I found a 1999 Northstar Calibration Parameter Report that AJ posted in an archived post part way down the page here and of course it has the lookup charts I was looking for.
What are the units of "load" in the timing chart AJ has posted?
The timing map is labeled: Spark Advance Vs. Load Vs. RPM, Open Throt and the units are: gm/cyl RPM with the gm/cyl varying from 0.08 to 1.2 in steps of 0.04
I am used to seeing the timing map as Advance vrs Map and RPM with units for Map in KPa. varying from 20 to 100 Kpa.
Cylinder airmass (g/cyl) is a more direct measurement of the air (and fuel) in the cylinder and is a much better indicator of how much spark timing is needed for optimal burn given amount of air/fuel in the cylinder. The older use of kPa was more of an approximated load on the engine.
Cylinder airmass (g/cyl) is a more direct measurement of the air (and fuel) in the cylinder and is a much better indicator of how much spark timing is needed for optimal burn given amount of air/fuel in the cylinder. The older use of kPa was more of an approximated load on the engine.
Thanks. That makes sense. This has gotten me started.
OK - So now the task to make the 7165 $12P bin file resemble the Northstar Calibration Parameter Report that AJ posted a long while back ( see above post). I have gone through the Calibration Parameter Report maps one by one in Excel, transposing them to suit Tunerpro, proportioning them to what make sense and saving them to my $12P bin. For example, to convert from airmass (g/cyl) to MAP KPa, I called the low end of the airmass scale 20 Kpa and the high end 100 KPa and proportioned and interpolated to get even KPa values for the new map scale. Maybe not exact but close enough for a start I hope. I am about half way through and will post my results when closer to being done. I have noted that several of the 7730 $A1 bin maps are very rough when graphed in Tunerpro. My converted maps are quite smooth in comparison when graphed. Maybe this is why the engine has run poorly on the 7730 $A1. Comments are welcome. I am on the cusp of my knowledge/experience. I am making up any deficiency with motivation and hopefully not ignorance.
OK - I have copied AJ's data from the Calibration Report he posted into a variety of excel sheets, worked on them, and copied the results into the 7165 $12P bin using Tunerpro. Some comparisons between the Calibration Report data I will run in the 7165 compared to what is in my 7730 $A1 code: Main spark advance from the 7730 $A1
Main spark advance for the 7165 $12P using data from the Cal Report:
VE vrs MAP and RPM from the 7730 $A1:
VE for the 7165 using data from the Cal Report:
Timing advance vrs load and coolant temp from the 7730 $A1:
Timing advance vrs load and coolant temp for the 7165 from the Cal Report:
The ones created for the 7165 using the Calibration Report have more data points and are smoother. They are converted using excel and are intended as a starting point, especially the timing. I had to edit the XDF files to make the charts larger and transpose and interpolate the data to make new charts. Comments are welcome.....
Speaking in general terms, I was always amazed at how "ragged" some of the transitions were from one cell to the next, when looking at my own ECMs. Even factory tunes. This becomes even more apparent when looking at the graphs, as you posted. The smoother transitions would appear to make more sense.
Originally posted by Daviero: For example, to convert from airmass (g/cyl) to MAP KPa, I called the low end of the airmass scale 20 Kpa and the high end 100 KPa and proportioned and interpolated to get even KPa values for the new map scale. Maybe not exact but close enough for a start I hope.
There is no easy conversion between airmass and kPa as 100 kPa will be a different air mass depending on displacement, compression, intake/exhaust and camshaft. My LS4 pulls .72g/cyl at WOT as it delivers about 430 fwhp, so a 300 fwhp N* will be slightly less than that at WOT. So you should use .68 or something a little smaller as 100 kPa.
Wow... what a enormous difference in the shape of those plots. As you said, it's no wonder the N* wasn't running very well with the $A1 bin maps. Keep up the good work!
There is no easy conversion between airmass and kPa as 100 kPa will be a different air mass depending on displacement, compression, intake/exhaust and camshaft. My LS4 pulls .72g/cyl at WOT as it delivers about 430 fwhp, so a 300 fwhp N* will be slightly less than that at WOT. So you should use .68 or something a little smaller as 100 kPa.
Thanks for your input..... Yes I am making some assumptions that might be a bit off the mark, but are they conservative or risky? I thought that it was pretty safe. The A1 code only goes to 4000 rpm after which a VE adder is employed. The highest VE in the adder is 75% for all values of MAP, so I set the highest g/cc in the Caddi chart to 75% VE and proportioned the rest with respect to that value. If the VE in the chart is higher than it ought to be, the engine will be rich - better than too lean. I am counting in the learning of the 7165 on WBO2 to correct at least parts of the map and I can scale the rest and repeat. Here is what I dealt with:
$A1 Main VE and VE adder - the main VE chart is way smoother than it was originally after spending a lot of time on the dyno.
