Wideband 02 heatsink...?
04-18-2011, 07:24 PM
#25
I've been experiencing this issue. There is some recent information over on Innovate's forums here (posts #38 and #40):
http://www.innovatemotorsports.com/f...t=12655&page=3
The gist:
and
As for me, I'm going to experiment with disabling overrun cutoff in the MS and see if that takes care of E8. I think that will confirm what the Innovate guys are chasing.
http://www.innovatemotorsports.com/f...t=12655&page=3
The gist:
Hello,
I have recently spent some time investigating E8 issues on several devices including LC-1s. There is a lot of confusion about what causes an E8. The truth is that lots of failures will cause an E8. The E8 is a general sensor timing error. The direct digital pulse is not responding as expected. Here are some possible causes:
1) sensor out of temperature range.
2) hardware problems inside the DD device (typically opamp failure)
3) cracked sensor (usually caused by shock cooling the sensor, thus the need for extended bungs).
Now, assuming there are no hardware problems (cases 2 and 3 above) the most likely E8 cause is the sensor out of temperature range. Our customer service has mentioned "over cooling" in the past. The reason this can happen is due to the heater algorithm that we use. It is a PID loop. The original poster of this thread has the classic example of over cooling. When he was on the throttle hard the sensor heats up. Suddenly he let off the throttle. At that point our PID loop had what’s called integral windup. It’s an error condition that takes some time for the heater to correct. So while the sensor was cooling down (via the cold air while off the throttle) the Integral had to wind down before it could start applying the heater again. Meanwhile the sensor got too cool and could no longer take DD readings. At that point it threw an E8.
Now for the good news! We have developed a new PID algorithm that will respond to Integral windup faster. So it will be able to almost immediately start applying the heater to keep the sensor within operating temperature. In addition to that scenario it will handle the scenario where the sensor can over cool even when the sensor heater is on full blast. This scenario is not as common but is possible. The new algorithm will set the LC-1 into a re-heating mode and re-try readings. This would be a temporary E8 condition while the sensor is heating up. In the past the LC-1 would throw the E8 and give up.
The fix is in the form of a new firmware. I have been testing it for months now and would like to start beta testing it in the field. If any of you would like to try the new FW please let me know. Ideally I would like someone that can reproduce the E8 problem like the original poster described. After the new FW is installed I would like to know if it fixed the problem. I would like to stress extreme caution when upgrading FW on an LC-1 as if the download fails the only way to fix it is to cut it open which destroys the rubber overmold. Test the connection on you computer via LM Programmer and Logworks before trying a FW upgrade!! This is very important.
Please PM me if interested in the FW.
Scott VanDell
Innovate Motorsports
Hardware Engineer
I have recently spent some time investigating E8 issues on several devices including LC-1s. There is a lot of confusion about what causes an E8. The truth is that lots of failures will cause an E8. The E8 is a general sensor timing error. The direct digital pulse is not responding as expected. Here are some possible causes:
1) sensor out of temperature range.
2) hardware problems inside the DD device (typically opamp failure)
3) cracked sensor (usually caused by shock cooling the sensor, thus the need for extended bungs).
Now, assuming there are no hardware problems (cases 2 and 3 above) the most likely E8 cause is the sensor out of temperature range. Our customer service has mentioned "over cooling" in the past. The reason this can happen is due to the heater algorithm that we use. It is a PID loop. The original poster of this thread has the classic example of over cooling. When he was on the throttle hard the sensor heats up. Suddenly he let off the throttle. At that point our PID loop had what’s called integral windup. It’s an error condition that takes some time for the heater to correct. So while the sensor was cooling down (via the cold air while off the throttle) the Integral had to wind down before it could start applying the heater again. Meanwhile the sensor got too cool and could no longer take DD readings. At that point it threw an E8.
Now for the good news! We have developed a new PID algorithm that will respond to Integral windup faster. So it will be able to almost immediately start applying the heater to keep the sensor within operating temperature. In addition to that scenario it will handle the scenario where the sensor can over cool even when the sensor heater is on full blast. This scenario is not as common but is possible. The new algorithm will set the LC-1 into a re-heating mode and re-try readings. This would be a temporary E8 condition while the sensor is heating up. In the past the LC-1 would throw the E8 and give up.
The fix is in the form of a new firmware. I have been testing it for months now and would like to start beta testing it in the field. If any of you would like to try the new FW please let me know. Ideally I would like someone that can reproduce the E8 problem like the original poster described. After the new FW is installed I would like to know if it fixed the problem. I would like to stress extreme caution when upgrading FW on an LC-1 as if the download fails the only way to fix it is to cut it open which destroys the rubber overmold. Test the connection on you computer via LM Programmer and Logworks before trying a FW upgrade!! This is very important.
Please PM me if interested in the FW.
Scott VanDell
Innovate Motorsports
Hardware Engineer
For everyone else, I just realized that this forum does not have a private message (PM) feature. Ooops. So if you would like to try the beta FW please email me at:
svandell@innovatemotorsports.com
__________________
Scott VanDell
Innovate Motorsports
Hardware Engineer
svandell@innovatemotorsports.com
__________________
Scott VanDell
Innovate Motorsports
Hardware Engineer
Reply
0
0
04-19-2011, 11:28 AM
#30
The 1.20 fix sounds promising. It is a two-pronged approach:
1. Make the PID loop more responsive to rapidly changing conditions (they are probably programming some logic to zero out the integral term under certain trigger events).
2. Revise the algorithm so that when E8 conditions occur, the system tries to recover for a set time period (including a zeroing of the temperature control PID loop integral term) before giving up completely. The 1.10 firmware apparently gives up the instant that E8 conditions happen.
I suspect that for the past couple of years, Innovate, like most companies, has been in "batten down the hatches" mode. They probably downsized significantly. Now that sales are picking back up, they've brought on some help and are tackling some longstanding issues. The firmware release dates on their support page tend to confirm this suspicion.
Reply
0
0
04-19-2011, 01:45 PM
#31
Supporting Vendor
Join Date: Sep 2006
Posts: 2,332
Total Cats: 67
I suspect that for the past couple of years, Innovate, like most companies, has been in "batten down the hatches" mode. They probably downsized significantly. Now that sales are picking back up, they've brought on some help and are tackling some longstanding issues. The firmware release dates on their support page tend to confirm this suspicion.
Reply
0
0
Thread
Thread Starter
Forum
Replies
Last Post
Mikel
MEGAsquirt
4
09-28-2015 04:46 PM
