct circit to get a digital reading
____________________
Current rides - 2016 Triumph Street Triple Rx, 1994 Suzuki Bandit 400 VM, TGB 204 Classic 125cc
"with nothing left to lose, there is everything to gain. It's not the size of the dog in the fight, it's the size of the fight in the dog"

[nowhere.elysium]

Hey, don't apologise to me; I'm far from the first person to have a crack at doing this sort of thing

Have you turfed up the necessary code to do the interrupt-driven stuff for the rev counter and the speedo? It's not quite as simple as attaching it to a potentiometer.

Anyway; here I am, trying to sound all authoritative, and I never actually finished the damn thing. Suppose I should get off my arse and address that, really.
____________________
'10 SV650SF, '83 GS650GT (it lives!), Questionable DIY dash project, 3D Printer project, Lasercutter project

[The Shaggy D.A.]

Bah, blocked at work.
____________________
Chances are quite high you are not in my Monkeysphere, and I don't care about you. Don't take it personally.
Currently : Royal Enfield 350 Meteor
Previously : CB100N > CB250RS > XJ900F > GT550 > GPZ750R/1000RX > AJS M16 > R100RT > Bullet 500 > CB500 > LS650P > Bullet Electra X & YBR125 > Bullet 350 "Superstar" & YBR125 Custom > Royal Enfield Classic 500 Despatch Limited Edition (28 of 200) & CB Two-Fifty Nighthawk > ER5

[P.addy]

Secretly, nowhere is printing a hammer to beat you with.

[blade023]

[blade023]

Small update:

In the interests of not letting the magic smoke escape* from my beloved arduino mega (2560), I've been looking in to current draw. I worked out that with all LEDs lit at the same time (25), the draw is 183.8235 milliamps. This is very close to the mega's maximum limit.

(It's taken me a while to get round to all this calculating as it's more fun to just make stuff work! I just scared myself doing some of the calculations, which resulted in a value greater than 200mA by a fair margin, then realised that the lcd display is multiplexed, so only 1 digit is lit at once. A difference of 46 leds, and 25 leds all lit at once! Panic over..)

In a real world situation it's very unlikely that all LEDs will be lit at once, and if they are it will only be for a short time. As an estimate, in 6th gear, at 88 mph (assuming this is at about 10k rpm which I'm not sure about), that's 25 led's lit. Each LED will be drawing about 0.007352941 amps or 7.35 milliamps, which totals 183.8235mA. If I were to add anything else output wise, I'd need to seperate the mega's power from the output device's power supplies.

Still, knowing that I'm within the current supply limits is reassuring, and it also means I now know that I'd need to sort out transistors for anything more that I'd want to add. I have various ideas for extras once I've got the main parts working.

I've also ordered some proper value resistors for the leds (680ohm if anyone's curious), a frequency to voltage chip (for the rpm reading, seems the easiest way) and a hall effect sensor. Nowhere might remember from my posts in his thread that I was talking about using a reed switch and was told hall effect would be better.

*Magic smoke: stuff that makes the arduino 'go'. If this escapes then the arduino stops 'go'ing...
____________________
Current: VFR800 2008 13.7k and rising rapidly!
Sorn: 2003 FZS1000 30k(ish) looking to sell.

[mysterious_rider]

Really want that light but in red, and have it flashing side to side under my headlight like that 'kit' car with hasslehoff.
____________________
KAWASAKIIIII

[nowhere.elysium]

[blade023]

[blade023]

Nowhere; The simplest method of reading the rpm, is not to use interrupts at all, but use a frequency to voltage converter. that way you don't have to mess about with time calculations or whatnot, you just have an analog input measuring the voltage, and setting how that voltage range corresponds to the rpm. About 2 lines of code. I've ordered a f-v converter form RS Components (I highly recommend them for electronics btw) and here is the datasheet for it if you're interested. It's the LM2917M-8/NOPB, the datasheet is for 2 models.

Off the top of my head, I'd then do..

{code}
// the first line here reads the voltage from the f-v converter, the second line converts it to our rpm range
rpmVoltage = analogRead(rpmInputPin);
rpmActual = rpmVoltage / cleverNumber;
{/code}

To me that seems infinately simpler and more elegant than messing about with a second interrupt! There's already going to be one interrupt for the hall effect sensor mounted near the wheel. This will detect wheel rpm and convert to actual speed.

