That kit is something that AZ-D offers-apparently for some time now, which is why I didn't understand the hype over this post....
Printable View
Jus tto throw a spanner in the conversation, has anyone thought of ultraviolet light? Will it light up the white graphics on the cluster?
Just an idea.
But I do like what Nicholas posted up, looks great!
You know I've never thought about that but it would be kind of neat. Heck the gauges are already white; I'll bet even just a thin uv light bar or strip would make it pop. Hmmm...
Also it looks to me like the illuminated gauges I just posted are just decals that have all the numbers clear with electroluminescent strips behind it. Could be mistaken though...
The SpeedHut EL decals are most definitely NOT the same. Those decals are rather lackluster in person. They do not even closely match the typeface of the original cluster, are very monotone at night, bleed light through their edges, and don't solve the problem on non-backlit needles. Not to mention your still running a completely original cluster with all of it's disadvantages.
UV illumination is something I've tried as well. The red and yellow light up very nicely, but that is it. The white lettering, blue lettering, and the needles are all as dark as could be under UV light.
I havent had much to update about this lately, but I hope to within the next week or so have a full breakdown about the entire stock cluster, it's specific weaknesses, and what will specifically be improved upon with the new one.
Thread bump! Any news? :)
Farrar
As with everything anybody does life slows you down.
What I do have...
Limited working software to drive the stepper motors for the needles. Still needs lots of refinement though and I'm trying to obtain specific driver IC chips for them which will make them work much nicer. Minimum order requirements: 100/pcs @ $6.50/each. That fine for producing 50 dashes, but not so fine for prototyping to see how well it would work.
Also have a lead on a plastics manufacturing company which has done low volume production for reasonable rates. Conveniently they are only 2 miles from my home so working with them should be easy (keyword: should).
I had also brought my spare cluster and the sample backlit decals to the Minnesota Fall Apple Orchard gathering to show how the design is currently working out.
Also now that I finally found the battery for my good camera I can get some high-res photos of the sample decals and more details about the stock cluster and writeup some more info on what is changing.
Stay Tuned....
Cool beans, Jim! I just wanted to make sure you hadn't given up on the project. :-)
Farrar
That's awesome Jim! Keep it up!
+1, im excited about this and i dont even have a D yet... :wiggle:
I have experience and a good source for rapid prototyping if you're considering using any sort of additive manufacturing for a prototype or mold.
Thread bump!
Any progress? 8)
Yes, still progressing. No new good details to report yet. But certainly still progressing.
On a side note, if anyone can figure out a way to obtain ONE or TWO of these chips...
http://mcrmotorusa.com/IC%20Drivers.html
Or compatible Switec M-S Quad Driver X12.017
It would be a spectacular help. They should be the *key* to a good fluid needle. However the only manufacturer/distributor I've found of them or equivalents will only sell by the case. A box of 100 at $6.50 each. I dont want to buy the whole damn case until I know for sure I can get the electronics to work right. They've told me their "dealer" network only sells their stepper motors.
So there's a little preview of the costs which will be involved in development of this. That box of 100 costing $650 will supply *ONE* component needed twice in each dash, enough for 50 dash units. Fortunately not many other components will have such a high minimum purchase requirement.
Also compatible is a VID66-06 or VID66-08 chip. These are all manufactured in China and seem to only sell in even larger quantities than I can get the other one for. They are all cross-compatible.
***BOUNTY***
If anyone can find me a small quantity (or individual) of these for a reasonable price, I will reward you with $20 or a more substantial discount on the finished product.
There seems to be a bunch of stepper motor ic's on eBay available in small quantities. I know nothing about this stuff but thought perhaps you could have a look there.
Search eBay for stepper motor ic
I have easy access to the actual stepper motors. They are sold by everyone to replace the original bad ones on mid-2000's GM vehicles.
What I'm looking for is a source for new IC driver chips which effortlessly drive the stepper motors in a smooth fluid motion. I've found several compatible chips, but no source to buy a small quantity of them for prototyping.
http://mcrmotorusa.com/IC%20Drivers.html
MS Quad Driver X12.017
VID66-06
VID66-08
mouser.com seems to have a bunch of driver chips available individually.
Small Update...
Dedicated "specific" driver chips are assumed to be not available right now, based on my refusal to buy a very large quantity without knowing if I can get them to work right.
On that note, I have progressed with programming a chip to drive the stepper motor directly. Its a crude program, one for forwards and one for reverse, nothing linking the two or varying how fast it moves and nothing to stop it (except the stepper's physical stop).
BUT IT WORKS.
As I have time on Sunday and Monday to further refine the program, I'll take a video and post it up.
Jim,
Have you contacted any of the Driver chip suppliers to see if you can get developmental "Samples". Usually "off the shelf" component suppliers will work with you even if your only a small fish. Was there anything available at Digikey,Newark,Allied or Mouser?
This is a very specific type of driver chip, probably very uncommon with hobbyists. None of the common major electronics distributors have anything even close. I've contacted MCR and Juken Swiss Technology about samples. MCR said no, minimum order is 100 at $6.50 each; and there was no response from Juken.
In other news, after this weekend's foray in my first extended go with PIC programming....
I now have a basic control of the stepper motor destined to drive the needles for the gauges...
http://www.youtube.com/watch?v=d9Y4iUdh8A4
This is running a preset program which is turning it 300ยบ in each direction, with a 1 second pause between each rotation.
