Street datalogging

[sltous]

I have been interested in adding some data logging to my car but I am primarily expecting to use the car (and the data logging) on the street rather than on track.  Primary goal is some pre-diagnosing problems so I'd be interested in GPS logging of engine parameters (oil pressure, oil temp, water temp, rpm, TPS, etc) probably with a button to 'tag' unusual behavior for review.

 

For instance: over the weekend I was on a nice drive but found myself forced to stay at very low throttle due to other cars on the road and I was experiencing some kangarooing around 0% throttle where the drivetrain drag would slow me down then applied throttle would accelerate and the cycle would repeat.  Ultimately was able to rev out the engine on a passing section and regained confidence everything was OK but it would have been nice to have a record I could review at home.

 

The car has an unlocked MBE9A4 ECU so it may be that the easiest thing to purchase is an AIM or similar data logger and connect over CAN.  I have been interested in the AIM PDM or similar products so I could get additional data streams for 'free' from a weight and wiring complexity standpoint.  My uncertainty is if AIM software is well suited to street use as the mapping will certainly not be able to find a track for mapping.  It would be a fun addition if the software could recognize the same sections of road and allow me to compare various time spent on the same section over days/weeks/months but this is not a primary goal for me compared to recording for potential problems.

[ashyers]

If you have an MBE9A4 you should be able to look engine parameters there with no issues. You could install a switch to "tag" your data stream if an event occurred.

[sf4018]

The Aim Solo2DL with OBDII connector is gonna be the cheapest Aim option ($700) if you have to have GPS info tied to the data. You can get it to learn a custom route/track, I've never used that feature though. You'll need the MBE9A4 programming software to configure what data is sent out on CAN, then configure Aim to get it. The AiM software is Race Studio 3, you'll have to define a custom CAN protocol on it, it's not too hard. I can send you mine if you need a template.

 

If you don't need the GPS, I would think your ECU software should have a built in data logger.

[Croc]

It should be easier than that.  Set the AIM Solo 2 software to read your Caterham as having a Noble ECU.  Connect the AIM Solo via the J1962 (OBD2 style) plug.  Then it will work as Noble also used the MBE9A4 ECU.  Some of the CAN data channels are spurious but you can filter for that with some study.    

 

Then the data read out can be visualized along with the GPS data as follows:

 

 

 

I set mine for track but you can set a one way route analogous to a hill climb or rally stage.  

 

I don't think the AIM software is smart enough to set triggers for pre-defined limits but I have not played around with it to that level of sophistication.  If you experience kangarooing then note the time and go back to review the data based on that time stamp.  

 

For the kangarooing, keep in mind that for some Caterham models some of the problem was being caused by a weak throttle return spring.  

 

 

[sltous]

Thanks for the thoughts on the AIM software.  I had not considered thinking of street use as a 'rally,' that may help me understand what I can expect from the software.

 

With respect to the kangarooing as a symptom of a soft throttle spring, the roller barrel ITBs seem to have a pretty hefty return spring so I don't think this is my issue but it may be work checking out.

[sltous]

Is there any difference between the Pin 6 CAN Bus High and Pin 14 CAN Bus Low at the OBD2 plug and Pin 9 CAN Bus High and Pin 8 CAN Bus Low at the 36 pins J2 Delphi connector?

[sltous]

On 2/20/2022 at 5:00 PM, sltous said:

Is there any difference between the Pin 6 CAN Bus High and Pin 14 CAN Bus Low at the OBD2 plug and Pin 9 CAN Bus High and Pin 8 CAN Bus Low at the 36 pins J2 Delphi connector?

I inspected with a multimeter and it looks like there is approximately 1.3-1.6 ohms resistance between OBD2 plug pin 6 and 36 pins J2 Delphi pin 8 (CAN high) and similarly 1.3-1.6 ohms between OBD2 plug pin 14 and 36 pins J2 Delphi pin 19 (CAN low).  Although I might usually would like to hardwire for better performance I don't see an obvious improvement over use of the OBD2 plug as I could always cut the cable and hardwire so I have chosen an AIM Solo DL with the OBD2 plug.  Hoping to test relatively soon,

[sltous]

Followup:  I picked up a Solo DL and an OBD2 cable, after remembering I had to actually turn CAN on in Easimap following the instructions on https://www.aim-sportline.com/download/ecu/racing/MBE_9A4CAN_101_eng.pdf I was able to connect wires correctly and get a first data log.

 

I am beginning to wonder if the kangarooing is in fact due to the ignition switch, however, and would like to try temporarily bypassing it with a toggle switch, is there any information on the current rating needed for that switch?  I was getting several (2-5) ohms of resistance in the switch and when I took it apart the insides are looking pretty corroded.

 

 

[Croc]

8 hours ago, sltous said:

...is there any information on the current rating needed for that switch?  I was getting several (2-5) ohms of resistance in the switch and when I took it apart the insides are looking pretty corroded.

 

That does not look good.  There are wiring diagrams in the downloads section of this forum.  Also the Caterham owners manual sometimes had a wiring diagram included.  

