Date:         Mon, 12 Sep 94 15:38:02 bst
Sender:       Vanagon Mailing List <vanagon@vanagon.com>
From:         A.C.Erskine@lut.ac.uk (Andy Erskine)
Subject:      Re:Computers on Board


>I built as part of a student project oh, many moons ago, a diagnostic
>program that worked on the old, OLD, Timex-Sinclair 1000's. I might even
>have my write up around somewhere.
>
>The program gave a decent approximation of RPMs, Dwell, and some
>other functions I cant remember. It used a 'black box' that hooked onto
>the expantion connector and the points wire of the distributor. It did
>not have enough memory for me to implement the diagnostic functions
>that I had envisioned, nor was the timing accurate enough. It ran off
>of 12 volts, was rigged as hell and impervious to water and grime,
>and used a small 12 volt TV for output.
>
>If I were to to it again today, I would try to use a cheapo 8086
>portable and implement the following functions:
>
>Tach
>Dwell
>degrees between firing of successive cylinders
>degrees between firing of same cylinder (timing fluctuation)
>Inductive pickups on all four spark plugs. These could be cross
>	referenced with the points to indicate no-fires
>TDC sensor with firing informaton to indicate advance. Reports
>	advance curve under various driving conditions.
>Oxygen sensor, mabye to indicate mixture under various conditins.
>
>If you have any suggestions, send them my way, or post them. As the list
>stands right now, the black box would not be very expensive. All it would have
>to do is read the various sensors, digitize the results, and serially 
>send them to the host PC. The host PC would be able to do all the analysis
>of the data in real time, giving you information as you drive, or in the shop.
>
>All of the sensors, (esp. the four inductive pickups) would be more expensive
>than the 'black box' reading equipment. The pickups, O2 sensor, TDC sensor,
>and some mounting brackets would cost about $100.00. I cant cost the rest
>of the system, and the computer would be up to you. The hardest part would be 
>making all the logic operate reliably on 12 volts for long periods. 
>
>I have always been intrigued by the big computers at service bays, but
>I realized that the snesing hardware is trivial. All of the cost of these
>machines, mostly leased, is the vehicle tune-up database that they carry.
>
>So, any ideas?
>
>--- Chris Chubb (cchubb@ida.org)


One diagnostic method that has been used on large (that's BIG marine or 
power generator size) diesel engines for years is to look at cyclic speed 
fluctuation. This is the variation in crankshaft rotational speed with 
crankangle. You can work backwards from this data to generate a cylinder 
pressure diagram from which you can determine all sorts of mechanical and 
combustion problems. This facility can now be built into ECU's so that your 
car can shut down and tell you what is wrong before anything catastrophic 
happens  (or more likely, it will tell the dealer what is wrong so he can 
charge you an arm and a leg to tell you what is wrong). I work on something 
similar and know at least one major car manufacturer intends to introduce this.

Andy E.

A.C.Erskine@lut.ac.uk