I think I've pretty much sussed the design for a display (aiming for 256 LEDS max, although even this could be upgradable), with pluggable modules (e.g. numeric RPM or plug replacable bar graph) I think the controller/display could be done for around £60 (chips comming in around £30 + £60 for LEDs and other sundries). Programming won't be a problem, just something that will take time.

The only bit I'm concerned about is inputs at the moment and designing them to be robust enough for automotive use. I guess I can get RPM sensors OK but I guess I'll need to sort out analogue inputs for water temp e.t.c. Anyone tried this before?

The aim would be to just have an open source digital dashboard with configurable options so that say, you could plug in a bar graph module and set for bargraph mode (e.g. rpm) or display as a 4 digit display, or have the option to show the warning lights. Keeping things modular would mean more wires, but also a non fixed layout.

Sounds like a cool idea! Once you have it going You could multiply the LED outputs so you could sell it to the bling brigade so the under car LEDs could change colour according to speed or engine revs! Or how about reversing the LED alphanumerics so you can have a head up display! This has lots of fun applications as well as the options to design your own cool and groovy dash board, of course the price for the bling drivers would be about 3 times the dashboard drivers

Just thought of another application, when you exceed the speed limit you could get a random LED flash on the numberplate to confuse speed cameras!


Caber

i gues you're going to use pic's to do the clever bit... most of them have have a/d ports that you could compare to most thermo-resistive devices... quite easy to calibrate also with boiling water / ice...

oil pressure... now that would be a chalange... unless you could compare the output to a known source... still...

it wouldn't take much more work to implement a usb port and write logging data out... to a pda of some sort..


thinking thinking!!!!

and... you could implement a protocol stack to read gps data from a bluetooth GPS... then you'd be cooking with gas...

anyone up for a group project!!!

Yes, someone on here was developing their own instrument cluster, but I cant for the life of me find the relevant thread. It had LED's, a black facia and a brushed aluminium type surround with a very nice looking overall shape.

here you go, found it..................

thread

I'm sure I have some circuit diagrams somewhere fro a universal digital dash!!

Just one observation regarding the use of LEDs. Having used my Evodash for the last six months I find that the LEDs practically dissappear in bright sunlight, but are blindingly bright at night.

I can manually turn the brightness down at night (an automatic level controller from the lighting circuit would have been nice) but the sunlight problem still remains.

Ultra bright LEDs might be an answer but maybe LCDs would be a better option?

quote:

---

Originally posted by WanchaiWarrior
here you go, found it..................

thread

---

quote:

---

Originally posted by caber
Sounds like a cool idea! Once you have it going You could multiply the LED outputs so you could sell it to the bling brigade so the under car LEDs could change colour according to speed or engine revs! Or how about reversing the LED alphanumerics so you can have a head up display! This has lots of fun applications as well as the options to design your own cool and groovy dash board, of course the price for the bling drivers would be about 3 times the dashboard drivers

Just thought of another application, when you exceed the speed limit you could get a random LED flash on the numberplate to confuse speed cameras!


Caber

---

I made this for my brother



High output blue LEDs on the left (low level), high output red LEDs on the right (high level).

Had adjustable low and high limits on every parameter. This is just the display board PCB btw, I then had a mainboard which had the filtering circuits on. Only trouble is I updated the screen to often so when the fuel sensor was bouncing around the displayed figure would change too often also making it difficult to read....
I really should have finished the project. I ran into problems with my code going beyond the far call barrier so everytime I added new code I had to recode the last few functions (which were pushed beyond the far call) to cope... Which was a PITA!!!

Have you had a look at the megasquirt schematics?, theres inputs and reference outputs on there that would do the job so could give you some ideas.
Paul.
[quoteThe only bit I'm concerned about is inputs at the moment and designing them to be robust enough for automotive use. I guess I can get RPM sensors OK but I guess I'll need to sort out analogue inputs for water temp e.t.c. Anyone tried this before?

The aim would be to just have an open source digital dashboard with configurable options so that say, you could plug in a bar graph module and set for bargraph mode (e.g. rpm) or display as a 4 digit display, or have the option to show the warning lights. Keeping things modular would mean more wires, but also a non fixed layout.

what processor did you use... you could have run that with a series of 3 pic's for under a tenner... wasn't all the cost in the lcd displays???

for the fuel gauge it seems to make more sense to display the average of the last twenty samples minus the top three and bottom three... (ave of 14 mid points) and a sample interval of 5 seconds (that should get you round corners and over the humpback bridge)

yeah - did one ages ago - it's in my car & has passed SVA (actually twice now...another guy also has one). Mine used a FPGA but a micro would be a more suitable basis. Yes LEDs are not so good in bright sunlight - but easy to dim for night use (I have a 64:1 brightness range & this seems fine).
You won't have much trouble with interface electronics - for temp, oil pres. and fuel tamk I just used a load resistor to a 3.3V supply rail & an ADC. For tacho I used a 2 pole RC filter.
Bob