Osiris Reflash After a Disconnected Battery
11-09-2010, 08:10 PM
#1
Osiris Reflash After a Disconnected Battery
This issue has been raised before, but my experience conflicts with the conventional wisdom. Most of the posts I've read here say that a disconnected battery will not affect the reflash, but that has not been my experience.
I recently performed some work on my car that required disconnecting my battery for about a day. For reference, I have a 2005 Infiniti G35 coupe with a vortech supercharger installed. Mods include the 3.12 pulley, 600cc injectors, walbro 255lph fuel pump, cosworth manifold, and berk hfc.
After I reconnected the battery, I noticed that my car runs leaner and that the high speed fuel cut reacts differently. Specifically, cruising AFR drifted to the high 14s / low 15s, and, even during relatively slow speed coasting, AFR will quickly spike higher (this used to only happen at highway speeds). The maps in my Osiris tune are still there, as was verified by switching to my anti-theft map. However, the mix does not richen up as it should when engine load and RPM increase. Engine code P1283 (running lean) has popped up a few times as well. I have not observed any vacuum leaks.
Strangely, I've noticed that as I drive the car more, the car seems to be learning its way back to a richer fuel mix. If I hold the revs and engine load constant (as displayed on my gauges), I can see the AFR adjust within a few seconds. I am not comfortable getting into boost, however, since I cannot be sure if the mix will richen quickly.
I raised this issue with my tuner about 2 weeks ago, but he has had trouble getting a response from Uprev (I guess they were away at SEMA). Perhaps someone on here can chime in. Has anyone else experienced this?
I recently performed some work on my car that required disconnecting my battery for about a day. For reference, I have a 2005 Infiniti G35 coupe with a vortech supercharger installed. Mods include the 3.12 pulley, 600cc injectors, walbro 255lph fuel pump, cosworth manifold, and berk hfc.
After I reconnected the battery, I noticed that my car runs leaner and that the high speed fuel cut reacts differently. Specifically, cruising AFR drifted to the high 14s / low 15s, and, even during relatively slow speed coasting, AFR will quickly spike higher (this used to only happen at highway speeds). The maps in my Osiris tune are still there, as was verified by switching to my anti-theft map. However, the mix does not richen up as it should when engine load and RPM increase. Engine code P1283 (running lean) has popped up a few times as well. I have not observed any vacuum leaks.
Strangely, I've noticed that as I drive the car more, the car seems to be learning its way back to a richer fuel mix. If I hold the revs and engine load constant (as displayed on my gauges), I can see the AFR adjust within a few seconds. I am not comfortable getting into boost, however, since I cannot be sure if the mix will richen quickly.
I raised this issue with my tuner about 2 weeks ago, but he has had trouble getting a response from Uprev (I guess they were away at SEMA). Perhaps someone on here can chime in. Has anyone else experienced this?
Last edited by - bigc -; 11-09-2010 at 08:12 PM.
11-10-2010, 01:58 AM
#3
New Member
Join Date: Sep 2009
Location: Italy
Posts: 193
Likes: 0
Received 0 Likes
on
0 Posts
If you disconnect battery there is a reset of learning values of fuel a/f.
May be you need some times to go back to a "normal" situation...
Usually there are no so big "fluctuations" in values but may be...
May be you need some times to go back to a "normal" situation...
Usually there are no so big "fluctuations" in values but may be...
11-10-2010, 01:26 PM
#4
What work did you do on the car? Anything that could have affected the fuel pressure?
And as ITOzann stated, the ecu "learns" over time and retains those values. Disconnecting the battery for a long time probably cleared the learned fuel trims. I wonder if it is possible that the tuner did not clear the learned fuel trims before tuning your car, and now that they have been cleared your tune is off? You would have to ask Uprev if that is even possible though.
And as ITOzann stated, the ecu "learns" over time and retains those values. Disconnecting the battery for a long time probably cleared the learned fuel trims. I wonder if it is possible that the tuner did not clear the learned fuel trims before tuning your car, and now that they have been cleared your tune is off? You would have to ask Uprev if that is even possible though.
11-10-2010, 02:10 PM
#5
^ i had to run a power wire to my trunk for air suspension. the battery was disconnected while performing the electrical work.
yeah, i am still pushing to get some information from uprev, but, as of yesterday, my tuner still has not heard back from them.
yeah, i am still pushing to get some information from uprev, but, as of yesterday, my tuner still has not heard back from them.
11-10-2010, 07:00 PM
#6
Disconnect the battery and you end up back on map 1 and also have to relearn your fuel trims (50-100mi of driving). I don't know what map you normally drive on (there are 5) but that is something to keep in mind. I have my valet map in slot 1 for this reason -- fuel cut at 3krpm and I know to switch to the correct map 
There is nothing else that disconnecting the battery is going to do to an uprev tune.
There is nothing else that disconnecting the battery is going to do to an uprev tune.
Last edited by djamps; 11-10-2010 at 07:01 PM.