Main VE from Northstar calibration report (part of a post by AJ and archived)
massaged main VE combining both above charts - see the logic notes at the bottom of the chart.
I did the a similar task to adapt the main timing charts thinking that not too much advance is a safer place to start. I am not professing to know what I am doing - but am doing just what seems reasonable. So as always, comments and suggestions are welcome.
I don't think you're main and adder tables can total over 100. It will just max out and adding more fuel won't work. I don't think you are using obd1 anymore but it may be giving you false information if you are using it as a reference.
I don't think you're main and adder tables can total over 100. It will just max out and adding more fuel won't work. I don't think you are using obd1 anymore but it may be giving you false information if you are using it as a reference.
I am using OBD1. I have been running a 7730 with adapted $A1 code but was never satisfied. I am changing to a 7165 with $12P-OSE and using an Ausie NVRAM board and changing the tables to incorporate the tables AJ posted back in March 2011. I am running a 1997 Vin Y Northstar. It may seem counter intuitive to go from the 7730 to an older 7165 pcm but I decided to avoid the expense of dyno tuning an LS1 pcm. The closest dyno is now a day's drive away anyway. The 7165 idea comes from here: https://www.fiero.nl/forum/F...2/HTML/119333-4.html and is as fast as a 7730 and can learn. The Ausies seem to have this all sorted out. The $12P code does fueling all with one main VE chart, no adder chart as far I know. The adder was the old $A1 way of doing things. It was used as a reference only. Edit: The challenge is how to best adapt the OBD2 tables for the 1993 - 1998 Northstar to the OBD1 $12P-OSE code. That is what I trying to do. ------------------ Daviero - 88 N* GT
[This message has been edited by Daviero (edited 12-03-2013).]
We have all of the OBD2 Northstar's open for tuning.
This is the spark maps from a 2003 or the 04 PCM
I am going the other way AJ - not to a 2003 or 04. I am taking your tables for 1993 to 1998 and adapting to $12P-OSE. You posted these back in 2011 and after an exhaustive search I found them. Thanks for posting them by the way :-).
I am going the other way AJ - not to a 2003 or 04. I am taking your tables for 1993 to 1998 and adapting to $12P-OSE. You posted these back in 2011 and after an exhaustive search I found them. Thanks for posting them by the way :-).
Your email asked for timing tables.
OBD2 is much cheaper to tune than OBD1.
As far as a dyno tune goes you need to consider a few things/topics first. The type of dyno is very very important. If the dyno is not a eddy current or load dyno (the expensive type) it will just make WOT pulls. This will only allow tuning in WOT and it will give you HP/TQ numbers. Dynojet is a popular non load dyno
On an Eddy Current dyno you can hold the RPM/MPH and run through the MAP/Load range. This allow you to dial in the VE and MAF tables very fast, but depending on how far out it may take several hours to get it perfect. This is much easier than tuning on the street that limits you on correctly populating the cells.
I street tune every car. I have a local dyno I can rent to get the HP numbers and that is all it is good for. I did run into an issue the other day in a Corvette. I needed to dial in the PE or WOT EQ table. It was raining and I was unable to tune the car. The car spins the tire from a cruise at 45 on dry pavement. I had to keep the car 2 days extra before I was confident that the AFR was correct at WOT
Most street tunes take me 20 or so burned modification. This is why I say OBD2 cars are cheaper. The time it take to burn a chip vs using the tuning software I have that does not charge per car.
The 7165 ECM and the WB O2 sensor are all in. Checked and double checked the wiring and pinning. I really wanted to have this figured out with the minimum of help but that is not to be. I need help.
The car starts but is so very rich as to barely run and now I have a CODE 51 memcal error and cannot download to the ECM or upload from the ECM to TunerproRT5.
A few notes about what I have done to try OSE$12p on the Northstar: The OSE$12P VE table and a few other tables have been edited to reflect the stock Northstar tables. The Kinjflow injector variable is set to 0.274 L / g per sec, I think this is right? The car starts but is so very rich as to barely run and now I have a CODE 51 memcal error and cannot download to the ECM or upload from the ECM to TunerproRT5. The modified bin downloaded to the ECM successfully at first, but it won't do it now. I am using the "OSE12P XDF" file and the "OSE 12P Petrol with calc VE ADX" file.
The original memcal on the NVRAM board is the 6 cyl board that was originally in the 7165 ECM. My understanding was that only the knock filter would be used unless the ECM went to limp mode? My intent was to erase and reburn the memcal after I arrived at a decent tune.
Any comments or suggestions? I am out of ideas on where to look / what to do.
