Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Electronics designing for 1970s automotive OBD

  1. May 30, 2017 #1
    Hi to all!

    I´m Kai from Finland and me and my step dad are working to make an OBD system for 1970´s Chrysler vehicles. The company itself made one during the era and we would like to have same kind of system, only with 2010 equipment. (so no fridge- size computer for my shop)

    We´ve solved most of the continuity and voltage measurements, but there are a few tacks we are fighting with.
    First problem is the ampere metering for starter motor: it should be able to test amperes from range 0 to 700 for cranking current evaluation and alternator output. I know there are shunts to be used in parallel with the ampere meter but they seem to have very limited ampere suppression.

    Second question is that we need to get multiple information on one wire: there are only 7 wires on the OBD port for 50-60 different functions. What can be done to separate multiple analog signals from one wire? Most of the components being tested have separate grounding circuits if that helps anything. I´ve heard of multiplexing but don´t know if there´s and easier way to get into same result.

    We were thinking of using a PC board in series between the OBD port and the laptop. I do the schematics and building and my step dad does the programming for the system diagnostics but neither of us is an electric engineer so I thought this might be the place for help.

    Thank you for reading and hope to hear from you!
     
  2. jcsd
  3. May 30, 2017 #2

    anorlunda

    User Avatar
    Science Advisor
    Gold Member

    :welcome:

    OBD = on board diagnostic , correct?

    You may be right. I easily found 500 Amp 50 mV shunts online, but not 700 amps. But I did find 1000 Amp precision resistors. Because the currents are so high, safety is a big concern. A melted component carrying 700 amps is a serious hazard. Do you know the length and guage of the wire supplying current to the starter? If so, you should be able to find the resistance of the wire and use that as the shunt. Fewer components means safer and more reliable.

    In today's world, converting your analog signals to digital may be the easiest way to go about it. I think others here at PF can advise you on that.
    If you stay with analog, you must guard against induced voltages by using twisted pairs or coax cables.

    If you use a standard platform such as Rasberry PI, you may find many devices, boards, and mountings suitable for auto applications. I recommend that you post your questions on one of the forums dedicated to that platform. If you're lucky, you may find others who have done similar projects.

    Good luck on your project. Be safe.
     
  4. May 30, 2017 #3
    Thank you for replying. I do know that the wire that is spliced to the OBD wire is after the shunt of the vehicle´s own ammeter. My step dad pointed out using a microprocessor in the P.C board that converts the signals into digital form and therefore, more convenient for him to program into. I´m not sure which platform he was planning to use but I trust him with what he´s going to do.
     
  5. May 30, 2017 #4
    Last edited: May 30, 2017
  6. May 31, 2017 #5

    anorlunda

    User Avatar
    Science Advisor
    Gold Member

    As I said in #2, the easier way in today's world is digital. It is also probably much easier to forget the idea of one wire and to use a separate twisted pair for each signal.

    But I also said that you can get more practical advice on an antique car forum than here on PF. You may be able to find someone who already did what you want to do. I stand by that advice.

    Lycka till
     
  7. May 31, 2017 #6

    Averagesupernova

    User Avatar
    Science Advisor
    Gold Member

    What you want to do is a pretty big undertaking. I am not sure you grasp the complexity of it.
     
  8. May 31, 2017 #7

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    This morning at your photobucket link i was able to see several pages of that Chrysler diagram but now i only get one.
    There are three unlabelled pins in your diagnostic connector , on a page i can no longer see.
    Which i guess are for communication by computer.

    You'll have to get that computer's communication protocol from someplace if you want to talk to the computer.

    Starter current is the voltage drop along this wire


    MoparDiag.jpg


    which appears to be a red #6
    which is 0.3951 milli-ohms per foot at room temperature.
    That'd be 0.277 volts per foot at 700 amps, a bit more at engine compartment operating temperature.

    I suggest you trace out the wires in your diagnostic connector and make your own drawing of where each comes from.
     
  9. May 31, 2017 #8
    Good morning to all!

    I tried to discuss about the matter on a vintage Chrysler forums, but even with the schematics as evidence for the analyzer existance, all the members said there never was an OBD plug-in on cars of the era.

    Even with OBD programmers forums, they are more about coding and programming existing units and don´t seem to know much of the hardware or the electronic engineering of the systems. That´s how I ended up here.

    I checked my photobucket link and at least I saw all of the 9 pages I uploaded. They should be free for all to browse.