11-10-2010, 09:43 PM
#7
map 1 is my regular driving map. i'm in map 1, as verified by the flashes using the cruise control button. i drove 450 miles from LA to SF after the battery was disconnected, and have probably put 100 more miles on the car since i've been back in norcal. there has been an effect on my uprev tune.
Trending Topics
11-11-2010, 08:48 AM
#10
My guess to the O/P...something else is going wrong if the tuner properly flashed his tune onto your ECU. Software doesn't just change on it's own.
11-12-2010, 05:31 AM
#13
Registered User
11-12-2010, 08:01 AM
#14
Registered User
Join Date: Aug 2010
Location: So Cal
Posts: 34
Likes: 0
Received 0 Likes
on
0 Posts
Chris, I've disconnected battery for extended periods of time (several days) due to an issue with taillight mod... as far as I can tell my Osiris tune has not be affected. However, I cant remember if I have monitored a/f after disconnect (I have data logged) several times, just cant remember if it was before of after the taillight issue.
I was just at Church's yesterday and told Shawn I was pinging occasionally and we both thought it might be the crappy CA fuel (blends change seasonally) - but maybe in fact I have the same issue? It's going to be hard to tell right now because I was reflashed again (pulled timing a little and some other things), but when I have some time I'll battery disconnect and monitor to see if anything changes.
I was just at Church's yesterday and told Shawn I was pinging occasionally and we both thought it might be the crappy CA fuel (blends change seasonally) - but maybe in fact I have the same issue? It's going to be hard to tell right now because I was reflashed again (pulled timing a little and some other things), but when I have some time I'll battery disconnect and monitor to see if anything changes.
Last edited by CustomGsedan; 11-12-2010 at 08:03 AM.
11-12-2010, 08:28 AM
#15
Thanks for posting up Mike. Oddly enough, I mentioned this to Randy a few weeks ago, and he noted that he had seen the same thing with a car he was working on. Ran fine before the battery disconnect, but was unstable afterwards (he did audio work). He took the car over to Church and said Shawn observed the same thing...
11-12-2010, 08:56 AM
#16
Let's think about this for a second -
As I understand it flashing an ECU completely overwrites the tables that were flashed into it's memory previously. So if you flash your ECU with a custom Uprev tune you are effective overwriting the values in all of the tables that the ECU references to determine fuel delivery, timing, etc. So how on earth could you "lose your tune" if it is flashed into memory? There would be nothing else for the computer to revert back to. If the ecu truly lost it's memory, wouldn't the car just not run at all?
The only way you could "lose your tune" is if the tuner used Real Time Tuning and then forgot to flash the tune permanently into the ecu's memory. Here are a few blurbs from the Osiris user's manual that help explain what I am talking about:
Understanding the ECU architecture and how UpRev modified ROMs work.
Before we get into how to tune using Osiris you need to have a basic understanding of the ECU architecture.
At the core of the ECU is a processor that uses 3 types of memory:
-
Flash-ROM (internal to the processor). This is Read Only Memory, where the ECU stores its operating software as well as the tables that are used to calibrate/tune the ECU’s operation. This area can only be written to via an ECU reflash (Osiris). The flash memory has a limited number of rewrites that can be preformed before the memory starts to degrade. Once the flash memory has begun to degrade the ECU will have to be replaced. A minimum of 200 flashes can be preformed according to the CPU data sheet, but twice that is likely possible (we flashed the ECU on our development 350Z more than 400 times without any memory degradation).
-
RAM (internal to the processor). This is Random Access Memory (read and write) that the ECU software uses to store variables that change in real time such as sensor readings, environmental conditions and changes over time, as well as calculated values that are used for the software control models that manage engine operation.
-
Non-volatile EEPROM (external from the processor). This non volatile memory is used to store things like idle adjustments, ignition adjustments, learned idle air volume, VIN, and other static values that vary from vehicle to vehicle. You should not have to deal with the EEPROM, but it’s good to know it’s there.
The most important feature that UpRev adds to a ROM when it is patched is taking the critical tables that are used for tuning (ie fuel, ignition, throttle, rev/speed limits, etc…) and modifying the ECU software so that it loads these tables into RAM and looks the tables up out of RAM as opposed to the flash memory. This relocation of the critical tables allows for real time tuning (RTT) without the need to reflash the ECU for every adjustment. It also makes map switching possible since the values can be changed on the fly.
Due to limited RAM space on the ECU, there are certain data values and tables that have to remain static and cannot be placed in RAM for RTT, such as target idle rpm and cam advance. These values can only be changed via modifying the ROM and doing a reflash. These static tables/values can’t be changed via map switching either.
Another UpRev ROM feature is the addition of map switching and storage space for alternate maps. Map switching allows for multiple maps to be stored in the flash memory so that they can be loaded into RAM at the user’s request (via cruise control switches).
