Megasquirt Support Forum (MSEXTRA)

I have simulated a method of a switch input to engage Limp Mode.

I don't have EGT, but have an input set up as EGT. Then when the Temp is out of range (due to me changing the voltage of the input) CEL is thrown, and Limp Mode is engaged. Thus far, I have only simulated this with an open input ADC13.

This is my actual problem: Limp Mode is indeed entered, and the "Limp mode" indicator lights in TS. The EBC is properly set to output "0", the Rev Limit operates correctly, BUT the Limp Boost Limit (kPa) is not operating (no fuel cut at 105kPa).

So, the basic issue is that the Limp Mode Boost Limit is not working. Any way to fix this?

I can take a log, if that would be helpful. Please tell me what you would like to see. MSQ is attached.

Attached is a log made when CLT out of range set limp mode.

Note that Boost Target is 102 (no EBC), at WOT, Boost goes to wastegate of 150 kPa, but the Limp mode Boost limit is set at 105kPa, No cut occurs.

Is this a known bug in the limp mode? Is there a work-around I can put into the code?

DNM

Why are you still using such an old version of firmware?

64bit wrote:Why are you still using such an old version of firmware?

Why should I believe that 1.5.1 has fixed this bug?

1.4.1 is generally a great build, and 1.5.1 does not address any features I desire.

Do you have evidence that boost protection is indeed properly activated in limp mode in 1.5.1?

Let's put it this way.

They probably aren't going to update an old firmware.

So, update. And if it's still an issue... Submit a bug report for the new firmware.


Crazy idea, huh?

Just for the fun of it, I upgraded to 1.5.1. This is not fixed in that version.

In limp mode, put a low Boost Limit, and it gets ignored. The RPM Rev Limit is properly set.

DNMakinson wrote:Just for the fun of it, I upgraded to 1.5.1. This is not fixed in that version.

In limp mode, put a low Boost Limit, and it gets ignored. The RPM Rev Limit is properly set.

Can you post the MSQ and data log on 1.5.1 too?

Here I have car in Limp Mode by virtue of EGT high. In the log, first you can see that the Lower Rev Limit is working. Next you can see that the EBC is dis-engaged (wastegate boost only); and that the Boost limit of 105kPa is being ignored.

Hmm... I think I am seeing your mistake, but I will wait for Matt to comment to avoid showing my ignorance of the boost/EGT features.

EGT stays at 975-992 degrees for that whole log (both 1.5.1 and 1.4.1), and no EGT shutdown actions are enabled.

If you want to use the sensor check to trigger limp home mode instead of the EGT failsafe, set the minimum EGT much higher - something like 800 degrees.

The issue is not triggering limp mode. Limp mode is triggered. The issue is that when in limp mode, the lower boost limit is not engaging.

I am faking the EGT to cause limp mode to be engaged.

The issue exists. It does not matter what triggers limp mode.

DNM

Actually, I did not say I was faking (faming) limp mode with a bogus EGT and limit, but that is what I did. That was so I could demonstrate the bug.

Double Post

In searching the 1.4 code for cel_overboost, I get nothing that looks like it is used. The duty limits otoh are being used for limp. If you are trying to implement a safety here, I would use the boost duty limit based on your open loop characterization until the developers can implement the cel_overboost.

Ted. Exactly. There is a professed feature that simply does not exist. The parameter is set in 1.4 and then never used. I found that same as you.

EDIT: I fear it is not a matter of when developers implement, as I cannot seem to convince them that it hasn’t been implemented already.

EDIT2: I guess I should let this rest as I’m getting a bit frustrated.

Ted75zcar wrote:In searching the 1.4 code for cel_overboost, I get nothing that looks like it is used.

Correct. It is now added for the next beta.

James

Thank you, James.

[/thread]