I will gladly send my Northstar OSE$12P bin to anyone who wants to see/review it and to those who want to use it if/when it is figured out.
Sounds like the nvram corrupted. Pull the jumper on the NVRAM board for the bin A/bin B selection. I have found that the new boards have issues with the bin A side corrupting but the bin B side works.
Thanks for responding Chet - I was hoping you would. I've had no joy though. I made sure all Map B variables were the same as Map A and pulled the A/B jumper. The bin downloaded from TP5 without error on the first try - my hopes were high. I started it and it idled rough but not stupid rich like before, but the data connection to Tunerpro became lost so I shut it off. It ran for maybe 10 seconds. I reestablished the data connection and found another Code 51, Memcal Error. Just for interest, I started the car again and its back to stupid rich. I cannot download the bin to the NVRAM any more either. Seems the "B" side is corrupted too? This installation has seemingly been easy for others - I don't know where my troubles are coming from. So I still need help to sort this out........
One question about the fueling also: I've read that the injector batch fire is one fire per cycle which for the Northstar would render the Kinjflow variable = 0.274 L Disp / Grams Per Sec In the TP5 Item List, "Sync PW < Min Flag" is "double fire" - if double fire, the Kinjflow variable = 0.137. Is it single or double fire per cycle?
------------------ Daviero - 88 N* GT
[This message has been edited by Daviero (edited 01-13-2014).]
Thanks for responding Chet - I was hoping you would. I've had no joy though.
One question about the fueling also: I've read that the injector batch fire is one fire per cycle which for the Northstar would render the Kinjflow variable = 0.274 L Disp / Grams Per Sec In the TP5 Item List, "Sync PW < Min Flag" is "double fire" - if double fire, the Kinjflow variable = 0.137. Is it single or double fire per cycle?
The latest batch of NVRAM modules from VL400 seem to have an issue with corrupting the ram. I have had several that have done just what you have happen. Luckily I can re-flash the ram chips. I have found that the jumper needs to be set to BIN B to work correctly. This is what I am running in my car at this time. Your ram may have corrupted both images when it went. If you want, send me your nvram and adapter/memcal and a copy of the bin you are using and I can re-flash the ram for you. I can then test it on my car to make sure it will communicate and flash through the aldl port. Granted I may not be able to run my engine on your bin, V6 vs V8 may be an issue. You still have my email address yes?
Double fire fires the injectors every crank revolution or on every fourth ignition pulse on a V8. Single fire is once for every two crank revolutions or every 8th ignition pulse. The single fire mode is mainly for use with very large injectors that are unstable firing every cycle(crank revolution) at low open times, so they fire on larger shot every other rotation allowing the injector to work correctly. This is mainly for boosted applications.
I have found the injflow calc/spreadsheet to not be of much use. I adjust the raw values instead using reasonable deduction and experience. What size injectors are you using?
That would be fabulous Chet! I will send you an email.
The injectors stock Cadillac 16.71 lbs per hour. I was using a starting Kinjflow variable = 0.274 but after seeing the double fire notation in the Tunerpro items list I am thinking that might be rich. In the end the Kinjflow number has to make the injector duty about 80% - correct?
Update: I revised the maps in the OSE$12P bin using Tunerpro-5 and the Caddi stock calibration data. The stock Caddi data has more data points in the charts so to use the extra data I changed the chart sizes where appropriate for the 7165 by editting the XDF file too. I thought the finer data set would make the maps smoother and better running. Seemed reasonable at the time. Apparently you can NOT do that. VL400 (PCM Hacking) advised me that the charts have to stay the same size, so NO changing the XDF file for chart size. Just waiting now for another flash tool now so I re-flash my NVRAM chip with my newest version of the Northstar adapted $12P bin file. My old Willem seems to have packed it in. Hopefully I can post about success soon.
Keep slugging it out Daviero... if it were easy, then everyone would be doing it!
LOL. I have to keep telling myself that! I've got my new chip burner now but been so busy with work that when I go to the shop, it seems like I am in someone else's garage not mine. Its the old paradox: you either have the time 'cause you're not working or the money 'cause you are. Sound familiar? I'll hopefully post about success soon......
Hooray! Got my Northstar running on the 7165 ECM and an NVRAM from PCM Hacking.net. Some of my initial problems were from my adaptation of the Caddi data to the 7165 bin. I also had problems with my chip burner. Non issues now. I only idled it for a short while, but it was really smooth and even. The wait for spring tuning and driving will be a long one...... It seems this is an improvement on the $A1 coded 7730 ecm.
