Project "Better" Stock Clocks....

I, like many, really want to have a more functional cluster. I have an 80 650 maxim with pretty basic gauges. I have a few warning lights and the tach, speedo, odometer and trip....

I, like many, want to swap my forks out to something more modern and will eventually have to figure out a way to make the stock speedo work correctly. And my bike also suffers from the tach drive leak syndrome... Where if it isn't leaking now... it will be soon.

I really don't think that a fully digital cluster "fits" on the bike.... Yeah you can get them mounted but, really, they stick out like a sore thumb.

So here is my idea:

I'm going to make my own gauges. I'm going to use the stock bits and pieces but all of the internals and faces will be one-off or hand made. I'm a mechanical engineer who dabbles in electronics so I figure "why not?"

Features (its a long list and getting longer!):

All micro-processor controlled
Sweeping needle based gauges will remain (by far the best gauge is a sweeping needle)
Again, Stock looking, stock enclosures
120 or 140 mph scale
12k tach scale
Back-lit carbon fiber gauge faces
Small multifunctional display
digital tach signal (removal of the cable system)
digital speed signal (removal of the front wheel driven cable)
Programmable "smart" shift light
built in time clock
3 trip computer (for gas tank range, maintenance, and other)


Features being thought about...
Head temp info
Oil temp info
Oil pressure info
Custom colored back-lit


SO! Where to begin?! Well right now I'm sourcing a set of stock clocks for parts salvaging and have been solving "How do I do..." type questions. I have been shopping around for parts but for the most part, this project is in it's infancy...

 

This should be interesting. Keep us posted.

 

Might not fit your requirements but I always thought this setup was elegant:

 

ok... So I have been putting a lot of thought into this project. I got most things worked out. The number one pain in the butt is how to detect RPM not using the cam. Right now, I would like to make a one piece plug with a stock o-ring to drop into the tach drive assembly bore in the head and completely blank it off... This way there is no worries about leaks later. I can either detect RPM by an inductive pick-up on one of the coils (like a lawn mower tach or some bike retrofits) or I can use the stock timing wheel and place a photo sensor in the to see the RPM... The former is easy, but kinda ghetto. The latter is harder but to install but easier to circuit and have work repeatably... I might try both.

Also, the gauge "drives" or movements will be small R/C car/plane servos modified to travel more than 180*. I own a subaru WRX and when you turn the key, the gauge move the full motion before settling in to the 0 readings, which I think I'm going to steal for these.

I think I'm also settled on the "features being thought about". I'm going to include a outdoor air temp reading and a 80 mile range warning light (I don't have a gas gauge!)... The oil temp and pressure is a nice thought but there is no easy way to implement them.

I may modify the stock clocks enclosure so that the little divit in between the two gauges is where the read out screen will go. And a might remake the enclosure in carbon fiber...

I'm trying the keep the budget below $200....

 

Anyone work with Waterjet cutting machines?????? I just came up with a great idea for the gauge faces... I was going to silk screen them but this blows that idea out of the water!!!

Anyway if anyone can run a waterjet and convert cad drawing to usable cam progies for it let me know!

 

double post

 

I'm just glad to be rid of the 85MPH speedo... plus you gotta have a clock.

 

Not to hijack, but would that instrument cluster work on a 650 Midnight? I REALLY like the better speedo, and with a 750 tank with a sending unit...

 

Yeah yeah rub it in that you have a nice cluster with a clock. I have a wrist watch strapped to my bars...

Two things. One, 150mph for these bikes is a bit much, if you didn't have that nice 270 degree needle wrap, those number would be really squashed in there... Two, man you need to put some miles on that bike! :wink:

 

Dang... Hit my first road block. The water jet place states "the gauge faces need too fine of a jet to be made" Apparently the .030" jet causes nasty over shoots and unwanted radii everywhere. I was excited about the back-lit design that I had in mind, so now I'm kinda bummed that I need to redesign... I really don't want to silk screen the numbers but I may have to.

