How can I force the PCM to stay in closed loop mode all the time ?
03-28-2005, 04:20 AM
#1
Staging Lane
Thread Starter
How can I force the PCM to stay in closed loop mode all the time ? Specificly, for the 6.0L truck motor.
What is it that causes the ECU to switch from closed loop to open loop mode ? TPS sensor ? a throttle-switch? MAP output?
I want to force the ECU to stay in close loop all the time and simulate the narrowband o2 sensor using a wideband system. That way, I can force whatever AFR I want at any RPM/TPS/MAP condition.
If I need to simulate the TPS, MAP ignal, etc. that's not a problem.
Thanks.
What is it that causes the ECU to switch from closed loop to open loop mode ? TPS sensor ? a throttle-switch? MAP output?
I want to force the ECU to stay in close loop all the time and simulate the narrowband o2 sensor using a wideband system. That way, I can force whatever AFR I want at any RPM/TPS/MAP condition.
If I need to simulate the TPS, MAP ignal, etc. that's not a problem.
Thanks.
03-28-2005, 08:28 AM
#2
Moderator
iTrader: (11)
Join Date: Mar 2002
Location: East Central Florida
Posts: 12,605
Likes: 0
Received 6 Likes
on
6 Posts
I think you may have a lot of work to do there.
The PCM wants a switching O2 waveform and
tries to dither the mixture to get cross-counts
of a certain density and controls based on the
duty cycle high/low.
You could lower the closed loop enable temp
and suppress PE by putting the enable MAP &
TPS very high.
But it seems to me that, if you're going to be
trying to control the mixture by the back door,
jacking around the WBO2 voltage as feedback,
you're going to run into other problems. Like,
the closed loop feedback time is well longer than
you want, for fast events (like rolling on power
enrichment in a tenth of a second, by lowering
the O2 feedback, it's take several seconds for
the STFTs to bring you up from 14.7:1 to 12.8:1
by their slewing). In the meantime you're likely
pinging like mad. Bypassing the "real" fuel calcs
that a commanded enrichment would use and
making the trims do the work, I don't expect will
be satisfactory in the end.
The PCM wants a switching O2 waveform and
tries to dither the mixture to get cross-counts
of a certain density and controls based on the
duty cycle high/low.
You could lower the closed loop enable temp
and suppress PE by putting the enable MAP &
TPS very high.
But it seems to me that, if you're going to be
trying to control the mixture by the back door,
jacking around the WBO2 voltage as feedback,
you're going to run into other problems. Like,
the closed loop feedback time is well longer than
you want, for fast events (like rolling on power
enrichment in a tenth of a second, by lowering
the O2 feedback, it's take several seconds for
the STFTs to bring you up from 14.7:1 to 12.8:1
by their slewing). In the meantime you're likely
pinging like mad. Bypassing the "real" fuel calcs
that a commanded enrichment would use and
making the trims do the work, I don't expect will
be satisfactory in the end.
03-28-2005, 08:55 PM
#3
Staging Lane
Thread Starter
Thanks for the info.
Actually, simulating the switching signal of the narrowband o2 sensor it trivial
But that's the easy part.
What I was hoping was that the PCM would re-learn its fuel tables based on the new o2 data. So that it would not rely directly on the o2 sensor loop for mixture control. Just for re-learning the tables. Sounds like that may not be directly possible.
In which case, I can just make my own PCM and take over the injectors. All I have to do is snoop the TACH, MAP, TPS, MAF signals + a wideband. I would probably still have to simulate the narrowband to prevent the "real" PCM from throwing an error code. Possibly add load resistors across the old PCM's injector connectors.... Hell, I'll use them for triggers
These are hi-impedance injectors, right ?
A bit more work than I was hoping for. But such is the price for fully predictable closed-loop tunability, I guess.
Actually, simulating the switching signal of the narrowband o2 sensor it trivial
But that's the easy part.What I was hoping was that the PCM would re-learn its fuel tables based on the new o2 data. So that it would not rely directly on the o2 sensor loop for mixture control. Just for re-learning the tables. Sounds like that may not be directly possible.
In which case, I can just make my own PCM and take over the injectors. All I have to do is snoop the TACH, MAP, TPS, MAF signals + a wideband. I would probably still have to simulate the narrowband to prevent the "real" PCM from throwing an error code. Possibly add load resistors across the old PCM's injector connectors.... Hell, I'll use them for triggers
These are hi-impedance injectors, right ?
A bit more work than I was hoping for. But such is the price for fully predictable closed-loop tunability, I guess.
03-28-2005, 11:49 PM
#5
Staging Lane
Thread Starter
Well yes and no... the stock PCM will still try to go for stoich. But it won't be connected to the injectors. My PCM will lie to it and tell it everything is ok. Then my PCM will run the injectors at whatever AFR I want. Basically, I am getting in between the PCM and the injectors, like a piggyback controler. But, in addition, I will fake out the stock PCM to believe everything is normal. That way, it won't pull back the timing, etc. Really, all I have to do is simulate the stock 02 sensor signal.
I THINK there are two o2 signals ( pre and post cat ) that I need to do. No big deal.
If I pull this off, the stock PCM will run happily along THINKING it is controlling the injectors, but actually just doing everything else it is supposed to do.
Since my closed-loop is wideband based. I can target any AFR I want.
I THINK there are two o2 signals ( pre and post cat ) that I need to do. No big deal.
If I pull this off, the stock PCM will run happily along THINKING it is controlling the injectors, but actually just doing everything else it is supposed to do.
Since my closed-loop is wideband based. I can target any AFR I want.
03-29-2005, 12:46 AM
#6
FormerVendor
Join Date: Jan 2003
Location: Melbourne, Australia
Posts: 525
Likes: 0
Received 0 Likes
on
0 Posts
Sounds like one hell of a project! Good Luck!!
Writing your own code for the stock PCM to do what you want is not impossible as long as a wideband interface with electical characteristics the PCM can handle is available. I imagine it would take up less space than the narrowband O2 code, but i am only guessing.
Again, good luck!
Chris...
Writing your own code for the stock PCM to do what you want is not impossible as long as a wideband interface with electical characteristics the PCM can handle is available. I imagine it would take up less space than the narrowband O2 code, but i am only guessing.
Again, good luck!
Chris...
03-29-2005, 02:45 AM
#7
Staging Lane
Thread Starter
Writing your own code for the stock PCM to do what you want is not impossible as long as a wideband interface with electical characteristics the PCM can handle is available. I imagine it would take up less space than the narrowband O2 code, but i am only guessing.
Actually, the wideband part is trivial for me. I work for Innovate ( LM-1 guys ). I might use an LC-1, or I may just build the wideband controller into the PCM and avoid analog signals alltogether. I can probably run the digital wideband and PCM code on the same processor
Maybe an Atmel Mega128We'll see what develops
