Modern cars have sensor information available in the diagnostic plug, but older cars don't have neither the plug or the info.
But in some cases they have some sensors or some mechanical device that could help to provide important info with small tweak, for instance breaker points can provide RPM info with a small electronic help.
I believe that this can have several approaches.
From a point of view of a classic car owner that likes the car for what it stands for I believe that the electronics are welcome to assist maintaining the car as it is, and to confirm the running conditions.This means that all the readings should be available in a proper way, and in the classic car universe that means a dial with a needle.
There are "new" sensors that can be fitted in classic car to help improve their lifespan, and to improve the efficiency in a way that will be good for the "pocket" and for the experience of owning a classic car.
For those who consider that a small change a blasphemy this ain't for them.
But in classic car world we also have the "racers". Guys that love to improve the engine and everything that they can put their hand on. For those the only solution is to complement the sensor readings with actuators.
In a first approach this aim for classic car owners with a "realistic" budget.
The readings that I consider essential in a car:
Temp: air, water and oil
pressure: oil , manifold
rpm: engine
To display the information A LCD with a background image of the dial and a servo with a needle on the top of the display.
A switch to change the background image.
I believe that the arduino can read 6 sensors and actuate a servo without any trouble.
For the workshop days we could put all this information in a android to help tune the engine with a opened bonnet. For MOT, track day, exhibition event, what ever.
Very poor quality images of what i pretend!
Oil pressure
As a first approach I have 7 readings and that would means 7 dials. And some times there is no space or that isn't just "proper" to fit in a classic dashboard.
But it can be done. I believe that your suggestion is more simple to implement but the trade off will be a very large instrument cluster. Some folks like that approach a dashboard full of dials
My first milestone is to build a tool to help Classic Mini owners of forumini.forumeiros.com to maintain their cars.
I will try to translate electronic/programming talk to "petrol heads" on a DIY section.
The deal is you will help me to help others to enter in the arduino world.
My second milestone is to actuate fuel injectors and ignition coils with arduino
milestone 2.5 - Infotainment - I'm not very keen on infotainment the sound I love the most is induction noise. But nowadays is required for younger people!
The only thing I really need is radio with mp3 and bluetooth to answer phone calls to always keep the hands in the steering wheel.
Third milestone actuate "race" electronics like electric superchargers, power jets or extra set of fuel injectors or nitro.
The things that a weekend racer love.
Most of the analog gauges you mention would actually be better as simple warning lamps, and I suspect the coolant temperature is the only thing which owners would actually need from day to day. What would potentially be very useful though is a 'master caution' sort of indicator which would warn you whether the car was all green (all Ps and Ts in expected range), amber (no harm from driving it but it may be low on fuel or not up to temperature) or red (some fault which requires immediate correct).
You could consider using bluetooth to present more detailed information on a smartphone when the driver goes look for it.
Hi PeterH.
Generally speaking classic cars already have warning lamps. Drive a classic car is more than to drive a car.
It's different. Is more to experience a car that you really like, because you spent time and money to keep it running.
Most of the times the owner actually rebuilds the entire car thats why he needs to be aware of more component status than a "regular" car owner.
Even in a day to day use. Because we can be doing "strange" things like for instance to test a waterless cooling liquid or test different different thermal grade spark plugs,etc.
Believe me, we need more than water temp even in a day to day usage.
I know very well what it's like to drive a classic car, and dismantle it down to the last nut and bolt and rebuild it to better than new, and to develop and tune and maintain it when it is streched well beyond its original performance envelope.
I still think you're mistaken about the utility of most of those gauges. All you actually need to know 99% of the time is that all the important parameters are within acceptable limits. I mean fuel pressure and temperature, air on/off temps, voltage levels, EGTs, WI levels and flow, transmission temps, charge cooler temps, brake temps, oil pressure/temp, coolant temp, AFRs and so on and so on. Nobody actually needs to watch these minute by minute unless you are addressing a specific problem, and then you instrument the thing you're working on. Sure you can log everything, but you don't need to watch these things while you're driving. You really just need an alarm if something goes out of bounds. (In my case that alarm will automatically dump the boost and switch to conservative maps and hopefully save the engine as well as scaring the pants off the driver.) I mean - if you have a gauge showing manifold pressure (I do) then what are you actually going to do about it? In practice if something goes wrong then you inevitably weren't looking at it at that precise moment and need to look at the data logs to work out what was going on anyway. That's my experience. But it's your car. If you want a lot of gauges, there's nothing stopping you from adding them.
PeterH, we both say the same. I'm trying to develop ONE dial, that most of is time will be displaying water temp, I presume.
But, if required that same dial will display another reading with a push of a button.
I like very much your suggestion, to give a priority to a reading that is going out of limits and change the background color and display it. That gives even more value to this project.
When we reach the milestone 2. All this readings will be useful!
I think actually changing some sort of parameter if the target value goes over or falls out of acceptable range is more important than a warning LED- but either will serve the same function.
one being automated, and one being a warning to adjust something manually, like manifold pressure controlled by a boost solenoid or just a warning LED bringing your attention to the boost-creep that is occurring, and advising the user to adjust their boost control accordingly.
i am also going to be working on something like this. i registered to the arduino boards so i can see what people have come up with before my components and my Uno R3 arrive. i am also very happy to see many people interested in car electronic and auxiliary engine management as a viable use case for arduino projects.
