Thread: Questions before purchasing

First I would ask you to check out this thread, and specifically #6 (post linked to in the thread): http://www.mp3car.com/lcd-display/15...ml#post1488680

Originally Posted by os12tr

what is a basic schematic to wire in a variable resistor for measurement? (for example a fuel level sending unit)

Provide a bottom or top resistor, and create a voltage divider. Scale the resistor to give the highest range between 0v and 3.3v when your gauge is full to empty.



Your guage resistor will be either Z1 or Z2. The other resistor you select will complete that diagram.

Voltage Dividers: http://en.wikipedia.org/wiki/Voltage_divider

Originally Posted by os12tr

A basic schematic for a potentiometer ( a pressure sensor)

As long as it is between 0v and 3.3v, wire it straight up. If it is not, then attach another voltage divider to it (a potentiometer is a variable voltage divider) to scale the output voltage.

Originally Posted by os12tr

A basic schematic for a hall effect sensor ( a crank or cam sensor)

This requires on the special boards I mentioned in the first thread's link. Polling this data is not reliable, so I have made a board that looks just for that signal

Originally Posted by os12tr

A basic schematic for a on off signal (such as an input for turn signal indicator or check engine light indicator)

On would give you 12v, off would give you 0v. On some cars this is reversed, but basically there will be ground, and something much higher than ground. Use a voltage divider, and wire it in like the potentiometer

Originally Posted by os12tr

Is it possible to use dash command with the fusion brain?

I have no idea. There are plugins for RR and CF as well as standalone software both available from me and 3rd party.

Originally Posted by os12tr

Is there a document or tutorial for the questions above?

Thank you for your help.

The forum has a lot of this information already, but it can be hard to find. I try to answer questions here as fast as I can as soon as I see them.

Note that if the crank/cam sensor is Hall Effect, then it will be a digital output. But many have changed to reluctors which have the usual AC output and usually require a "negative going zero crossing" detector to provide the digital output. Hall sensors are analog as is the reluctor, but Hall Sensors as used in cars are a Hall Sensor Module that provides a digital output. (I love the old Hall is digital versus reluctor is analog and hence Hall is more accurate argument. In fact the opposite is true!)

And yes, these need to be interrupt sensed for ignitions etc, else clock a timer if wanting a rate or count function.

Thank you for the replies. I'm going to get some wiring diagrams for my car and see how things are wired to the factory dash.

Originally Posted by os12tr

Thank you for the replies. I'm going to get some wiring diagrams for my car and see how things are wired to the factory dash.

If you can find them for free, then go for it. But otherwise it might be useless. Generally they show block diagram wiring. So the fuel sender is a box with a wire coming out just like the wheel speed sensor is a box with some wires coming out. They do not divulge the technical aspects of what signal that wire is carrying, just the existence of the wire.

When I do this for customers, I bring out a large suitcase full of test equipment and a couple oscilloscopes with buckets full of parts to get the signal correct. IMHO, that is what you need if you are looking at doing this yourself. Some companies will rent oscilloscopes.

I can go to UTI and use alldata and mitchel for free.

mitchel - as in Mitchell CDs etc? Great stuff! Especially for our PC-based documentation.

Just re pulse signals like indicators/flashers etc...

Older vehicles have thermal-type analog gauges for fuel, temp, oil-pressure etc. Even older ones have electro-mag gauges (that bump & flutter..)
These AFAIK always have voltage regulators for the fuel & temp. (Oil gauges are often across the raw +12V)
But those voltage regs are usually electromechanical (thermal) pulse types akin to older flasher cans. (Brilliant technology - mechanical(!!) but stable, reliable, often adjustable, and makes gauges work with a reverse connected battery without damage!)

They typically pulse at about 2Hz to give an average voltage of ~8V.

Sensing issues are solved with an RC filter, ie thru a diode & resistor to a resistor-capacitor combination that provides a time constant of a few seconds or more (enough to give a suitably averaged figure).
The same can be used to sense if an indicator/flasher is on, but because you don't want the "is on" signal lasting for seconds after cancellation, the RC is usually set to 1sec or less and the circuit merely senses a voltage above (say) 1V etc as opposed to wanting an accurate voltage from gauge sensors.

Someone recently posted that such gauges are current types - which is true. But IMO that is not an issue - there is still a voltage (even if pulsed) across the gauge or sensor/sender. (Was it 2k1Toaster, or one of the other greats that posted that? I meant to comment or PM at the time, and now can't find the thread.)

I'm about to do similar for an electric fan sensing circuit and an indicator/flasher-reverse lamp combo (equivalent to stateside stop/flashers etc).
My temp gauge is a classic vibro-regulator with thermal gauges as used from the 1960s until "electronic" gauges replaced them, though many new vehicles still use them (though I wonder if the e-mech v.regs have been replaced with electronic regs?).
Now I feel confident that what I say is correct - ie, can get a voltage output from a (pulsed) current meter - however I can't recall actually doing it... But I'm certain I have seen it even if I didn't do it. But if confirmation is needed...

The only drawback is that the series diode (that stops the "sample and hold" capacitor discharging back through the gauge or sender) means a voltage drop of ~0.6V, or ~0.3V if a Schottky diode etc.
Whilst that isn't a problem for voltage-level or set-point sensing (as used for cooling fans etc), it does effect gauge reading linearity.
It may simply mean the gauge voltage is simply compressed from the ~0.6V diode drop, but it might also add non-linearity.
However often sensors and their gauges are non-linear anyhow and hence require mapping.
And good designs will usually provide for mapping - whether by calculation or a look-up table (with iterative calculations) - even if it is a linear relationship.


Sorry of the above bamboozles you, and it has no impact if you have electronic gauges etc, though it may be useful for flasher sensing. But this reply is also in part for others.
2k1Toaster probably has this covered elsewhere (sorry - I'm way out of date re the FB) and should be able to provide calculation details for the capacitor value (where its time-constant R is the input resistance of the FB etc, as opposed to the series input Resistor which is only to limit the capacitor's inrush current and should be well under (say) one-hundredth the FB's input resistance).

I won't go into why this RC hardware solution is (IMO) superior to uPC signal averaging, but IMO unless the pulsed input is interrupt sensed, it isn't difficult to see the programming complications.



PS - for stop/flasher and similar combinations (like parker/clearance - flashers etc), I was recently made aware of exLED's TPC v.2 2Color (2Way) Turn + Position {clearance/parker} (also TPC v.3 for R-LED) modules.
I also added some comments about its use on the relevant thread on the12volt's drl relay.

Thank you for all the info. I'm going to have to do more research after I get wiring diagrams for my car. 2003 toyota camry le 4cylinder.

I have everything figured out except how to display and store odometer readings. Any suggestions would be helpful. Thank you

Originally Posted by os12tr

I have everything figured out except how to display and store odometer readings. Any suggestions would be helpful. Thank you

The method we use is taking the wheel sensor and counting the pulses. If you know the outer diameter of your tire, and how many pulses per revolution, you can work out the distance driven on a pulse by pulse basis. The wheel speed is usually a hall effect sensor, low small signal AC, that we filter and apply a crap-top of gain to and get a nice square pulse we count and send out over USB CDC as a standalone odometer board.