Car hard to start below freezing after sitting outside all night

[bryantrinidad]

Updates on freezing temps

[Hondata]

Thank you.

What I see is that it starts with an injector pulse of 12ms whereas for all the other datalogs where it does not start the injector duration is 40ms. It seems like on the PRB datalog it started with a 18 ms pulse but then did not seem to continue to run, whereas when it did start the pulse dropped down to 5ms fairly quickly.

So the PRB/PRA makes no difference, and the open/closed loop does not either. May as well go back to your normal calibration.

For the next test can you key on, datalog and wait until the o2 sensor starts to read, then start?

Or you can just assume it is very rich and limit the 'Cranking base injector pulse' table to about 40 ms at colder temperatures (the final pulse will be adjusted for the 1000 cc injectors and will be less than 40ms).

But I would prefer to have the lambda logged for the next test.

[bryantrinidad]

I tried but I don't know if I did it right.
Here's a long log of cold start and towards the end I changed the ECT.
My only solution is still changing the ECT to start the car at below freezing temp,

[Hondata]

The o2 sensor was not working upon startup but it was not your fault. I found that the o2 heater will not operate when the ECT is below a certain temperature.

So I have added the minimum temperature for the primary and secondary o2 sensor heaters to this software release.

I suggest you change the minimum temperatures to something lower than the ambient - perhaps 0F. They are in the closed loop tab of parameters. Incidentally the default temperate is 40F but I don't think that this is the magic 40F which tells the ECU not to start.

The next time it is cold then try switching on the ignition for a minute or so and see if the o2 sensor will start to read. To me it still seems the cranking pulse is a bit high and a lambda reading at start will help diagnose this.

[bryantrinidad]

Thanks for looking further into it. Here's a datalog with ignition on for a min. I don't think it read the 02.

[Hondata]

In that datalog the battery voltage is so low the sensors are starting to glitch.

[bryantrinidad]

.

[bryantrinidad]

Above freezing cold start, sitting for 2-3 days.

[Hondata]

29F battery min 6V, sensor glitches yes

40F battery min 9V, sensor glitches no

The 40F datalog show the engine running fairly rich after start so I still think you need to remove fuel from both the cranking and the temperature correction.

[bryantrinidad]

it still says its rich even with adjusting cold cranking fuel. I dont think it reads the afr right away after it started.
It started because its above 32F. I will try again if it drops below zero. Thanks

[Hondata]

Keep reducing the cranking fuel - it looks rich. Generally I keep halving the cranking fuel until it is too lean. It needs big changes.

[bryantrinidad]

Okay. So just to make sure. Under Fuel Trim in Cranking Fuel Trim?

[bryantrinidad]

So I have a bunch of cold start here below 32F. I changed the cold cranking fuel trim lesser and lesser as I go along.
I know of someone also reported the IAT going high as well. It did on mine in one or two of the datalog while trying to start it.
Its still kind of weird that once the ECT hit 32F, it has no trouble starting. The last datalog I did was when the ECT got warm for starting the car over and over until it reached 32F, it does say 35F on the datalog because I forgot to click the record and I shut off the car once it started. I had to redo the datalog at 35F.
Wouldn't you say its kinda off that once the ECT is above 32F, its starts with no problems regarding on what is done to cold cranking, fuel trim, or tuning. I know I am not the only one with this problem.

[bryantrinidad]

Last 2 datalog, only attaches 5.

[Hondata]

All datalogs apart from '32F CC 75' show sensor glitches and a battery voltage going as low as 6V.