Megasquirt vs CHem

[ottawa rogue]

I'm going to need a new cal this winter as i'm planning a T2 conversion on my '86 2.2.
would i be better off using CHem to burn a new cal, or would something like Megasquirt be a better way to go for tuning flexibility?

I'm not going for all out performance as this is my DD, mainly looking to run 20lbs or less boost.

[Clay]

then go ChEM. If your looking for all out performance and tunability, then a stand alone would be the way to go.

Personally, I see the ChEM as THE way to go on a car like what your talking. If nothing else, its a stepping stone. Get the car running right with a good cal, then if you want more down the road you can always upgrade.

I run a blueberry code on my GLHT with super results. Its fairly straight forward, and has a lot of tuning capability in it.

[ottawa rogue]

I'm waiting for my account to be activated on the CHem website, sounds like i need to do some poking around over there

[ottawa rogue]

how "user friendly" is chem to use?
the thing with Megasquirt, is that it looks an awful lot like some of the tools i use now at work in terms of the interface etc.
is CHem a setup where you burn a chip and stick it in to see what happens, or is it more of a tune as you go/on the fly setup?

[MiniMopar]

It's awkward and buggy, but it odes certain things well. D-Cal is also awkward and a bit buggy. The two together get you pretty far, but it's not plug-n-play. Blueberry60 is tuned for Geoff's Shelby Z and has a lot of weird stuff that I had to undo to fix driveability issues. Also, not all the tables are fully understood. BB is much better than it was, but if you have been following along the past few years, you'll know what I mean.

I'm not saying you shouldn't do it. I'm just saying that it is probably not as easy to tune as one of these systems that are designed to do exactly that.

[ottawa rogue]

cool, that's what i was looking for. i may check into having someone burn e a cal once i get things going, but it's looking like megasquirt may be the way to go for what i want.
once i hear from the chem moderator, i should know.

i guess the big thing for me is the megatune interface, it's alot like some of the heavy truck software i use at work. Hell, if i can diamondlogic to run relaibly, this shouldn't be too bad:D

thanks for the info guys

[Clay]

It's awkward and buggy, but it odes certain things well. D-Cal is also awkward and a bit buggy. The two together get you pretty far, but it's not plug-n-play. Blueberry60 is tuned for Geoff's Shelby Z and has a lot of weird stuff that I had to undo to fix driveability issues. Also, not all the tables are fully understood. BB is much better than it was, but if you have been following along the past few years, you'll know what I mean.

Hmmmm I never had an issue with the blueberry60. but Im running basically the exact same setup he is. Ran low 8s in the 1/8th mile w/ slicks "out of the box", more or less. Had to bump up the launch RPM, but that was it.

[MiniMopar]

So you are running +20s and a 3 bar, then?

[Clay]

yup, +20s, 3 bar, same head, same intake, same turbo, same displacement, etc.

[SpoolinGLH]

I get discourage to about the Ladybug code. I am good with computers,wiring,cars in general. But I DO NOT know where to start when looking at moparchem.. its very frustrating, thats why alot of people arent on the forum. I would love to burn my own cal but looking at it is mind boggling

[Aries_Turbo]

I wrote up a quick tuning guide for calibration work back in one of simons threads in here. id search for turbovanman as the author in this section i think. I'll look for it later. it makes it easier. most of the tables you can ignore.

Id go the Chem/D-cal route.... Biggest reason? the factory knock detection electronics. they work. MS just doesnt have quite as good of a setup as far as ive seen.

Brian

[MiniMopar]

yup, +20s, 3 bar, same head, same intake, same turbo, same displacement, etc.

That would be why. I ran that setup on my Daytona for a bit. It barely has enough fuel to start when it is warm out. Scale that cal for +40s and your car won't ever start the first time in the summer. He also messed with the advance tables in a way that was causing audible pinging under medium boost/high RPM in some cases...particularly when it's still on the cold advance tables. If you are using a grainger and have cast pistons, be very careful. He also renamed most of the tables and changed a lot of the scales. That means you can't compare his cal to any stock cal without hand-editing a TBL file for the old cal. There are lots of errors in the assembly with respect to TBL and CALX generation. Several of the tables are completely wrong while others are missing from one or the other. Also Chem2 only supports CALX files while D-Cal only supports TBL files. This closes the door for Chem2 for a lot of things outside of Ladybug.

I'm not trying to scare you or to bash Geoff. He did a lot of really tedious work and Blueberry has come a long way. My point is just that you can't expect to be able to just sit down and make a cal. It just doesn't work that way. I've spent months doing a lot of trial and error to fix these little issues. WOT fuel and spark is the easy part. That is basically what the Super 60 cal is. If it's a race car, then you are fine. If you need to drive it to work every day, it needs a lot of tuning.