[sltous]

Thanks Croc.  I have been concerned by a wiring choice made somewhere along the line where effectively the 12v line to all of the switches was hot even before the key was inserted (allowing me to turn on lights, even spin the starter with the push button etc. without the key) so I was starting by correcting that and I am beginning to suspect it was made as a consequence of a dodgy ignition switch.  I pulled a spare keyed switch, fabricated up a column mount, and wired it up in place of the Caterham part.  The Caterham part looks like it is a Lowe and Fletcher 2201 locking cylinder attached to a Lucas 157SA switch, the one I had on hand has some very nice threaded connector lugs instead of the male spade connectors on the Lucas part.

 

Testing to come but initial reactions are that it behaves exactly as a switch ought to, fuel pump primes, the headlights/turn signals/starter motor all run (only when the key is turned) etc.

[sltous]

Collected some data logs on a drive this morning.  All screenshots are taken at the same zoom level and show Speed, RPM, and TPP only.

 

I believe there is one that shows the TPS performing correctly and two with significant evidence of problems in the TPS.  I plan to look for corroded electrical terminals at both ends of the TPS wiring but is this an obvious 'you should replace the TPS'?

Edited March 6, 2022 by sltous

[sltous]

The TPS and the plug both appear to have OK terminals (at first look) however the TPS is missing the locking part of the plug.  I also noticed that the wiring for the TPS is not grounded/shielded which was a clear requirement when I installed a distributor-less ignition system previously.

 

Do I need (a) a new TPS with the locking bar (b) grounded wiring (c) something else?

[ashyers]

I'd watch the TPS live with the engine off and see if you can cause it to go wonky while working the throttle. If you can, I suspect it has a dead spot.

[sltous]

Thanks ashyers.  I tested the wiring by depinning the mating plug, connecting one end of my multimeter to the actual pin where it connects to the TPS and the other end of my multimeter to the actual pin where it connects to the ECU and gave the wire a good healthy wiggle.  All three wires passed with flying colors climbing from 0.2 or 0.3 ohms to a peak of 0.4 or 0.5 ohms with the same behavior from all three.  Additionally once the 36 pin main plug was connected back to the ECU the ground pin where it mates to the TPS has resistance of ~0.6 ohms to the chassis ground.  I think all of this is OK, the crimped connections have good seals and unless I add solder to the crimps I don't see any realistic way to improve this part of the wiring.

[ashyers]

Log the TPS with the engine off and ECU on. A nice gradual sweep of the throttle give you a quick indication of the condition of the guts in the TPS. You shouldn't see any spikey data in the log.

 

As far as soldering terminals, I'd strongly suggest you don't. Unless you're damn good the solder will wick into the wire, embrittle it and you'll induce a failure.

 

Andy

[sltous]

Thanks again ashyers.

 

I took a trace with the engine off over the course of approximately 1.5-2s.  Same AIM data logger / CAN bus setup as before.  Without the engine running I don't see any evidence of noise in the data. 

 

 

I also did not see any evidence of noise in the TPS data today for approximately the first 30 minutes.  According to my histograms the TPS was at >70% throttle just 0.1% of the time in the first 30 minutes (all up-hill) and at >70% for about 2% of the time (all down-hill) for the last approximately 20 minutes.  I don't see a viable explanation for clean TPS -> RPM -> Speed correlation some of the time and terrible noisy TPS -> RPM -> Speed a small portion of the time unless there is a bad electrical connection somewhere.  Happy to get alternate thoughts on where the problem may be.

 

Here I have the TPP spiking between 0->100->0->100->0 while the car is decelerating (sharply) and RPMs are dropping on the left side of the trace and on the right side I see the TPP having my expected behavior of increased TPP -> Increased RPM -> Increased Speed (I am assuming a fixed gear for the last but not for the throttle -> RPM)

[ashyers]

That last log doesn't look good. There is a possibility the TPS is dropping off the low side of the sensor pads when the throttles are pulled closed decelerating. Check the TPS setting and try your test while pushing the throttles closed and see if you get the bad signal. If not I think you're looking at either a bad TPS or wiring. Is the TPS a standard street car part?

Edited March 7, 2022 by ashyers

[sltous]

That's a very interesting thought, you may be on to something there.

 

TPS has part number PBT(GF20+GB20) which might be the same as 97vb - 9f832 - ab.

[sltous]

I have the roller barrel throttle bodies and there isn't any obvious place to push to overshoot the bottom of the TPS back around to 100% but I think it does match the behavior I'm seeing in the datalogs.  I'll think about it some more and try to do a more scientific test.

 

I took another deeper look into the datalogs, trying to determine if there was a noticeable change in road roughness around where I was having problems (nothing obvious here) and added in the Throttle Voltage layer.  It looks like the throttle voltage is not dropping to very low voltages at the same time as the TPP signal is spiking.  Do I possibly simply need to go into Easimap and adjust my throttle voltage min and max?

 

Edited March 7, 2022 by sltous

[Croc]

5 hours ago, sltous said:

It looks like the throttle voltage is not dropping to very low voltages at the same time as the TPP signal is spiking.  Do I possibly simply need to go into Easimap and adjust my throttle voltage min and max?

 

 

Interesting plots.

 

Just to clarify - this happens when engine properly warmed up?  Does it happen cold?  Does it happen with ignition on but not engine on?

 

I would not play with Easimap settings yet as (assuming I understand the history here) it worked fine on those settings up to a point in time then you experienced the issue - do I have that correct?   If so then changing Easimap would just mask the underlying source issue?