It is important to note that the only the tables that are loaded into RAM can be changed via map switching. That means only the real time tunable tables are updated when a map is switched. The static tables/values, such as target idle rpm and cam advance, can not be changed via map switching.
From the tuning section:
...Continue to make passes and adjust the RTT parameters as needed until you are happy with the results. Once you have a finished tune you will need to save the RTT file from Cipher so that it can be permanently flashed onto the ECU (the values that are changed in real time with Cipher are only stored in RAM and the tune will be lost once the ECU looses battery power).
As I understand it flashing an ECU completely overwrites the tables that were flashed into it's memory previously. So if you flash your ECU with a custom Uprev tune you are effective overwriting the values in all of the tables that the ECU references to determine fuel delivery, timing, etc. So how on earth could you "lose your tune" if it is flashed into memory? There would be nothing else for the computer to revert back to. If the ecu truly lost it's memory, wouldn't the car just not run at all?
The only way you could "lose your tune" is if the tuner used Real Time Tuning and then forgot to flash the tune permanently into the ecu's memory. Here are a few blurbs from the Osiris user's manual that help explain what I am talking about:
Understanding the ECU architecture and how UpRev modified ROMs work.
Before we get into how to tune using Osiris you need to have a basic understanding of the ECU architecture.
At the core of the ECU is a processor that uses 3 types of memory:
-
Flash-ROM (internal to the processor). This is Read Only Memory, where the ECU stores its operating software as well as the tables that are used to calibrate/tune the ECU’s operation. This area can only be written to via an ECU reflash (Osiris). The flash memory has a limited number of rewrites that can be preformed before the memory starts to degrade. Once the flash memory has begun to degrade the ECU will have to be replaced. A minimum of 200 flashes can be preformed according to the CPU data sheet, but twice that is likely possible (we flashed the ECU on our development 350Z more than 400 times without any memory degradation).
-
RAM (internal to the processor). This is Random Access Memory (read and write) that the ECU software uses to store variables that change in real time such as sensor readings, environmental conditions and changes over time, as well as calculated values that are used for the software control models that manage engine operation.
-
Non-volatile EEPROM (external from the processor). This non volatile memory is used to store things like idle adjustments, ignition adjustments, learned idle air volume, VIN, and other static values that vary from vehicle to vehicle. You should not have to deal with the EEPROM, but it’s good to know it’s there.
The most important feature that UpRev adds to a ROM when it is patched is taking the critical tables that are used for tuning (ie fuel, ignition, throttle, rev/speed limits, etc…) and modifying the ECU software so that it loads these tables into RAM and looks the tables up out of RAM as opposed to the flash memory. This relocation of the critical tables allows for real time tuning (RTT) without the need to reflash the ECU for every adjustment. It also makes map switching possible since the values can be changed on the fly.
Due to limited RAM space on the ECU, there are certain data values and tables that have to remain static and cannot be placed in RAM for RTT, such as target idle rpm and cam advance. These values can only be changed via modifying the ROM and doing a reflash. These static tables/values can’t be changed via map switching either.
Another UpRev ROM feature is the addition of map switching and storage space for alternate maps. Map switching allows for multiple maps to be stored in the flash memory so that they can be loaded into RAM at the user’s request (via cruise control switches).
It is important to note that the only the tables that are loaded into RAM can be changed via map switching. That means only the real time tunable tables are updated when a map is switched. The static tables/values, such as target idle rpm and cam advance, can not be changed via map switching.
From the tuning section:
...Continue to make passes and adjust the RTT parameters as needed until you are happy with the results. Once you have a finished tune you will need to save the RTT file from Cipher so that it can be permanently flashed onto the ECU (the values that are changed in real time with Cipher are only stored in RAM and the tune will be lost once the ECU looses battery power).
Last edited by mx594; 11-12-2010 at 09:03 AM.
11-12-2010, 08:57 AM
#17
It sounds to me like your tuner forgot to flash the tune into the ROM memory and you lost it when you disconnected your battery, or something else changed. Check your fuel pressure. It could also just be a bad O2 sensor. Are you using a third party wideband or the stock wideband?
Last edited by mx594; 11-12-2010 at 09:04 AM.
11-12-2010, 09:09 AM
#18
^ thanks man, i really appreciate that post! i'm still on the stock o2 and haven't thrown a code for a failing o2 sensor. that could be an issue though. the aftermarket wideband i have on my a-pillar uses a bosch sensor and is only about 6 months old.
11-17-2010, 10:18 PM
#19
Update: Chalk this one up as 95% me being paranoid and 5% fuel trims slightly off. I took the car for a spin this morning while Rob from ZCG (my tuner) rode shotgun and logged data. With the exception of a few minor tweaks, he said the car was still performing in line with the original tune 
I guess after seeing my fuel going all over the map during the first few hours following the battery disconnect I couldn't get it out of my head that something was off.
I guess after seeing my fuel going all over the map during the first few hours following the battery disconnect I couldn't get it out of my head that something was off.
Thread
Thread Starter
Forum
Replies
Last Post
