How long to closed loop mode?

[MR_Loyale]

Does anyone have any ballpark times it takes the EA82 to switch from open loop to closed loop? I am well aware that it varies depending on the temperature.

 

 

So for example, say it is 30 degrees farenheit. Or is the idle bump down more indicative of this? Can closed loop mode happen BEFORE idle bump down?

Edited December 29, 2016 by MR_Loyale

[skishop69]

It's actually based on engine temp for OBD1. Well, coolant temp and O2 cross counts, but ECT has priority. Depending on the software in the ECU, that can be anywhere from 100 degrees to 140 degrees so there is no set time. Since no one has cracked the older Subaru BIN files in the ECU, we don't really know where they set their temp in relation to O2 cross counts.

[DaveT]

When I was recently doing a bunch of testing, I saw that the O2 sensor signal started doing its cycling soon after the idle dropped to normal speed.

 

I could use a prom burner to read the code from an ecu memory chip. But would anyone be able to decipher it?

[MR_Loyale]

Probably the idle down is when it happens. Makes sense.

 

The reason I ask is that I have two Loyales and both of these exhibit the following behavior exactly:

 

 

 

Morning is cold. I go out start up with no warm up and drive off. I go down the hill where I live and stop at the bottom. The I turn right up a very long hill.   Shifting from first to second is no problem. I accelerate then go to shift into third and the engine lurches as if to die and I have to hold the clutch and feather the pedal to keep it alive briefly and back to second to crawl up the hill at a max of 30 mph. All the while the speedsters are on my rear honking their horn. This is repeatable and has happened regularly to me. By the top of the hill in seconds, it shifts normal and heat blows hot.

 

This lurching only happens if I do not warn up until it has idled down. If it has idled down, no problem. Having had the white loyale since new, I can say I don't recall that ever happening until recently. Could this be a relation of open loop mode and a restricted cat such that open loop cannot compensate? Just guessing here.

 

I was always told it didn't matter to warm up the engine first.

 

The cars: 93 Loyale FWD 183k miles, 92 Loyale 4WD on stick 185K miles.

Edited December 29, 2016 by MR_Loyale

[MR_Loyale]

When I was recently doing a bunch of testing, I saw that the O2 sensor signal started doing its cycling soon after the idle dropped to normal speed.

 

I could use a prom burner to read the code from an ecu memory chip. But would anyone be able to decipher it?

 

Worth a shot. Send me the BIN file. What is the processor in the ECU? Knowing this and the opcodes might give a clue,

[DaveT]

I'll dig one out and get the numbers, pull the chip tonight.

 

 

I just get in aND go. Having that big of a power drop seems unusual to me. I have 3AT s. I tend to delay up shifting a bit longer when cold. These engines are made for high rpms, not low rpm high tourque.

 

The CTS can cause weird issues. MAF, check exhaust for no restrictions. Check for air leaks into intake.

[MR_Loyale]

I'll dig one out and get the numbers, pull the chip tonight.

 

 

I just get in aND go. Having that big of a power drop seems unusual to me. I have 3AT s. I tend to delay up shifting a bit longer when cold. These engines are made for high rpms, not low rpm high tourque.

 

The CTS can cause weird issues. MAF, check exhaust for no restrictions. Check for air leaks into intake.

 

 

Is there a test for CTS? I don't think they are very accurate even when they are working properly. That is a good suggestion.

[DaveT]

There are a couple of numbers in the FSM. I have a few more recorded at home.

[DaveT]

I started a new thread for the ECU investigation.

[skishop69]

Just because the data shows it switching doesn't mean it's in CL or that it's even actually switching. Some manufacturers built in algorithms for the ECU to follow under certain circumstances, so it may show switching, but it's not. This only applies to OBD1 systems. I've done a lot of OBD1 tuning and I can tell you that OL is based on ECT first and then O2 cross counts. The smaller the engine, the lower the ECT for OL as it warms up the CAT quicker, but every manufacturer is different. Therefore, time to OL cannot be calculated precisely as it depends on manufacturer coding, engine temp and cross counts. As for reading the BIN file, you can't do it without deconstructing the OS coding so you can read the hexadecimal format the BIN is written in. Manufacturer coding is different across the board, so you can't just pull the BIN file, plug it into a tuning program and expect the program to read it. Even if you could open the file, all you'll see is the hexadecimal log which does no good without the coding to read or change it. As far as the coding goes, you'd need someone who actually knows the old Subaru coding or someone who is a coding expert to decrypt it. I'm not trying to dissuade anyone from trying as I'd love to see it hacked, but it is going to be a huge undertaking that requires expertise.

[MR_Loyale]

Just because the data shows it switching doesn't mean it's in CL or that it's even actually switching. Some manufacturers built in algorithms for the ECU to follow under certain circumstances, so it may show switching, but it's not. This only applies to OBD1 systems. I've done a lot of OBD1 tuning and I can tell you that OL is based on ECT first and then O2 cross counts. The smaller the engine, the lower the ECT for OL as it warms up the CAT quicker, but every manufacturer is different. Therefore, time to OL cannot be calculated precisely as it depends on manufacturer coding, engine temp and cross counts. As for reading the BIN file, you can't do it without deconstructing the OS coding so you can read the hexadecimal format the BIN is written in. Manufacturer coding is different across the board, so you can't just pull the BIN file, plug it into a tuning program and expect the program to read it. Even if you could open the file, all you'll see is the hexadecimal log which does no good without the coding to read or change it. As far as the coding goes, you'd need someone who actually knows the old Subaru coding or someone who is a coding expert to decrypt it. I'm not trying to dissuade anyone from trying as I'd love to see it hacked, but it is going to be a huge undertaking that requires expertise.

 

I think I have enough expertise to give it a shot with the BIN file. I could make a snarky post listing my 30+ years doing hardware and software, my BSCS with minors in mathematics and physics, USAF stint in hardware, my published articles in my field going back to 1986 or even highlight the article I wrote about hacking the palm pilot binary format in 2001 or brag about being a software engineer who did a stint at Microsoft but that should not be necessary.  No one is paying me and I am not asking permission. Hell I will probably even fail though I haven't even specified what it is I am intending other than poking around.

 

Let an old man have his fun will ya?

[DaveT]

I was hoping someone would have that sort of experience. I've written a tiny amount of assember code, some BASIC. I'm more of a hardware guy.

[DaveT]

I was hoping someone would have that sort of experience. I've written a tiny amount of assember code, some BASIC. I'm more of a hardware guy.

[skishop69]

C'mon now. I was just pointing out what we'd need . I used to have some experience in this, but that was many, many years ago and I'm way too rusty to even begin to take a crack at the code. The tuning portion  is another story. I'm glad you have the expertise necessary to try because I really would love to see it done and my limited experience is nowhere near your level.

 

"If we shadows have offended, think but this and all is mended..."

[MR_Loyale]

C'mon now. I was just pointing out what we'd need . I used to have some experience in this, but that was many, many years ago and I'm way too rusty to even begin to take a crack at the code. The tuning portion  is another story. I'm glad you have the expertise necessary to try because I really would love to see it done and my limited experience is nowhere near your level.

 

"If we shadows have offended, think but this and all is mended..."

 

 

Thats all good but I am still unclear as to what we are trying to do. So you might want to mozy on over to the  ECU thread, download the hex editor and open up the ROM file. Starting browsing for stuff that looks "tuner like".  If we have a running car and know this ecu works, we can get another UV ROM chip, alter our bin file, burn it into the ROM, change chips and see how it blows up. LOL.