In fact, thinking about it, I may at some point get another of these f-v converters and hook that up to the speedo to read that in the same way. No interrupts at all! and there's no chance of (due to timing issues) a discrepancy in the rpm or speed readings. As long as the voltage converts to rpm properly.

There's just one thing that I need to get my head around. There was me thinking I'd hook up speedo + to f-v +input, speedo - to f-v -input... then hook up the f-v g - arduino g, and f-v output + to arduino analog input.....

https://www.kenhawley.co.uk/arduino1.png

Seems it not so simple. Looking at the datasheet there are examples of how to wire it up, and a couple of them are relevant to rpm reading. But until I take the time to read through it and figure out what I need to wire up where, I'm a tad confused. I've never really got into wiring diagrams, so I'm a little in the dark. Hah, and you think school is over when you leave! I'd have paid more attention if I'd have known how useful all this sheet is..

[Xenocide]

Can't you just use the arduino's analogue/PWM inputs or do they not offer enough range?

You could try taking a feed off the sweep feed going to the servo thingy in the bike's tach housing rather than the signal that drives it, it must already be being converted by that point to get a stable signal for the needle.
____________________
KTM LC4 400 | Honda VTR1000

[blade023]

Using the pwm pins (which I hadn't thought of) I would be relying on the arduino's timing for an accurate reading. It's doable but would be better with a 555 timer chip The rpm needle itself from what I've managed to find out is connected to a servo which has a frequency signal to tell it how far to move across. This is in mechanical rpm gauges and makes sense as it's the easiest and apparently cheapest way to do it. That's the signal I'll be using. afaik there's no conversion between freq sig and rpm needle moving, unless you know different? I'm yet to implement so I'm not certain yet.
____________________
Current: VFR800 2008 13.7k and rising rapidly!
Sorn: 2003 FZS1000 30k(ish) looking to sell.

[orac]

i am not sure why you are thinking of using analoge, take a feed from either the coil feed or an induction coil around an HT leed, there will be only 0.05 seconds per rev at an idle of 1200 rpm. depending on firing order a simple count command once the singal has been suitably cleaned.

it maybe worth having a small slave circuit to monitor the data.

i have been thinking about a similar project and would have a the RPM and speed taken by a small slaves and allow the master to take the data to over the i2c bus. that way the master can control the display/s and not having to do things like distance time calculations which take time.
____________________
Current rides - 2016 Triumph Street Triple Rx, 1994 Suzuki Bandit 400 VM, TGB 204 Classic 125cc
"with nothing left to lose, there is everything to gain. It's not the size of the dog in the fight, it's the size of the fight in the dog"

[RhynoCZ]

Looks great so far, but I can't really imagine this being on a actual motorcycle, vibes will ruin the thing in a few seconds
____________________
'87 Honda XBR 500, '96 Kawasaki ZX7R P1, '90 Honda CB-1, '88 Kawasaki GPz550, MZ 150 ETZ
'95 Mercedes-Benz w202 C200 CGI, '98 Mercedes-Benz w210 E200 Kompressor

[The Shaggy D.A.]

[blade023]

orac: The reason I was thinking of using analogue was that it's constant regardless of timing discrepancies. I could use a 555 timer with the arduino I guess, which would solve that. It also means the code is much simpler, as it's a simple analogRead(tachPin) function followed by a conversion from V to RPM instead of a whole mess of code. There are several ways of doing it though, and I've been thinking about a few. One thing I would like to do in the future (especially having seen the price of the tiny arduino boards without pins that look ideal for small tasks and wiring up to pcbs..) is as you said; let slave chips handle things like counting rpm and just send the filtered value to the master controller.

RhynoCZ: Yes, vibrations are an issue in some cases, which is why I ordered a hall effect sensor (also on other's recommendations) for the speed sensor instead of using the reed switch I had. I hooked up the reed switch and tested a simple rpm program which didn't seem quite right. I then hooked up the hall effect sensor, and the rpm was smooth as silk. I could have just used h.e. in the first place but that's not my nature. I have to see for myself, it's fun experimenting! As for the rest; vibrations won't really be an issue as it's all solid state. As long as the connections are good (my soldering skills have progressed enormously) everything will be fine..