I will use the BenF circuit to transform ignition points in a "digital sensor".
I will start by rectifying the wave that is generated by the ignition points to a tension that is in the digital signal range.
I also bought a arduino TFT lcd to display the readings.
I'm waiting for the mail delivery
It might be safer to use that signal to switch a transistor rather than feeding it to an Arduino input directly - that way if anything goes wrong you have blown a cheap transistor rather than destroyed your Arduino.
In file included from /usr/share/arduino/libraries/TFT/TFT.h:7:0,
from sketch_oct03a.ino:1:
/usr/share/arduino/libraries/TFT/utility/Adafruit_ST7735.h:28:26: fatal error: Adafruit_GFX.h: No such file or directory
compilation terminated.
I changed the path in the include to the proper path and then there are a lot of classes and variables inside the blocks with old fashion declaration! They weren't ported to the proper syntax by the RPM packager maintainer!
I use fedora 18 and arduino compiler 1.0.5!
I when to GitHub downloaded the most recent versions and they don't work out of the box!
Worst, some classes tft.h can instantiate!
If someone uses fedora and can give me a clue. It will be very nice!
First, great project. I'm a (relatively) new VW Ghia owner and eventually would like to have computerized monitoring as well. I don't "need" anything but the speedometer, but I would sure like to know more about what's going on back there. Especially as an absolute novice mechanic. Sometimes I don't get the community's hesitation regarding aesthetic or functional decisions that are perfectly feasible ideas. I say, carry on good sir.
Second... Not that I'm an expert by any means, but I know much more about electronics now than I do cars. Do you have a project blog somewhere? I'd love to read along and chime in when I can.
Third... Have you downloaded or Gitted the GFX library? It's a separate thing from the device library. Adafruit built a one-size-fits-all graphics lib they use for all their displays. You shouldn't have to go changing paths in files. AFAIK, there shouldn't be paths in files.
SirNickity
Your car is very "cool". I love boxer engines. When I was a young boy the car who pushed me to automobilia world was the 512 BB.
I dont know nothing about the quality of your car bodyshell, but the engine is "bulletproof".
In Europe, lots of VW boxer owners replace the original engine for a 911 unit. The 911 (953) has a 3l engine with a 4x4 transmission.
Just a thought ... I don know your car chassis!
second - Can track this project from a classic car owner point of view in TMF
Third - neither, I used the libraries available in fedora's RPM package.
But there are, just open a .cpp file
Nevertheless is a good approach that simplyfies a lot
Arduino compiler has serious issues with Fedora, the package maintainer doesn't keep them updated.
The C files and headers are useless.
I downloaded the files from the Arduino repository and everything went fine from that moment on.
This week plans:
Build a break out plug to transistorized ignition (need to buy hardware)
to make final circuit in a prototyping board (buy a smaller one)
Help wanted to:
Build a code assembly to represent a analog dial in the TFT
bichomini:
Today I realized that I need an lambda probe circuit to analyze tension variation!
A narrowband sensor would be useless for practical purposes, and a wideband sensor is going to require a fairly expensive controller. Maybe 'tension variation' is a mistranslation since I can't see how it relates to AFR ratios, but in any case AFR ratios are something that you're only going to need to know if and when you're retuning the engine after a significant modification. It's certainly possible to do (if you're willing to spend the money) but I can't figure out why you would want to.
For "tension variation" I meant that there is a tension value change in the sensor. Depending how rich or lean is the mixture. That's how it relates to AFR. I disagree with you that this value is only useful after a significant engine change. To have this ratio displayed in the tft, it can help to "spice" a track day!
When I'm on a track the last thing I want to do is waste attention on things like lambda readings. Even something as fundamental as a low oil pressure warnings needs a high intensity light in order to gain my attention. But then you already know how I feel about the utility of all these readings you're proposing to display.
I think the site you linked to works on the assumption that there is a wideband controller built into your car's ECU. Does your classic car already have a wideband sensor and an ECU capable of driving it? If not, you're going to need a wideband controller. They can be quite expensive. The vast majority of people who have their engines modified are sensible enough to have them tuned professionally, and of course the professional tuner will provide their own equipment which is of far better quality than the sort of stuff you and I would buy. A small minority who do their own tuning will borrow the equipment rather than buy, because you only need them while you're actually tuning and that is something you only do after modifying the engine and usually do and finish pretty quickly. Only people who have a recurring need to tune their engine would find it sensible to buy their own equipment. I use mine a lot, but unless you're changing the spec of your engine and then retuning it yourself I can't see why you'd need it. You certainly don't want to be watching it when you're on the track.
PeterH:
I think the site you linked to works on the assumption that there is a wideband controller built into your car's ECU. Does your classic car already have a wideband sensor and an ECU capable of driving it?
(...)
You certainly don't want to be watching it when you're on the track.
Another great help from you. I was looking from the sensor signal analisys point of view. Never gave importance to the need of driving it. Now, I'm developing the warm up circuit to heat the lambda probe. And prior to that I will read the signal.
Next week my classic car will have the lambda probe, and I hope that one day Arduino will be able to drive it.
This electronic circuit is for the "do it yourselfers" not for hardcore racers.
From a competition point of view, this is more meant to those who drive the car till the race track.
When reach there open the bonnet, fine tune the engine to have a little more power. And in the end of the event undo everything to return home!
Need help to build the circuit to warm the lambda probe and the circuit to read it.