 

For the record, I follow this thread, but that last post only vaguely makes sense to me. Interesting, though. Wish you well, will be interested in your designs. My speedo needle snapped shortly after I got the bike, and I wouldn't mind ideas on new custom gauges.

Fitz's are nice, though. Seriously, fitz, where is that cluster from?

 

What is more impressive is the other XJ in the background, and the OTHER bike also pictured to the right.

 

Sorry about that. I guess because I am so wrapped up in the project, I forget others don't know what the hell I'm talking about.

I was going to have some carbon fiber gauge faces cut on a water jet. The numbers and "ticks" would have been completely cut out except for the centers of the zeros, eights, six, etc. Then I was going to have a thin white backer so that the gauge faces would have been natural carbon fiber with white numbers. At night, the numbers would have been back-lit so the white numbers would be yellow or red. On the tach, I was going to have the back-lit color change (when driving aggressive or really revving the motor) from yellow to green at 7k and red at 9.5k.

Without the numbering cut out.... all of this becomes nearly impossible.

 

That sounds like it would be a pretty awesome instrument panel. Even with silk screening you can make a pretty awesome gauge and even customize it to a degree.

Have you looked to see if there are any alternate cutting options available at your disposal? I'm not sure if this would be an option (due to availability of a commercial location willing to do it or due to pricing) but perhaps it could be laser cut?

 

hmmmm.... Didn't think of that! I'll have to call around.

 

Not to steal your thread but I am wondering about Maxim gauges. Are they readily easy to take apart just for cleaning the interior *glass* portion and painting? I've read bigfitz52 tutorial on repairing & cleaning the speedometer issue(s) with a seca but has anyone worked on the round maxim gauges? Can this be done with the gauges mounted?

On both my 550 & 650 the front clear portions just look "older" and I would like to paint the lines & needle. Or would I be getting myself into too much trouble...

 

Thanks for the tip!!! I was able to source a shop to laser cut the faces.... with one small hickup. They will not touch carbon fiber.... Between the fumes and the burning epoxy messing the laser head up the guy said it will never happen. But, he can do them in thin aluminum.

So, the way I see it now is the faces will remain back-lit and they will just be gloss or semi-gloss black instead of carbon fiber. The project marches on.... Such is life.

Oh, and I'll report in how hard the gauges are to take apart once I get my hands on a set. I can't imagine that it is all that hard.

 

wouldn't it be easier to just buy one
http://trailtech.net/vapor.html

 

way cool idea, tuned in for sure!!

 

there is always an easier way... :wink:

 

A teaser of what I had in mind.... Maybe if I have time tomorrow, I'll extrude the drawing to get the full effect.

 

Really liking it. Have you considered adding some sort of logo to the bottom portions of the gauges? If you are spending all that money going custom, why not make it TRUE custom? Add in a personal touch, or maybe a stylish model, like 'XJ650' on the speedo and 'custom' or 'racer' on the tach?

Just personal preference, but I would put in something extra.

Hell, even if you could find a digital clock to put in the bottom of the tach, or maybe get a 750 tank with the fuel sender and put a fuel gauge in the bottom of the tach.

No matter what you do with it from here, it does look good so far. Just be sure the gauges are a matte finish so you don't blind yourself in the afternoon sun during an eastward travel!

 

The lower portions may be covered by the stock gauge enclosures... I haven't cracked one open to see if I can take them out.



Either way, the speedo will have a small lcd screen displaying information. The tach with have the shift light just below the "6" or maybe off set... not sure yet.

I was thinking about getting some tuning forks in there some where.

 

A few renderings of the faces...

Plain Jane


With tuning forks



The tuning forks version may not work.... BUT if I can get it in there, I think its a winner!

 

I like the tuning the forks. if you can pull that off it will look really nice. are you keeping the odometers in the speedo housing or moving then elsewhere?


