Looking for good references on signal conditioning...

All,

Does anyone know of a good reference on signal conditioning in automotive environments?

I need to construct some input circuits for a gasoline engine management system that I am working on. I have found some IC packages (and read their data sheets) and I have also found some very simple approaches using only a single resistor and a couple of capacitors (though this was used on a system that had an external Analog to Digital IC package, external that is with respect to the main processor).

In my case, I am using the Motorola MPC555 which has all of the AtoD stuf built in, but I don't want to run "raw" sensor output signals directly in to my MPC555 (for fear of melting it).

FYI - if you couldn't tell fom reading my request, I am not an electrical engineer nor do I develop embedded systems for a living. I spend my working hours as a medical imaging software jockey and I love cars.

Any suggestions? Thanks in advance for any pointers,

-Paul

Reply to
Paul Helmuth
Loading thread data ...

Not automotive in particular but the series of four that I wrote for Forthwrite (a magazine for Forth enthusiasts) dealt with real world interfacing techniques. Many of them will stand you in good stead. You should probably read it in coinjunction with a good electronics reference book like Horowitz and Hill's epic tome.

My series is on my website at:

formatting link

The basic techniques have stood me in good stead in far harsher regimes than automotive engine bays.

--
********************************************************************
Paul E. Bennett ....................
 Click to see the full signature
Reply to
Paul E. Bennett

Do not forget to look up protecting Auto electronics. The car is a hostile place for any circuit.

Reply to
Neil Kurzman

You won't find any good references for the info you are seeking, sensors and a/d's vary so widely that a generic reference book can't be written. The best info is from the datasheets for the sensors, they have typical interface circuits. For many sensors there is nothing wrong with wiring them directly to the A/D input of the 555, as many sensors today are made to do this (the sensor itself has an internal buffer with low source impedance). The single resistor and cap you saw in the datasheet was probably for anti-aliasing, but if your running the 555 a/d fast enough, you may be able to delete them too. Replacing the cap and resistor with a opamp would be needed if the sensor has a high source impedance (>2.5 Kohms for instance), or if you need a more accuracy (if the a/d has high leakage current), or you need a dc bias on the sensor output to accomodate the range of the a/d, or if you need a more complex filter, or to gain up or gain down the sensor output to the range of the a/d input. The more compatible your sensor output is with the a/d input the less signal conditioning you need. By the way, the 555 is a very complex chip, not the best micro to start out with (but it does have a nice floating point unit!).

steve

Reply to
steve

automotive

their

only a

system

stuf

in to

electrical

working

I think you need to look a bit deeper into the issues involved than Steve suggests. There are two problems - the first is making the sensor signals compatible with the A-D input on the processor which Steve addresses. The second issue is the EMC environement under bonnet (UK terminology). The specs are available if you hunt but a good starting point is that any pin on your completed black box must withstand being sparked with an 8kV spark from an EMC test generator (ISO human body model). I've found that systems that are OK with the spark are usually OK for everything else and if thay fail the spark they fail in EMC test and in real life. At first this requirement tends to induce horror and panic but in fact you can often meet it with 1nF to ground on the input pin, 47k in series and then annother 1nF to ground. Put the filter components as near the connector as you can - use 4 layer board with full ground plane if possible. (You can meet specs with 2 layers but it usually takes several iterations of board layout/filter values etc.)

If this design is a one off/low volume then buffer every signal into or out of the micro so you can replace a cheap buffer if something blows.

You can hire a spark gun cheaply (£50 a day or less in the UK) and its worth doing if there are any penalties (like wrecking your engine) if the system fails in service.

Good luck !!

Michael Kellett

Reply to
MK

Except he already has.

Obviously Steve hasn't worked with anything other than simple, cheap consumer grade stuff where the environment is more benign. I already know that Paul Helmut (the OP) found my series of articles quite useful for his purposes even if they didn't offer him a completely readymade solution (he emailed to say so).

As I have dealt with environments that are extremely demanding I tend to look at the interfacing of sensors and actuators quite closely. I know that some of my I/O designs have withstood more than their specifications (due to an accident in the testing phase where someone set the test equipment incorrectly). The basics are, "manage the energy properly and you will not lose your semiconductor devices".

--
********************************************************************
Paul E. Bennett ....................
 Click to see the full signature
Reply to
Paul E. Bennett

Not true, but rereading the original post I obviously misunderstood the original question (stupid me!). Your web pages were nice, but I have spent 4 hours in meetings discussing only wetting currents so I hope the OP knows what he is getting into.

Reply to
steve

After such a lengthy meeting (one just this one topic?????) I suppose we can all forgive the lack of thought you gave to your response.

Now that you have stated that wetting currents was a topic of this meeting, why, we will all want to know, did it take 4 hours?. Even when we were discussing re-use of some old, dirty, train auto-couplers it didn't take us that long to determine the appropriate wetting current minima, even running some testing on the auto-couplers concerned.

--
********************************************************************
Paul E. Bennett ....................
 Click to see the full signature
Reply to
Paul E. Bennett

meeting,

us

running

Hello Paul,

Calm down - its only a news group !

(And if we really wanted to pick holes in each others work Steve could always ask you why there is no common mode HF filtering in your input circuit (fig 1 first article).)

Michael Kellett

Reply to
MK

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.