Commented ATZ command returns no data

[devoncrouse]

When testing the module as-is, and setting a poll on rpm, I get OKs for the sets, and then empty objects for all of the poll results:

{ value: 'OK' }
{ value: 'OK' }
{ value: 'OK' }
{ value: 'OK' }
{ value: 'OK' }
{ value: 'OK' }
{}
{}
{}
...

Based on testing/reading, it looks like the problem is the ATZ (ELM327 chip reset) being commented out. Is there a reason for that? Things work for me when it's un-commented.

        //self.write('ATZ');
        //Turns off extra line feed and carriage return
        self.write('ATL0');
        //This disables spaces in in output, which is faster!
        self.write('ATS0');
        //Turns off headers and checksum to be sent.
        self.write('ATH0');
        //Turns off echo.
        self.write('ATE0');
        //Turn adaptive timing to 2. This is an aggressive learn curve for adjusting the timeout. Will make huge difference on slow systems.
        self.write('ATAT2');
        //Set timeout to 10 * 4 = 40msec, allows +20 queries per second. This is the maximum wait-time. ATAT will decide if it should wait shorter or not.
        //self.write('ATST0A');
        //Set the protocol to automatic.
        self.write('ATSP0');

[devoncrouse]

I wish I could make my OBD adapter generate fake data without being hooked up to the car...

[EricSmekens]

That might be it. I normally set up another laptop with OBDsim. https://icculus.org/obdgpslogger/obdsim.html It's kinda hard to set it up for the first time, but it's worth it. It just fakes your laptop as an ELM327 device.

I will try this out this weekend too, might be the case. In general, the whole module should be remade with callbacks maybe, or in any case, it shouldn't have timed polling. But that's the only way to get much results in a second I believe, as with OBD, you have to poll.

When I have time, I'm going to look into this, but time is spare. :/

[devoncrouse]

Yeah, I hear that. Thanks for the tip on obdsim; that'll help. And I'm getting what I need out of the module now more or less, so no rush.

[timkendall]

I'm having the same problem :\ Tried the above solution of un-commenting self.write('ATZ') and still no luck. What device are you guys using (I'm using this)?

[EricSmekens]

What car (year/make) are you trying it on? Does an app like Torque works with it?

[timkendall]

Tried a 2002 Acura TL and got no data. Tried it with a 2009 G37 and got data like this

Debug: Found device: OBDII (00-0d-18-28-08-11)
Debug: Found device channel: 1
{}
{}
{}
{ value: 'OK' }
{}
{ value: 'OK' }
{}
{}
{ mode: '41', pid: '0C', name: 'rpm', value: NaN }
{ mode: '41', pid: '05', name: 'temp', value: 70 }
{ mode: '41', pid: '04', name: 'load_pct', value: NaN }
{}
{}
{}
{}
{ mode: '41', pid: '0C', name: 'rpm', value: NaN }
{ mode: '41', pid: '05', name: 'temp', value: 70 }
{ mode: '41', pid: '04', name: 'load_pct', value: 22.265625 }

...while running this

btOBDReader.on('connected', function () {

    this.addPoller("vss");
    this.addPoller("rpm");
    this.addPoller("temp");
    this.addPoller("load_pct");
    this.addPoller("map");
    this.addPoller("frp");

    this.startPolling(1000); //Request all values each second.
});

Obviously something is a little off (why would it only get a few intermittent values?). And no, I don't have an Android phone to try Torque or a similar app with. Any idea? I was thinking it could be the baud rate but I don't see how to change that.

Thanks for making this library btw, super cool!

[timkendall]

Hmm, weird - it seems to work consistently with the 2009 G37 if I connect to the device before the car is on.

[EricSmekens]

I don't think it's the baudrate, you are getting some of the messages, so that seems to be good. Might be something with the polling, we only see rpm, temp and load_pct returned. Maybe you could try with:

this.requestValueByName("vss");

That way we can see if it's not a timing issue.

Yeah, the library could use some more revision. I made this librarby in my early JS days, so there may be a few things that are strange. Especially on the polling thing.

[timkendall]

Hmm alright, yeah I'll play around with it a little more.

[EricSmekens]

Closing issue, will make a list of upcoming changes soon, and this will also be mentioned.