Edit: manual g my autocorrect.

 

Not without breaking the glass I'm pretty sure.

You could get a vynil in carbon fiber and lay it over the aluminum piece. Then you get the look.

 

has anyone thought about the ratio change if you were to change the gauge face decals,,, since it will have numbers at different degrees than originally?? cause the way i understand this thread is we want to change just the decal and not the actual mechanism(s) behind "the scenes". at least that's where I thought this was going...

I, myself, have been pondering a change on my cluster but if the numbers are not in the same location as before, then it won't correctly read the right output (tach & speed).

www.productiondecals.com

they have a whole variety of gauge face decals, yet none for XJ series
waiting on a response from them too about this similar question as well

 

No, it sounds like he's planning on redoing the whole insides too, with microcontroller + servos. In which case there would have to be quite a bit of calibration going on...

 

BA! I would still have to cut out all the letters and those decals never look all that great.

I was looking at the parts drawings of the tach and speedo... from the looks of it the lower raised area covering the bottom of the gauge is part of the gauge and not the enclosure... so that may be promising and open up some more opportunities.

This is correct! A whole new set of guts with the stock enclosures.

A quick update. I'm waiting for a hall effect sensors... They are key to getting the speed and RPM information.

A simple lesson of how I see this working.

A hall effect sensor senses when a magnetic field is near, and when it does, it sends a signal that is easily interpreted. Just like a switch turns a light on

No magnetic field = 0 volts
Magnetic field = 5 volts

"But Rob, how will you read the engine speed?"

It should be fairly easy to fit a small magnet on the timing wheel under the left side crank cover on the motor and a hall effect there. Every time the motor spins once the hall effect will see a magnetic "pulse". Using these formulas in the mircoprocessor we get RPM:

RPS = Number of pulses/second
RPM = RPS x 60 seconds

"Ok, well what about the speed?"

Same system, same application, different sensor and magnet location. This time the sensor will be located on the inside of the fork (on the rotor side) and the magnet will be stuck in one of the rotor bolt heads. We get speed using these formulas:

WPS = Number of pulses/second
DPS = WPS x ((3.1416 x Wheel height in inches)/12)
DPH = DPS x 3600 seconds in an hour
MPH = DPH/5280 feet in a mile

I'm sure I'll have to play with the sampling rate (pulses/time unit) because at lower speeds, under 5mph we'll say) the pulses/second may not be greater than one. So effectively the computer will think that there are 0 pulses/second because the wheel does not make it all the way around in 1 second. Remember, the sensor is on or off and the computer does not know anything else.

Another way to correct this is to use two magnets and divide the MPH by 2.

I have a few more ideas to read RPM, but I will try them as I have to.

I will be starting the printed circuit board design soon for the brains of the project. And have been sourcing parts (temp sensor, hall effect sensors, SD card socket, Real Time Clock circuit, LEDs, etc.)... I still haven't had luck getting a hold of the gentleman who apparently has a set of gauges for sale...

For those curious, here is a sample of my printed circuit board handy work... This one is a wireless temperature and humidity sensor node for data logging.

 

This is really cool to read about. Would love to read about how the stock gauges work as well, would help me diagnose mine.

 

I just had my gauge cluster apart. It appears that those pieces are removable however I believe they are glued in and holding the glass. So you'd probably have to make some kind of ring that would fit in to hold the glass.

 

Well then here you go: http://xjbikes.com/Forums/viewtopic/t=15303.html in detail and with pictures.

 

Brief update... I have goodies on the way...

A nice little screen I'll be testing out for the display... It's got a built-in micro sd slot, which will be nice because I'll be storing mileage, trip and setting parameters on it. I bought a real time clock module rather than making one... I wouldn't have saved more than a buck or two making it and it would have taken some time so...

I think I finally got a set of busted clocks for my "parts bin"

I have been toying with some of the code but no "real" progress...

 

this thread blows my mind, yet i can follow it completely. i love it!!