Next step is to control its movement and position with a potentiometer. This will simulate the oil pressure, engine temp, and fuel level senders; and provide complete proof-of-concept with everything I need to do. If I am successful in doing this, I will likely not even need the dedicated driver chips I talked about earlier.
Progress is here.
Looks good. Even looks fast enought to use for the tach. Another guy getting into software to replace our 70's hardware :giggle:
And it should be able to go at least twice as fast as this, with proper acceleration curves built into the program. (That part is still beyond what I've accomplished). This seems to be the fastest it can go reliably without missing a step.
To finish my next goal, I need to divide an 8 or 10 bit ADC code into the number 150 or less, then write some if/thens to add/subtract X (the difference between Y [the current position of the stepper] and Z [the ADC code, the desired value of Y]), to/from Y.
My first attempt at reading how to do this (the chip doesn't actually divide) had my head spinning faster than that motor. My first reads on the if/thens were at least mostly comprehended.
To get your prototype working, i suggest you use a higher level cpu/platform to experiment with the hardware. I've had amazing experiences with Netduino or TinyClr, both C#, .Net capable, step through debugging, full programming support. Their prototype boards are about the size of a deck of cards.
Just shift the 8 bit value to the right one place to divide by 2. Hence you will get a number from 0 to 127.
For 10 bit shift it 3 times for the same 0 to 127 number.
That is how you divide in assembler code with simple CPUs.. To divide by other exponents, you can manipulate a few shifts and add the two values.
Thats a bit more complex than I need for this project. I should be able to control both the Volt/Temp gauge with a single PIC16 chip, both the Oil/Fuel gauge with a single PIC16 chip, the Tach with a single PIC16 or 18 chip, and the Speedo likely a single PIC18 chip.
This whole project is a start-from-scratch learning for me. Too easy to go over my head so I'm keeping each step very simple. Starting with the simple gauges then working my way up.
Ordered 6 .30" 7-segment green LED displays for experimental use for the odometer. A Vacuum florescent display is going to be DOA, as there are no commercially available ones in the package/size I need. Not to mention that in general the LED displays are several times cheaper AND have the ability to be built in different colors. There are .28" ones available as well, but I'm going to try the .30" ones first. FYI the vertical opening on the speedometer for the odo display is .28".
Color update - Backlighting WILL be available in 3 colors; Green, Red, and Blue. These are the most commonly requested colors. Green backlit will have alternate colors for all the alternately colored labels (red, yellow, and blue). The red and blue backlighting will wait and see how it looks before determining those other colors. Odometer colors will be green (normal), with red and blue optionally. If it is ever desired to switch colors of the backlight (like a person's different preference if the car is sold), the unit can be returned and colors switched for a small fee. Unfortunately RGB 7-segment displays are not available in the extreme small size needed, thus a user-adjustable color can not be an option.
After this project is done (at least a year from now) I may consider building a replacement A/C decal in the unavailable most-desired style (Type 2 or 3 here), LED backlit, which will have the option to plug-in to the new cluster for matched dimming. Color options will be the same as the cluster.
For proper 'old school' display, this would be way cool:
http://blog.onlineclock.net/wp-conte...tube-clock.jpg
nixie tubes are totally sweet... i don't think theyd match the interior though :wink:
If I may ask a question out of ignorance: why did you decide to use multiple ICs instead of a PIC with multiple inputs and outputs like an Arduino? I ask only because I have seen a BMW gauge cluster driven by an Arduino and I think it had the same number of gauges as our cars. Thanks, and sorry if it's a dumb question!
Farrar
That's a good question. If you have dedicated simple routines than it's better to use many small micros. With large complicated software, you make a simple change in one routine and a bug creeps into something you never intended to change. It also makes circuit board layout much cleaner. These small micros cost from 50 cents to a few bucks. If you use interrupts then that can change the timing anytime you make a software change. We run into this problem at my day job. We share a 700 pin FPGA on a circuit board and when one user makes a change in that FPGA it messes up a lot of other logic timing.
That makes perfect sense. Thanks, Dave!
Farrar
As of right now, no they will no black out. I need to keep consistent parts to try and keep this affordable for people. Too late to back out now anyways Adam, you already commited on one back at the beginning of the thread! :evilgrin:
I will ask if it's possible for a blackout face when getting the main decals printed. If its not horrible, I may let people choose it for an additional cost.
Dave pretty much answered it all. However these chips do have multiple in's and out's. While my software is still in its infancy, the side gauges will be sharing a single PIC with the software loop just going back and fourth updating the gauges. Really easy since none of them have consistent rapid updating like the tach and speedo will. The tach is going to need to interpret the odd-fire RPM signal and update the stepper position sometimes before it even reaches it so the needle doesn't stop-start or overrun the speed. The speedometer will need to keep several odometers in addition to rapid updates.
So the gauges will be lit at all times when the key is on? Even with the visible graphics? Will the graphics be color matched or translucent clear/white for the various color options?
No they will not be lit at all times. Only with the parking/headlights. IF I made a blackout version I would need to modify the back lights to be on extra bright during the day.
Only what is white right now will still be translucent white. Normal illumination color for those will be green. Other colors will be their normal tint and illuminate in those colors. See image in first post of this thread.
Right now the other color options will only be able to do the white lettering (everything green in that first post photo). Once the first set of decals are made I'll play around will all the colors.
Along with the PIC progamming I'm doing now, I'm also making the vector graphics for the decals. Doing much of this work myself should help keep costs down. I'll post visuals of them as I get them completed.