Gauge Design
 

Hi All,
In the next couple of weeks, I'll be having the first prototype PCBs made for a gauge project I've been tinkering on. It's evolved to include the following:

-intake manifold pressure
-intake air temperature
-exhaust gas temperature
-oil pressure
-oil temperature
-accelerometer
-input for linear analog WBO2 (e.g. Spartan Lambda sensor)
-peak hold
-SD card logging
-small (~2" diagonal) color LCD display
-~$250 price point, including all of the sensors except wideband

The idea is to have a simple but elegant interface to many of the common performance/tuning/diagnostic parameters with a minimum of clutter.

The sensor interface will be in the engine compartment. The display in the passenger compartment will have provisions for mounting at various angles but be small enough to be embedded somewhere if one's feeling ambitious. User inputs will be via a single panel-mount rocker switch to handle switching screens, clearing peaks, configuration (color selection, gauge ranges, colors, etc).

The firmware will be open source (Arduino compatible), and I'll probably break out a couple of free pins for custom input/output applications (fans, water injection, ...?). Also scheming on reading from (and maybe writing to) the Megasquirt via RS-232, but I don't have one to test so no promises there.

I'm interested in everyone's thoughts. Is there a decent market? Would I compete at this price point? Think I'd get 100 backers on Kickstarter? Does it cover the bases feature-wise? Any sensors I should substitute or not include at all? Ideas for incorporating other market segments (e.g. diesels, V8s, maybe motorcycles)? Words of wisdom for a first time entrepreneur?

Thanks for reading and for your input,
Mike

A 2" display will either display ONE gauge or it will be unreadable inside a car.

This.

An Arduino to MS driven display really needs creating though, the market could be huge.

Ideally you want the display to be circular to fit in a standard sized gauge hole IMO, with a single button to allow you to scroll though pre-determined ouptuts.

James one of the devs for MegaSquirt has created a programmable gauge (communicates via MegaSquirt proprietary CAN commands, but aesthetically (IMO) it's not great, the choice of colour for the readout is odd - I've never seen a car interior lit up in amber, so it matches nothing, and there's no cover for the internals. It looks very DIY.

Very good. I'll definitely keep the size in mind. The gauge is going to be either 2.2" or 1.8", I'm waiting for the latter to arrive for testing. I've programmed some screen layouts including both single sensor analog-style gauges and higher-density bar graphs (for testing refresh rates), but won't really be focusing on this until I get a prototype PCB order in as those lead times seem to dominate my critical path.

I've seen the MegaView, MegaCluster, and MS-CAN-Display. Is there another one? There's also the PLX LCD line, but that's focused on pre-OBD and OBDII vehicles. I'm thinking black backgrounds, white letters/numbers, and a variety of user-selectable colors for needles/graphs that can change depending on whether the headlights are on. Sound okay?

Thanks for your input, guys. I'm all ears.

This one:

Megasquirt MSEXTRA / MS3EFI • Small dash gauge (View topic)

If you need any help, just let me know, not sure what I can offer in assistance other than an ECU in my car tho.

I really want a multifunction gauge I can replace the dummy OPG with.

BTDT:

https://www.miataturbo.net/general-m...-gauges-59775/

As nice as that is, seems pointless adding other sensors when there are already perfectly usable ones there ;)

I like integrated solutions :D

I can relate to your aversion to the "carnival" look. I think your solution is pretty nice looking, well done.

Yeah, redundant sensors are a drag (and costly!). That may be reason enough to leave MS compatibility alone for now and cater to those with less brainy ECUs first. Thanks for digging up the other display post, I hadn't seen that. I like the simplicity of Jim's gauge and the ubiquity of the 52mm form factor.

IMO it's all the more reason to utilise MS compatability. It will be using OEM sensors (mostly), and has multiple ways to access the sensors output neatly - Serial port or CAN - rather than tapping into every sensor as would be required when using the stock ECU.

Cool beans..
But, let's see what the OP says about his design:

"-~$250 price point, including all of the sensors except wideband"

See, there's no "one size fits all" when it comes to these things..
Talking about MS compatibility, using existing sensors, and all that is really just utopic when you
a) do not have a sensor for whataver it is you want to measure - such as intake temps
b) have one of the many choices available for engine management, like the stock ECU, or Haltech, or Hydra, or whatever your favorite may be
c) have a vehicle that is not a Miata.

Calibrating a gauge system for a myriad of sensors available would be hell, wouldn't it?

Also, no one would even dream of investing in a system that would be available exclusively for Miata owners using MS for engine management.

I like integrated solutions, as well. That's exactly why I integrated my system into my car. ;)

Sensor are either a resistance (temp), or output a voltage WB/pressure sensor/TPS etc.

Allowing a form of configuration within the gauge for setting the resistance curve and what the input voltage is means you can account for any ECU/Sensor.

ECU integration is a nice to have granted, but utilising exiting sensors isn't. I don't see the point in duplication, it's just added complication, failure points, cost, and messes up the engine bay

So, you're saying I should have used the existing sensors for boost, EGT, oil temp, oil pressure, IAT, and AFR?

A one-size-fits-all approach to data acquisition and display is difficult to do in a cost effective and non-redundant manner, I think we can all agree on that.

The oil pressure sender is the weakest point in my supply chain and it would be great to just have a reasonably flexible voltage divider and have that one be a BYOS situation. Introduces setup complexity, but would also lower cost for those that would not install an extra sensor.

Never been in a German car?

I've got my own arduino-megasquirt-canbus solution working, I'm about to start putting together the gauge itself. I'm using Neopixel 5050's for at-a-glance reading of values I want to keep an eye on (blue = cold/low, green = nominal, red = hot/high). Nice thing about the PLX gauges is they also output a 0-5v signal which can be logged by MS. I was trying to stuff everything into oil pressure gauge's position using a 1.3in OLED screen, but with my new steering wheel it's obscured from my driving position..

I'm not a coder by trade, so I've had a fair amount of stumbling blocks while trying to make my code halfway decent. I'm writing up a howto guide right now (I think I'm the first one to actually get Arduino + Canbus Shields to work with Megasquirt.. not denying there are other at90can adaptations though..)

https://www.miataturbo.net/attachmen...ine=1385424949

Bittetech: Cool! If you haven't already, check out the videos here:https://code.google.com/p/ms-can-display/. I think you'll recognize the instrument cluster it's embedded in. The Google code page is not maintained, though, and I'm not sure what CAN hardware was used. Still, maybe there's some helpful code in there.

Congrats on getting your hardware talking and your display reading!

Great job, the picture looks familiar from my many google searches on the subject.

Do you have a blog or anything for your work?

Nice stuff, gotta keep us updated bittetech, that looks like what I want to do for an F1 style shift light system.

I've never worked with CAN or Arduino but my background is in C so if you need any help let me know.

make sure not to clean up your desk so you increase the risk of shorting out exposed circuits on trash and shit cause youre lazy.

...I was just going to mention that I have three desks in a similar state...definitely not going to admit that now.