Ladybug seems to have a lot of issues. Stick with Rob's cals. They pretty much "just work" and are documented better.

[ShelGame]

... Stick with Rob's cals. They pretty much "just work" and are documented better.

That's because my cals aren't "tuned" at all. They're all stock cals, simply scaled for X injectors and MAP sensor. The factory put in thousands of hours to get it to where it is. I don't see any reason to deviate from it straight away. Obviously, major changes to the engine can require major changes to the cal. I left it up to the end user to tune for his engine combo. Of course, I can do some custom tuning as well.

I'm just finishing up a new relocatable source for the '89T1 that will have 100% of the tables/constants ID'd and documented in D-Cal (well, I probably won't fully document the purge control, EMR, or EGR routines); ~70% for CHeM2. Most of the constants and a large number of the tables never need to be touched. The new source has all of the previous code "mods" now built-in. That is, they're not afterthoughts tacked onto the end of the binary and then hacked in - they're now assembled into the main-line code. Also, unlike LB and BB, the code mods can be enabled by a new config constant. So, you won't have to re-assemble the code to turn on the CE Flash, 2-step limiter, or hi-speed logger; you can turn them on in the binary. Much like the factory does with A/C and EGR. So, you won't lose your cal changes just to turn on a new feature. I have plans for some other new features as well.

I'm just starting to build the relo codebase for the 90/91 T1/VNT.

I'm not sure that I'll release the sources; but I will release new cals built on them. I need to finish testing the '89 T1 first, then they will be out. The '90/91 stuff a little later. After the '90/91 T1, I'll start on the 91 T3.

Eventually, I'd like to get to a point where I have a common codebase for both ECU's (well, the HEP-based ECU's anyway) so that the table usage is the same betwen them. As it is now, there are new tables and more complexity in each generation. I'm not sure it's really feasable (there's a lot of code that's hardware-specific).

It's really too bad we don't (yet) have a single binary editor that does everything we need to do with a cal...

[MiniMopar]

That's because my cals aren't "tuned" at all. They're all stock cals, simply scaled for X injectors and MAP sensor. The factory put in thousands of hours to get it to where it is. I don't see any reason to deviate from it straight away. Obviously, major changes to the engine can require major changes to the cal. I left it up to the end user to tune for his engine combo. Of course, I can do some custom tuning as well.

Yes, and we appreciate that. :) Much of what I have done has been to roll-back Geoff's changes back to CSX Stage II specs. Every time I do the cal gets better.


I'm just finishing up a new relocatable source for the '89T1 that will have 100% of the tables/constants ID'd and documented in D-Cal (well, I probably won't fully document the purge control, EMR, or EGR routines); ~70% for CHeM2. Most of the constants and a large number of the tables never need to be touched. The new source has all of the previous code "mods" now built-in. That is, they're not afterthoughts tacked onto the end of the binary and then hacked in - they're now assembled into the main-line code. Also, unlike LB and BB, the code mods can be enabled by a new config constant. So, you won't have to re-assemble the code to turn on the CE Flash, 2-step limiter, or hi-speed logger; you can turn them on in the binary. Much like the factory does with A/C and EGR. So, you won't lose your cal changes just to turn on a new feature. I have plans for some other new features as well.

That sounds great, Rob. You are definitely the king of the SMEC cals. You took the best of the 3 and concentrated on that. I have made some of Geoff's feature switchable by some TBL hackery (modifying branch instructions), but they are not as nice are using bit fields since I have to re-hack them for each new BB version.


Eventually, I'd like to get to a point where I have a common codebase for both ECU's (well, the HEP-based ECU's anyway) so that the table usage is the same betwen them. As it is now, there are new tables and more complexity in each generation. I'm not sure it's really feasable (there's a lot of code that's hardware-specific).

Yeah, my gut instinct is that it is probably not worth the effort. The '88 SMEC cal is close to the '87, but why would one use the '88 when the '89 is so much nicer? Beyond '89, I don't really have much to say. I've never owned a SBEC car. :)


It's really too bad we don't (yet) have a single binary editor that does everything we need to do with a cal...

Boy, no kidding. So many times I wished I had the source for D-Cal to fix some really annoying bugs and missing features in it. I started working on a portable (Windows/Linux) cal editor, but I just don't have that kind of spare time anymore. Intended to release it under GPL to finally end the lock-in. I'm not saying that I don't appreciate Derek's work. I just wish he'd allow others to enhance it and return them to the community.