    I know this would be much easier with CAN- lines (twisted pair cable) but, as I spoke with a Canadian restoration shop, they said their clients don´t want to pay or use time for aftermarket installation of the lines, rather they´d like to speed up the diagnostics progress by using the vehicle´s own existing system.
     
  10. May 31, 2017 #9

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    Well as interesting as this thread is

    upload_2017-5-31_23-44-58.png

    i'll try from my laptop in a day or two
     
  11. Jun 1, 2017 #10

    anorlunda

    User Avatar
    Science Advisor
    Gold Member

    Now I'm confused. What does that have to do with sending multiple signals on one wire?

    Is your goal to diagnose problems or to learn about cars/electronics/computers by building your own OBD?

    @Averagesupernova 's point is a good one. It will take you 2000 to 4000 hours of labor to design/build/install/test a system to send multiple signals over one wire in a car. Even so, you would learn much engineering. So once again, you need to focus on your objectives. Is it the end product, or the learning as you go?
     
  12. Jun 1, 2017 #11
    Do the schematics show with new link? http://s612.photobucket.com/user/Kaitsu93/library/

    The reason I´m undertaking this is because I am a Volvo mechanic and work with old Chrysler cars as a side job. I saw that those cars have a port for testing engine components and run basic tests. I thought it would be a lot quicker for one to diagnose a malfunction with a computer and doing something else meantime, rather than using a multimeter to test them one by one. I spoke with my Canadian colleague about the matter and he said what I posted above.

    However, some of the schematics show that one circuit splices into different components, such as circuit J2 14DB that goes through ignition coil, ignition resistor and EGR timer and solenoid. I know the solenoid can be measured for continuity and that´s enough info for that.

    But, the same circuit should also measure rpm from coil as well as primary ignition voltage. The same circuit also runs through ignition resistor that should be measured for resistance. the diagnostics plug only has one ground wire that can be used.

    That´s what I meant by measuring multiple things from one wire.
     
  13. Jun 1, 2017 #12

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    yes, Thank you very much !

    Pre OBD2 (1996) manufacturers were secretive about their system. And they used their own brand specific connectors.

    But they co-operated with tool suppliers.
    http://buy1.snapon.com/products/dia...CS(EAZ0025B01CRevB)/06 ChryslerOperations.pdf

    Every real world computer project starts with a signal list.
    List everything that's available on your diagnostic plug and figure out how to interface to the analog ones like starter current(google "High Side Current Sense" ),

    Then you'll have to research to find how that OBD1 computer communicates. RS232 can be done with just three wires but i have no idea what Chrysler did. Looks like OBD1 came into existence around 1983 - how would a bunch of practical engineers of the day done it ?

    Here's an outfit that makes interface IC's. I'd wager some old timer there can point you to how Chrysler's old timers did it.
    Note the references to RS232, which is what we all spoke in the early 80's.
    https://www.elmelectronics.com/products/ics/obd/
    Don't forget Arduino forums https://en.wikipedia.org/wiki/OBDuino

    Good Luck ! (I have a little 95 Dodge Spirit so am hoping you solve the interface puzzle)

    old jim
     
  14. Jun 1, 2017 #13
    Wow, thanks a lot Jim! It all seems to have gotten into history as Chrysler or Huntsville electronics division didn´t patent their invention, at least I can´t found it on US patent office web site. What I do know is that they planned to make add-on to the analyzer that could test and report body and convenience electrics as well, however, the computing and electronic engineering caught fire on mid- 1980´s so they did a total re- model that was called "OBD1".

    I´m sorry if after all the forums even this might not be the right one for this conversation, but this has been the most open- minded and helpful so far.
     
  15. Jun 1, 2017 #14

    Averagesupernova

    User Avatar
    Science Advisor
    Gold Member

    This thread is confusing to me to say the least. The op starts out by saying Chrysler had an OBD system in the 70s which I was unaware of and didn't find support to this claim on wiki.
    -
    I am not sure if the op wants to design the whole system or just communicate with an existing system that I was unaware even existed. Exactly what parameters are to be supervised? Measuring starter current is nothing that cannot be watched on a gauge.
     
  16. Jun 1, 2017 #15

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    It may not exist
    references i found to OBD1 start in 1983 model year

    this may be a plug for analog testing of "won't start" problems that plagued Chryslers in early 70's , notorious for burnt out ignition ballast resistors.