 

Circuit board is done... I made it more generic because I have to order in batches and I don't think I'll make ten XJ clusters.... or will I... hmmm. :wink:



Anyway... It will take 4 weeks to get these printed so I have some time to get the program written and get the enclosures, which I got for $20!!! Got most of the other parts in today but I had to replace my water heater so I didn't really have time to play with any of it.

 

Gentlemen....

A little demo video of the very basic design... The screen read out that is really blurry is printing "34 pulses per second", which is about 2040 rpm...

The really chunky "needle" movement is because the fan with the magnet taped to it is bouncing around and can cause bad readings. The servo movement needs to be smoothed out (maybe).

Let me know what you guys think.

http://www.youtube.com/watch?v=cjEpvbHCKIY

 

Just a little more work tonight...I didn't have much time, but I wanted to test the screen. It's SO small, the size of my thumb, and I was really concerned that I would not be able to read it at arms length. I love it! I can read it from 5 feet away with the larger font and 3 feet away with the smaller font. The video shows a test of the screen, first line is Odometer, second is trip "A", third is temp and forth may just be a status line (change oil, refuel, who knows....). Then the screen displays and counts a fake RPM in a small font to show the refresh rate.

http://www.youtube.com/watch?v=yOXfHDY5oS4

This picture was taken 36" away...

 

Ok... so as it turns out, the hall effect sensor I'm using can and will pick up all sorts of noise from the coils and plugs Trying to read to RPM off the timing wheel caused it to go haywire.

Back to the drawing board for now.

 

Quick update...

The wiring itself may have been the problem. The bike's ignition is so damn noisy that the unshielded wiring for the signal from the hall effect was picking all the high voltage noise. I'm changing it all over to shielded wiring and trying again.

If I can't get this to work I may just watch the coil primary wiring for the 12v pulse each time the coil charges.

 

Sorry for the lack of progress... I have been testing out ways to filter the tach signal. I have also noticed that the servos I'm using are too darn slow... and kinda big. So I'm in the market for some peppier servos.

 

servo as in Futaba R/C servo?

 

yeah... I'm using airtronics right now but they need to go. I'm going to get a set of micro Hitech's that are twice as fast.

 

Just be aware that when the coil fires there is up to a few hundred volts on the 12v coil negative, induced when the field collapses to fire the coil. How about using the noisy un-shielded system to your advantage?

You may be ably to pull a signal from the secondary by putting a few simple twists of wire around a high tension lead. (if you don't want AC throw in a tiny signal diode).

Good luck.

 

I haven't found any real useful advice or schematics for the inductive pick-up... Everyone say a few turns around the HT wire but when I do that I get nothing...

 

Here is a video of the tach response time and what not after I rewrote the code and am using a different method to pick up RPM... Getting close but I think that there is a bit of room for improvement with a faster servo. Let me know what you guys think.

http://www.youtube.com/watch?v=_RoXiPW3hWU

 

DANG!... I can't win!

So, this system shown above works on the bike using the timing wheel... works good actually... BUT the sensor doesn't clear the timing wheel cover!

Ok, no big deal. Either I'll buy a extra cover and modify it by making it deeper or I'll just have to machine a spacer to space the cover from the block.

 

How are you measuring it? Oscilloscope or multimeter?

 

At the time I was using 6 coils of unshielded wire around the HT wire with a big cap on one end of the 6 coils and then reading the voltage across the wire with a multimeter. The hope was that the cap would charge slowly and I would get a good overall voltage reading... It didn't work.

I guess I can go back and get rid of the cap and put a o-scope across the leads and set my trigger to rising (and pause after the trigger has been passed).

For a good signal, I can use a schmitt trigger with a 555 timer and hopefully filter any noise.

 

I'm watching this thread. Sounds like an exciting project thus far.

 

That setup works nice, but too big? Ever consider individual
Components or both tx and rx on the same side with a reflective stripe