MR2 Owners Club Forum banner

1 - 11 of 11 Posts

·
Premium Member
Joined
·
12,817 Posts
Discussion Starter #1
I was trying to do some logging today, and I keep getting big gaps in my data. Like it's dropping the connection or something. For example, logging a 2nd gear pull, it will log up to 2800rpm, then have no data until 4200rpm. On another pull it will be fine through that range, but drop everything from 6200-7000. I did four pulls total, and every one was missing at least a thousand rpm worth of data. It's an OBD Link LX logging to a Pixel 2 with, as I mentioned, Torque Pro. I have never had this problem before. Anyone seen this before? Any ideas?
 

·
Registered
Joined
·
5,948 Posts
Any chance you can try using a different device? That's all I can think of other than a loose connection at the obd2 connector or ecu plug.
 

·
Premium Member
Joined
·
12,817 Posts
Discussion Starter #3
Yeah, I can try with my tablet. That and wiggling the OBDII connector a little. Also I was reminded on Facebook that I don't have my CAN lines properly terminated with the 120ohm resistor that they are supposed to have. Seems unlikely that it would have worked perfectly for years without and all of a sudden would be a problem, but it's worth fixing anyway I think.

The thing that gets me is that it has worked up to now, so the more I think about it the more it seems like maybe just a loose connection or something. Or an update on my phone broke the bluetooth...
 

·
Premium Member
Joined
·
3,556 Posts
I've never had that issue with OBDII but I've had it on the dyno before from ignition noise.

Do you have your ignition filter caps on? are you running resistor plugs? and do you have proper CAN termination (measure resistance between CANH and CANL without the OBDII reader plugged in, it should be about 60ohms)
 

·
Premium Member
Joined
·
12,817 Posts
Discussion Starter #5
I've never had that issue with OBDII but I've had it on the dyno before from ignition noise.

Do you have your ignition filter caps on? are you running resistor plugs? and do you have proper CAN termination (measure resistance between CANH and CANL without the OBDII reader plugged in, it should be about 60ohms)
None of that stuff has changed since the last time it worked, BUT, it is true I don't have the proper termination. I didn't realize you were supposed to when I wired it originally, and it always worked fine so I haven't bothered to go back and fix it.

I did find a way to graphically show the problem however. I tried out the official software from OBD Link, and while it didn't solve the problem it does time stamp data bit differently making it easy to see the logging rate, or the delay between data points, depending on how you look at it. Here's a graph from a few minutes logging in the garage, engine off (so, this rules out ignition noise as well).

Typical delay is about .1s between data rows (logging 4 pids), but every 40-50 rows it jumps to over a second. It's not perfectly consistent, but there is a pattern to it.

70112
 

·
Premium Member
Joined
·
3,556 Posts
Hmm, that's a cyclical pattern. I think there's a decent chance it's a software issue on the adapter but i would probably power it with something other than the car's power supply to run a quick test.

Odds are the adapter knocks down the 12V to 3.3V or 5.0V so you can likely power it from a 9V battery for a 10 minute test but a power tool battery can also work really well and many of them can fit spade connectors directly.

Actually, before all that, have you tried it in a different car?
 

·
Premium Member
Joined
·
12,817 Posts
Discussion Starter #7
Actually, before all that, have you tried it in a different car?
My readily available other cars are on the older side (a 99 and an 03), and neither puts out more than about 4 PIDs per second, so they aren't great test subjects, but there is still some information there.

Here's what I get logging my wife's '03 Subaru with my Pixel2.
70196


To try to rule out a device (phone) problem I tried my old Galaxy S5.
70197


Better, but not perfect.

Back to the MR2, here's two more devices (old phone, and tablet).
70198


70199


The spikes aren't as high, but still present.

As a final reference, this is what it's SUPPOSED to look like, from one of my logs back in January.
70200


So, what changed? It certainly seems to not be vehicle dependent, and while it seems to be better with some devices than with others, changing devices doesn't take it back to what it was 4 months ago. So, random hardware problem with the adapter?
 

·
Premium Member
Joined
·
12,817 Posts
Discussion Starter #8
For what it's worth, I added the 120ohm termination at my OBDII connector, but it didn't make any difference to this problem. At least that's done correctly now...
 

·
Premium Member
Joined
·
12,817 Posts
Discussion Starter #9 (Edited)
Well, after all that (and more than a dozen emails back and forth with OBD Link support), they concluded that it was an android problem. So I downloaded their OBDWiz windows software and tried logging with my laptop (which luckily has bluetooth). Initially it appeared to be much worse overall performance as I was only seeing about 15 PIDs per second until (on a complete guess) I turned off the laptops WiFi, and after that I was getting the same 55-60 PID / second that I usually see on my phone. Graphing logging rate the same as before, it's not perfect, but it is a lot better than what I have been seeing. Unfortunately still not quite as good as I was seeing back in January though. I am using a different firmware on the adapter than I had then, so MAYBE there is a speed difference due to that, but I'm not terribly hopeful. Overall, at least I can log stuff again, but it's awfully frustrating. The laptop is a lot less convenient than the phone. And the performance is still a bit of a downgrade.

70290
 

·
Premium Member
Joined
·
12,817 Posts
Discussion Starter #10
I think I finally figured it out. The display on the phone has to be OFF. As soon as I lock the phone and the display shuts off, the 1 second delay every 3 seconds goes away. As soon as I unlock the phone, it comes back.

I think back in January when I was collecting OK data, I was logging for longer time periods and it was turning off on it's own. Last month I was just trying to log a single pull, and the display was staying on. And then in all of my testing since then, I never tried with the display off until accidentally trying it that way today.
 
1 - 11 of 11 Posts
Top