    I assumed the three unlabelled terminals near center were for RS232-like communication
    but now that the rest of the diagrams are available it looks as if they might be not even wired !
    Perhaps Kai will say for sure.
    upload_2017-6-1_15-14-55.png

    Which means he can do interface all analog , read the conditioned signals into a convenient microcontroller and program away. Good chance to learn something like an Arduino or Micromint hobbyist computer.

    Or just a panel with some lights and meters.

    I had hoped to learn the MOPAR OBD1 protocol for my old('95) Dodge but it looks as if his vehicles predate OBD1. So it goes.

    it will be straightforward to interface those analog signals.

    Good Luck guys, keep us posted ?
     
    Last edited: Jun 1, 2017
  17. Jun 1, 2017 #16

    Tom.G

    User Avatar
    Science Advisor

    The photobucket images you uploaded of pgs 8-166 and 8-231 clearly show no computer connection to the diagnostic connector, there are only the six analog testpoints and Ground.

    Error codes before the official OBD definition were frequently reported by flashing the Check Engine light. But If I recall correctly, even those only go back to the 1980's For some details on the 1980's technique see:
    https://www.freeautomechanic.com/diagnostictroblecodes8.html
     
    Last edited: Jun 1, 2017
  18. Jun 1, 2017 #17

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    I loved those late 80's to mid 90's Chryslers - turn the key four times and check engine light gives you the codes. Ones with electronic odometer displayed the codes there.

    How practical .

    old jim
     
  19. Jun 1, 2017 #18
    The only proof of the original computer can be found on digital newspaper archives. Google "Chrysler Electronic Engine Performance Analyzer" to see just some brief- views of the system.

    Those three blank connectors were used as a spare for future expansions as I mentioned above, such as AC and other not so important electrical circuits. However, it was dumped from the way of the OBD1.

    For supernova, I try to communicate with the existing system so that it will work on my and my clients cars (I have a 1980 Dodge Aspen for reference and bench- testing).

    I got some of the continuity measurements done, but I started to stumble on the circuit that measured resistor, RPM, EGR system and voltage regulator as I tried to do a test run with my multimeter. Even if the EGR timer or solenoid are broken and therefore not connected to ground via OBD plug, the multimeter signal still gets to ground from someplace else in that same OBD connector. I hope I explained the problem good enough:)
     
  20. Jun 1, 2017 #19

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    So Done.
    Hemmings says about 1979 Town and Country
    https://www.hemmings.com/magazine/h...1981-Chrysler-Town--amp--Country/3734471.html


    and from http://libwww.freelibrary.org/digital/item/zoom/39904

    MoparOBDZero.jpg


    Call it OBD-0 ?

    I have difficulty following the wiring diagrams page to page with some of them upside down and others blurred at the edges . You'll want your own simplified wiring sketches for each of the six signals available so as to know exactly what each voltage is telling you ... Bad ground connections , common on old cars where roads are salted in winter give truly confusing symptoms.

    I do hope you guys build one.
     
  21. Jun 1, 2017 #20

    Averagesupernova

    User Avatar
    Science Advisor
    Gold Member

    There is a huge difference between an engine analyzer and a computer that hooks up to an on-board computer to communicate with and read back stored codes. Engine analyzers of the day read back information in real time. They read things out such as RPM and dwell (google it if you don't know). They would display charging voltage, secondary ignition voltage, often had a timing light that got the signal off of a plug wire. They often would have a scope display that would show the secondary voltage at one or multiple spark plugs. You could diagnose and determine a lot of what was going on in an engine from the way the voltage would rise up on a plug wire. These machines had one or more vacuum gauges to aid in tuning. They would have load resistors for testing the charging system. I think you are confusing several different automotive diagnostic tools.
    -
    Chrysler was likely late in the game in engine analyzers. Here are some other examples:
    https://www.google.com/search?q=sun...p57UAhVqzoMKHdLAA8QQ_AUIBygC&biw=1366&bih=622
    https://www.google.com/search?q=all...p57UAhVn9IMKHXuqBMwQ_AUIBygC&biw=1366&bih=622
    -
    This was a popular one for do it yourselfers:https://www.google.com/search?q=pen...gC&biw=1366&bih=622#imgrc=_&spf=1496377528651
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted



Similar Discussions: Electronics designing for 1970s automotive OBD
Loading...