The Shaggy D.A.: "Main consideration.." There's a reason why I STILL haven't gotten round to plugging anything into the bike yet. I do have a multimeter which (weather permitting) I will use to test some things today, along with a voltage divider circuit I have in mind to get from 12-14v to <=4v. The only thing that troubles me, is that if the wires I test output a frequency of 1.21 gigawatt pulses (guess the film reference) the voltmeter setting on my multimeter will show a lesser voltage depending on the frequency. Kind of like how you can dim an led using PWM from the arduino.

If anyone can suggest a way of testing the actual voltage regardless of frequency variations, without using my arduino in kamikaze mode, I'd love to hear it. I'll have a search around and see what google thinks, but what I've come up with so far is to see if said wires I test will fry an led. Seems cheaper to lose one of those, over a £40 board I got for my birthday!

On a completely unrelated note.. This very early post is due to the fact that I didn't get a wink of sleep last night. I've no idea why, but I gave up in the end, and just sat around googling stuff for the project after going out for a bimble to the 24hr garage.
____________________
Current: VFR800 2008 13.7k and rising rapidly!
Sorn: 2003 FZS1000 30k(ish) looking to sell.

[orac]

converting voltage to RPM isnt that simple tho, you will have to eitehr run maths to do it, or use a table/array to get a refrence for relevent RPM.

using an i2c slave would mean the slave would do the maths and run any extranious code, then its a single command to get the data from the bus

for picaxe its as simple as

hi2cin [newslave address], (varable)
let pins = variable

this would make the pins = whatever the slave has defined the pins to be.

also you will need to filter any data coming from the engine electricly before it gets to any IC. do you have acces to a scope, i cant help but think to get somewhere close to the correct analogue reading vs rpm.

i dont know if the ardino has a pulsin either but that maybe worth looking into
____________________
Current rides - 2016 Triumph Street Triple Rx, 1994 Suzuki Bandit 400 VM, TGB 204 Classic 125cc
"with nothing left to lose, there is everything to gain. It's not the size of the dog in the fight, it's the size of the fight in the dog"

[blade023]

I might look into picaxe at some stage. I was thinking about signal filtering too. There's an input to the tacho (+ and - on the last one i looked at, which outtput 12v frequency pulse) which must be reasonably clean, as it drives the needle on the RPM gauge. Finding information on that isn't particularly easy but I did manage to unearth a few things. There are a few different needle moving systems, one of which drives a stepper motor (I *think*) from the frequency of the wires I mentioned. Often the voltage is dropped down to about 6v in the actual RPM circuit. I even found a description of a mechanical RPM gauge which sounded fascinating.

Another thing is that servos use frequency to dictate the position of the arm, and work on 3-7 volts, so that seems like a sensible way to make RPM needles move. Assuming the voltage conversion within the RPM circuit.

I don't have access to a scope at all, and looking at prices I certainly won't be buying one. I did wonder about using the arduino itself as an oscilloscope, but then again I'd have to make sure the voltage was nice and low. Simple enough as long as you know what the starting voltage is. So I could...

1) Use the arduino to figure out the voltage from the RPM gauge input wires..

2) make a voltage divider to drop that voltage to ardu-safe levels..

3) invent a time machine..

4) use the voltage divider with the arduino to make a simple oscilloscope and

oh look basil, I've gone cross eyed.

I need to do some testing and find the RPM signal wires. I'll pea roast as soon as I've done that.

Oh, and the arduino does have a pulsein function, which I'm going to try out once I figure out some code to go with it. I've got a hall effect test bench set up to simulate pulse generation for program testing.

12v batt --> pc fan with magnets --> h.e. sensor --> arduino
____________________
Current: VFR800 2008 13.7k and rising rapidly!
Sorn: 2003 FZS1000 30k(ish) looking to sell.