Yes thanks. Its been a couple years of struggle to get something better. The 7730 $A1 ran but had significant deficiencies. This setup has a VE Learn for tuning with a WB O2 sensor and communication to the ECM is through the ALDL link. No more emulator and USB cable. And you can switch between 2 tunes if you want, with tables to 9600 rpm. Forget that is an older ECM, it is just a better setup for what I needed. It is only OBD1 that's no matter - I don't need proprietary software to tune either. chetw77cruiser has detailed info about the 7165 ECM here. Finally the light at the end of the tunnel is here!
I read through your posts and this looks promising. When are you going to get it out for road test? I'm very interested in this ecm swap but don't understand all the programming and tuning stuff. I can do the mechanical and electrical stuff but when it comes to the tuning there is of knowledge that needs to be learned. When you get all the kinks worked out would you be interested in doing what Ryan Hess did, documenting the wiring and providing the bin file ?
Still not completed this task. I have procrastinated doing other things to the car like a spoiler 3rd brakelight and 13" brakes. The brakes won't do much good if it does not run though so I better get that figured out soon! Can't get TunerPro 5-RT to update the 7165 ECM and NVRAM now to start to tune the thing. Computer problems have to be the most humiliating problems. Hopefully a better post soon.
Interesting reading all your post. I have been building my F40 "N" swap for over twelve years. Humbling computers, but necessary for daily drivers or performance. Installed a 1998 vin 9 motor with 95 electronics and throttle body and twin T3 turbo's. A 95 OBD1 PCM with a Howell wiring harness, ran ok for a while but when I installed 42.5 lb. injectors runs way to rich. I installed a SMT8 piggy back it is nice to be able to adjust fuel delivery but is bayonet my scope to tune. Would like o know if you got you computer sorted out, sound promising. Another issue is it has always taken off in second gear. Would like to hear back from all. Thanks
Its been a long while since I posted - sorry about that. My Northstar run fabulous on the 7165. The wide band learn function is great to get the fuel chart set and the other features of the $12P code are great. Very tunable with Tunerpro. I wish I had of started with this instead of the 7730 and $A1 code. I spent a lot of time converting the data from the stock OBDII code to usable OBD1 format though. Minor things still to perfect like seamless start/stop of ac compressor, and cold weather warmup tuning. Overall I am very happy with the 7165 running $12P code and would recommend it.
Interesting reading all your post. I have been building my F40 "N" swap for over twelve years. Humbling computers, but necessary for daily drivers or performance. Installed a 1998 vin 9 motor with 95 electronics and throttle body and twin T3 turbo's. A 95 OBD1 PCM with a Howell wiring harness, ran ok for a while but when I installed 42.5 lb. injectors runs way to rich. I installed a SMT8 piggy back it is nice to be able to adjust fuel delivery but is bayonet my scope to tune. Would like o know if you got you computer sorted out, sound promising. Another issue is it has always taken off in second gear. Would like to hear back from all. Thanks
With a manual transmission try the Oldsmobile Series 1 computer. It's called the 1998 Oldsmobile Shelby on Alldata. Contact Ryan at www.gmtuners.com His username on this forum is Darth Fiero. I'm using this computer on my car. It's basically the best available for the '99 and older Northstar cars, as it's tunable via HPTuners, which means almost any dyno shop can tune it.
With a manual transmission try the Oldsmobile Series 1 computer. It's called the 1998 Oldsmobile Shelby on Alldata. Contact Ryan at www.gmtuners.com His username on this forum is Darth Fiero. I'm using this computer on my car. It's basically the best available for the '99 and older Northstar cars, as it's tunable via HPTuners, which means almost any dyno shop can tune it.
I was going to use this myself, and contacted Ryan too. Also looked at the LS ecm. Ryan advised he might be of assistance but was not familiar with Northstars or their programs. So I decided not to go down that road. The 7165 turns out to be a great solution imho. The bin I am running is essentially a stock calibration arrived at from a calibration file of the stock caddi ecm I found. (posted by AJ and way back in the archives) The $12POSE code support on pcmhacking.net. The 7165 independent tune-ability and the learn function is great. Thanks to Chet for posting 1227165 ecu better than stock that made me think this would also work on a manual trans Northstar. Since I have gleaned so much info from the community to get this far, I will share my bin with others who may want to PM me for it.
The only thing Ryan was missing for the Northstar were the MAF tables for the throttle-mounted MAF sensor. Because it's closer to the throttle than in most LS applications, the Northstar MAF is more accurate and figures more prominently into the MAF/MAP blend than it does with LS engines. However, this is similar to the way a 3800 SC is set up, so it's not without precedent. Because the LS/Shelby computer is tunable by HP Tuners, a local tuner can dial in MAF table quickly and easily.
Its the old paradox: you either have the time 'cause you're not working or the money 'cause you are. Sound familiar?
