I can 100% confirm that after deleting the purge valve I am no longer having any issues with my link ecu. I put plenty of miles and drove it long enough to where the purge valve would normally activate. Thanks @jdniss for solving this issue, hope this helps the rest of us having this problem.

 

No problems.
You may be able to achieve the same thing by just disabling the Purge output in the LinkECU - completely removing the purge valve mechanically shouldn't be necessary.
You could also either possibly disconnect the purge solenoid, or de-pin it from the ECU connector?

Personally, disabling or removing the system entirely would be my last resort - as it could cause pressure built up, or excess fumes from the tank, if not vented correctly.
That said, assuming the AEM systems were running with no Purge integration for the last 10yrs, maybe this isn't as problematic as I theorize.

The software setting in the Link to turn the Purge function off, would look like this:


I'd suggest running this by your Tuner though - as we saw AFR was tracking a little too far off AFR Target in some of @AFK-S2K's logs, likely because the Tuner "tuned around" the (unknown at the time) Purge solenoid periodically switching on/off.


There has been some great updates over the last few weeks from @AFK-S2K and castillaricardo, relating to more advanced EVAP controls in the LinkECU, to try and remedy the issue correctly.
https://forums.linkecu.com/topic/274...issues/page/2/


Regarding ECT Coolant temperature deviations;

Again, we can thank @AFK-S2K from bringing this other LinkECU thread back to life, and back on Link's radar:
https://forums.linkecu.com/topic/203...eading-issues/

As suspected, the vehicle the LinkECU startup map was based on, likely wasn't 100% standard..




This was actually quite interesting to read, and may explain some of the differences in relation to coolant ECT readings between the factory Honda ECU and the LinkECU.




My suspicion is that Adam from Link will request resistance values again from one of you over the next few days, so that they can 100% nail this issue and complete the following:

 

I am having the exact same issue. But my circumstances are a little different, I am still on the stock ECU, with a Vortech supercharger kit, including the black box and RRFPR. I get really bad buckling at low speed/low RPM. On track lately I’ve been getting random power loss that sometimes stays for a few seconds, or sometimes stays until the car is turned off. I can’t seem to pin the issue down.

 

Finally an update to this thread.

After a couple back and forth's with my tuner, i think he finally got the evap purge figured out.

Huge shout out to Jeff Evans. They guy is incredibly patient and is willing to help. I highly recommend him.

I think the screen shots speaks for themselves. its a series of conditions and a simulated PWM.

FYI fuel tank pressure sensor is for reference only. It is not tied to any condition of how the evap works.

 

Hers some more tips i found in regards to the ECT output to the dash gauge.
First thing to know about the factory ECT gauge on the instrument panel - each "bar" has a 10-12 second delay. That means it will light up after 10-12 seconds of when its proper duty % and Hz has be reach. This was first found by Modify, and i have confirmed this by testing in data logs.

I have determined when each bar lights up by setting temps to a low range that can be reached in normal operating temps, and then let the car warm up and watch the gauge. I would blip the throttle when each bar lit up so i could go back in the log and use that as a reference point. I then went back and used those output findings to test each light.

Long story short, in order to precisely have a bar light up at a certain temperature i would need more precision in the duty % and Hz values. Most likely to the hundredths or thousandths.
Since I cant figure out how to do that in link (or if its possible), i had to set a range for each bar. This keeps the next bar from coming on prematurely. Without setting a range, i found up to 8 to 10 degrees discrepancy in the set temperature and when the output would be at the correct duty % and Hz for that particular light.
Now, with the range, my output is within 1 to 2 degrees, which i have checked again via data log. This does not factor in the 10-12 second delay in the light.
what that means, is the output will be reached within 1 to 2 degrees of your set temp. but the light will not come on for another 10 - 12 seconds, so the temp could be higher or lower. While idling on warm up, the largest delta i found in set temp and actual light on was about 5 degrees, and this was in the colder range of 80 to 174F. For anything hotter than that it was within 1 degrees F. This doesn't go to say that in an overheat situation, temps could be rising much faster.

If you want different temp settings, than just change the temp's in the top row.
Do not change the output DC or Hz

My current lights are as the following:
Bar 1 - always on, cant change this, not even by making the output 0.
Bar 2 - on at 174F (link is weird and sometimes wont let me input the exact set temp i want, like 175)
Bar 3 - 190F
Bar 4 - 210F
Bar 5 - 216F
Bar 6 - 221F
Bar 7 - 226F

I also plan on making the check engine light come on at 230F

 

that's false. I've used many types of wideband with link.

 

Agreed, I've had no issues with a analog AEM and now using the 14point7 on CAN which uses the same settings as the Link wideband.