Northstar rebuild: Will style (Page 56/119)
Will DEC 01, 11:44 AM

quote
Originally posted by mrgrimes:
I write software for a living and once worked on a product that used the Delphi PCM32U (LS1 ECU), as a side project I re-flashed that ECU with my own boot loader and program code completely erasing the GM software. I plan to use this ECU on this project.



I know what you mean regarding the house. I bought two this year and I haven't seen the renovated one finished yet.

This bit about the ECM is very interesting... I'm curious to hear you talk more about this.
mrgrimes DEC 01, 04:00 PM
I have seen you on the GM ECU forums before. The ECU is the 332 based derivative developed by Delphi using a Moto 68332 (before freescale) for GM and that started production in 1999 and ended its run in 2005, but you probably new that. I worked for a company that was a vendor to GM and we had an relationship with Delphi where we would buy from Delphi the ECU only flashed with seed code. Using P&E micro BDM connection we would run out of RAM and flash the boot loader or program. For me to do this today using a junk yard ECU I need to pull a pin high on the flash chip and run my PC app that downloads s-record lines to the program running out of flash which in turn sets the address on the flash to program mode and writes each s-record line. I can flash any address in the flash chip.

I have all the documents for the CPU and peripheral devices, and harness connection diagrams. Back in the day I wrote the spark control portion of the product and tested the slew rate of the coils that we were using. I found that by running the coil at > 16 volts I could get the CPU to dwell 8 coils at above 11,000 RPM. In remembering that, I decided to purchase an ECU, hack into it and start writing a basic sequential fuel and sequential spark control using map, rpm, throttle, air temp and coolant temp. I got it all working on my test bench but my electric motor would only run the crank reluctor ring that I made, up to 6000 rpm.

I can configure the ECU to run as a 4, 6, or 8 cylinder but I have to use that Northstar crank pattern.
Will DEC 02, 11:01 AM
Sounds like you're neck deep in it. I don't think I'll ever get that deep into the actual programming side of the house, but I have lots of ideas for different algorithms to try in an ECM.

I've been curious about using the product of MAP and IAT for the load axis of a spark map instead of MAP by itself. I think that would more accurately represent the likelihood of detonation and result in a better ability to control detonation.

What I'd like to try when I get back from Iraq is to use the 0411 (which I believe is the computer you're talking about) with the code from a G-Van with Vortec 350, tuned with either EFI Live of HP Tuners (probably EFI LIve to leverage some other nutty projects I want to try). I'd use the Northstar DIS brick. If that works that's probably as far as I'd go with engine management on the '99- engines.

So I understand from what you said that the '00+ reluctor wheel is the same as the LS1 reluctor wheel. Does the Northstar continue to use two crank sensors to synch faster?
The positions of the crank sensors were tweaked slightly for '00. Do you know why that was done? For example, was it to make the "reference angle" variable the same between the two engines?

If I put the '00-'04 crank into my engine, would I be able to run it with the 0411 and LS1 program (and just tweak whatever amounts to a crank reference angle)?
That would upgrade me from DIS to coil/cylinder, but I'm just curious as I'm not about to change crankshafts for that relatively small benefit.

It sounds like you're building a pretty sweet engine and you've really cherry picked the best parts of the roller cam FWD engines.

Do you have flow numbers for the '05 heads? What oil system mods were necessary?
mrgrimes DEC 02, 01:04 PM
Ideally you would want to use the air temp at the intake valves whenever the throttle is > 25%. Because at greater than 25% the intake valve is the last air valve before the cylinder and it has the most effect on air flow. As you know pressure up stream and down stream along with temperature just before any orifice determine flow across any orifice. Without air temp you would be lost on your transient control algorithm, even with feed back from an output sensor. If the throttle is < 25% than the air temp at the throttle body is preferred.

In years past I have spend many a day calibrating my speed density code in order to meet California EPA chassis dyno tests. The best calibrations I achieve were if I ran a thermocouple into the intake runner right above the valve. From there I would create/calibrate an separate air temperature map (1 dimensional table) for the permanent manifold temp sensor based on the output of the temporary thermocouple. From a software design stand point having a two separate tables one for map and one for air temp is a better software abstraction and software architecture design vs a two dimensional table with map and air temp, even if the look up is faster cpu cycles wise for a two dimensional table.