[orac]

https://www.techsupplies.co.uk/epages/Store.sf/en_GB/?ObjectPath=/Shops/Store.TechSupplies/Products/KIT120
____________________
Current rides - 2016 Triumph Street Triple Rx, 1994 Suzuki Bandit 400 VM, TGB 204 Classic 125cc
"with nothing left to lose, there is everything to gain. It's not the size of the dog in the fight, it's the size of the fight in the dog"

[blade023]

Ok so I did some checking on the bike. I found this neat looking connector block to the dash which was surprisingly easy to get to, and looks very easy to tap a couple of extra wires in. I also found an wiring instruction manual for a gear indicator which taps into the exact wires that I want to use. It has wire colours for the fzs100 (2001) so I'm going by that.

In the photos below, numbering the wires on the connector block 1-8 on one side, and 9-16 on the reverse (second photo) like so:

[blade023]

btw orac, thanks for the link. It was interesting, but still pretty expensive. I was wondering about pic controllers while doing arduino stuff. Are they similar to arduinos?
____________________
Current: VFR800 2008 13.7k and rising rapidly!
Sorn: 2003 FZS1000 30k(ish) looking to sell.

[orac]

its normally just the way its programed and the language used to program it. the reason for arduino being so popular is the sheild system they came up with coupled with using C based language which alot of programmers already know.
____________________
Current rides - 2016 Triumph Street Triple Rx, 1994 Suzuki Bandit 400 VM, TGB 204 Classic 125cc
"with nothing left to lose, there is everything to gain. It's not the size of the dog in the fight, it's the size of the fight in the dog"

[blade023]

Well, there has been a bit of a crisis at home, so I've not had much time to do all the things I wanted to. I was planning on testing a few wires on the bike, making a voltage converting circuit, getting the 12v down to about 3-4 and inputting it into the arduino to start testing. Since my last post, I went round to a friends', and we started testing some wires with his multi meter. Then suddenly the neutral switch stopped working. I still haven't got round to sorting it yet (that was a couple of days ago) due to this crisis...

Our beloved cat of 17 years passed away after her final visit to the vet Wednesday morning. Mum was quite upset, and I'll admit it's hit me pretty hard. I made a memorial cross type thing for the grave. But as if that wasn't enough, the other cat we have, who's about 7 months old, decided to go out Thursday morning and not come back! He's never been out for more than a few minutes at a time (usually) or at most an hour or two. Especially since he eats a lot, he's always popping back in for food.

Everyone's been out searching, asking around, phoning around to try and find some info, but nothing. As I got half way through this post (started at 11:30, it's now 12:07), GUESS who's come back in! I'm so relieved. If he could talk, I'm sure he'd say something nonchalant like "sup..." as if nothing has happened.

So anyway, as soon as I have time tomorrow this project WILL resume. I've now got to the stage where I can plug in a couple of wires to the bike, and get some meaningful output, as I've made an LED shield for the 'duino, and added all necessary bits into a handy portable package. Once I've got the basic coding infrastructure down (about 80% done) it's plug and test time. Oh and figuring out where this fuse is for the neutral sensor so I can change it.

I'll post up another vid and/or some pictures hopefully tomorrow.
____________________
Current: VFR800 2008 13.7k and rising rapidly!
Sorn: 2003 FZS1000 30k(ish) looking to sell.

[blade023]

So I've pretty much finished the first version of the code, and I'm ready to do some field testing. Would anyone be interested in seeing the code? I'd be happy to paste it into a forum post, or maybe a .ino file would be better?

Also nowhere.elysium: If you're interested I could make a wiring diagram (pcb, breadboard or schematic..) I've been using Fritzing to prototype some stuff which is how I came to the final design of the LED shield. I still need to do one for the 7 seg display and the tacho/speedo input, but I may intergrate 1 or both of those into the LED shield. Until the testing is done I'm not 100% sure what components I'll need.

Out of curiosity, what bike were you trying to intergrate your dash with? And what wires/connector go to the dash? I know you were talking about picking up the pulses from the spark plug leads, but it may be a lot simpler to connect to the existing dash wiring as I'm fairly certain it's a dc signal, so only needs dropping down a few volts. Current draw would be minimal for the arduino too, so that shouldn't be an issue for the existing dash if you're not replacing it.

Testing should start soon after lunch; I'm hungry and we're in the process of burying a little furry friend. Have to do something though to keep myself occupied. Otherwise you get dust in your eye, etc etc..
____________________
Current: VFR800 2008 13.7k and rising rapidly!
Sorn: 2003 FZS1000 30k(ish) looking to sell.