For the 05 heads I do not have flow numbers but I will try to send you a picture of the ports. The oil mods are necessary because the front cam bearing oil supply passages from the block do not match up. The head over hangs the block where the timing chain cavity and front two cylinders start. The two supply passages can be accessed easily from behind the timing chain and plumbed directly to the output of the oil pump below.
mrgrimes DEC 02, 03:04 PM
for the reluctor pickup, the ASIC chip in the ECU that front ends the signal coming from the two reluctors also uses the CAM signal to determine what firing cylinder the engine is on when it starts to crank. It only needs one sensor to figure it out because of the unique crank signal pattern combined with the CAM phase will tell the ASIC what cylinder it is with in 26deg of rotation. I think is about 3 discret pulses from the reluctor wheel which if memory serves me right are about 8.x deg apart but have varying duty cycles. So the second sensor is only for redundancy if the fist one fails or the CAM sensor fails.

The block changed quite a lot in 00 so that may be why the are situated differently then before, but I do not know on this one.
Will DEC 03, 01:19 AM
I've torn a couple of '00-'04 engines apart. I don't think the block really changed *that* much. The crank sensor bosses moved a little bit, but that's all that was obvious.

There was a mount boss of some sort added to the lower crank case just below the crank pulley... That required a change to the front cover. That boss may have been added in '98 or so, though. The windage tray and oil manifold changed a couple of times in minor ways. The big annoyance with that was that they moved the alignment dowels also, so your oil manifold has to match your lower crank case.

'00 pistons were flat tops, and '99 had valve reliefs. The '99- and '00+ used the same head gasket, so the deck changes were a product of the '05 redesign. I'll have to snag one of those engines to take a look at it.

Could use an '05 block with the '00-'04 crank and not have the deck interface issues you're having? (IOW, were the crank sensor bosses moved again in '05)? Sounds like there are some big changes at the front end of the engine for '05. Do you have pics of your '05 heads? Or of the problem areas?
Nashco DEC 04, 01:40 PM
I have nothing to add, but wanted to say that I appreciate the dialogue on the interchange and ECM info. I'm taking notes!

Bryce
mrgrimes DEC 04, 11:34 PM
Thank for the info Will. I know the casting material changed on the +00 blocks too, but I can't remember where I heard that.

As for the 05 block with the 04 crank, the back end of the block, oil manifold, and pan are very different than < 05, so much so that it would not fix in my corvette. Add that I already modified my 04 oil pan for a dry sump and 04 and 05 oil pans are different. The 05 engine was off an AWD caddy, the oil pan had a big hole in it where the drive shaft passed through.

I will attempt to take some pictures of the heads this week end.
Will DEC 05, 12:52 AM
Oh! You're running RWD heads on a FWD block... no wonder you're having difficulties. I thought you were running '05 FWD heads on an '04 FWD block.
AJxtcman DEC 07, 03:00 PM

quote
Originally posted by Will:

Sounds like you're neck deep in it. I don't think I'll ever get that deep into the actual programming side of the house, but I have lots of ideas for different algorithms to try in an ECM.

I've been curious about using the product of MAP and IAT for the load axis of a spark map instead of MAP by itself. I think that would more accurately represent the likelihood of detonation and result in a better ability to control detonation.

What I'd like to try when I get back from Iraq is to use the 0411 (which I believe is the computer you're talking about) with the code from a G-Van with Vortec 350, tuned with either EFI Live of HP Tuners (probably EFI LIve to leverage some other nutty projects I want to try). I'd use the Northstar DIS brick. If that works that's probably as far as I'd go with engine management on the '99- engines.

So I understand from what you said that the '00+ reluctor wheel is the same as the LS1 reluctor wheel. Does the Northstar continue to use two crank sensors to synch faster?
The positions of the crank sensors were tweaked slightly for '00. Do you know why that was done? For example, was it to make the "reference angle" variable the same between the two engines?

If I put the '00-'04 crank into my engine, would I be able to run it with the 0411 and LS1 program (and just tweak whatever amounts to a crank reference angle)?
That would upgrade me from DIS to coil/cylinder, but I'm just curious as I'm not about to change crankshafts for that relatively small benefit.

It sounds like you're building a pretty sweet engine and you've really cherry picked the best parts of the roller cam FWD engines.

Do you have flow numbers for the '05 heads? What oil system mods were necessary?



That is funny stuff Will