I did get pretty far with a BIN->ASM script. It was just a Perl script that would using the TBL and LST file to push changes to the binary tables back into the source. That way you could tweak the code and keep your tuning. I've been using a binary diff tool, but it's a bit tedious.

Anyway, I felt the need to post to this thread to warn about BB. Someone could cause serious damage if they weren't careful with it.

[ottawa rogue]

That's because my cals aren't "tuned" at all. They're all stock cals, simply scaled for X injectors and MAP sensor. The factory put in thousands of hours to get it to where it is. I don't see any reason to deviate from it straight away. Obviously, major changes to the engine can require major changes to the cal. I left it up to the end user to tune for his engine combo. Of course, I can do some custom tuning as well.

I'm just finishing up a new relocatable source for the '89T1 that will have 100% of the tables/constants ID'd and documented in D-Cal (well, I probably won't fully document the purge control, EMR, or EGR routines); ~70% for CHeM2. Most of the constants and a large number of the tables never need to be touched. The new source has all of the previous code "mods" now built-in. That is, they're not afterthoughts tacked onto the end of the binary and then hacked in - they're now assembled into the main-line code. Also, unlike LB and BB, the code mods can be enabled by a new config constant. So, you won't have to re-assemble the code to turn on the CE Flash, 2-step limiter, or hi-speed logger; you can turn them on in the binary. Much like the factory does with A/C and EGR. So, you won't lose your cal changes just to turn on a new feature. I have plans for some other new features as well.

I'm just starting to build the relo codebase for the 90/91 T1/VNT.

I'm not sure that I'll release the sources; but I will release new cals built on them. I need to finish testing the '89 T1 first, then they will be out. The '90/91 stuff a little later. After the '90/91 T1, I'll start on the 91 T3.

Eventually, I'd like to get to a point where I have a common codebase for both ECU's (well, the HEP-based ECU's anyway) so that the table usage is the same betwen them. As it is now, there are new tables and more complexity in each generation. I'm not sure it's really feasable (there's a lot of code that's hardware-specific).

It's really too bad we don't (yet) have a single binary editor that does everything we need to do with a cal...



but does the same stuff apply to a LM cal? i'm not looking to swap over to a SMEC.

i guess what i'm wanting is a good "daily driver" cal that might gain a little economy in part throttle cruise on the highway (i drive 70 niles every day to and from work) and yet be able to compensate for the changes i'm wanting to make.
after rereading my first post, i guess i was kinda vague on my current setup....sorry guys.
current setup is as follows
bone stock '86 t1 log with no mods whatsoever aside from a '93 A-523 replacing the auto.

what i'd like to do
convert to T2 status with a late model t1 intake, a good sized intercooler and maybe a log header with a 3" exhaust all the way back. and put a bigger fuel pump in it.
i know i said 20 psi at the limit, but it'll probably end up around 14-16 psi.

does this sound like something a "tweaked" calibration coud handle, or am i gonna have to go all out with a heavily modified cal for this?

[MiniMopar]

If you go with a 3-bar MAP sensor and +20% injectors, you can use the Blueberry60 cal with some care. It is based on the '87 Turbo II cal and has been modified by the author for his Shelby Z. At present, it is the only show in town for open 87 calibrations.

[ottawa rogue]

+20s from where i'm at now, or after the retrofit? my current injectors are the 27pph units

and just how reliable would the bb60 be? i've read a few things besides your issues here and over at BM and the chem forum that suggests that it's still a little "buggy". i don't want to have to wrench on it all the time just to keep it running.(well besides the usual stuff:D )

[MiniMopar]

+20s as in 42pph.

The best thing to do is back-out his changes from stock from all the advance, cold start, and various other random tables. Then you should at least have a good starting point.

[ottawa rogue]

or maybe just start with the stock '87CSX cal and go from there?
from what it sounds like rob is saying, he just takes a stock cal and scales the fuel tables up to meet the new requirements, in a nutshell

[ottawa rogue]

well, if i can ever get activated over at chem, i'll be able to download the manual and see how bad it really is.
personally, i think he oughta make it so you could download the manual and look through it first before you joined just to see if you even wanted to mess with it.
it would save everyone some time

[MiniMopar]

The manual isn't all that helpful for what the real issues are. Like I said, calibrating WOT performance is one thing. Making it run nice all year around is trickier. There is some good, general info in there though.

The trouble with the stock cals is that the TBL files are very old and unmaintained. So by doing that, you are going back in time about 2 or 3 years. I suppose I could try to release a clean BB cal with stock curves in it. I'm still finding things, though.

[Aries_Turbo]

yeah he did change alot of stuff he didnt need to. when I made 2.5L changes to it I directly copied a decent amount of 89 T1 tables into it, replacing alot of the stuff he goofed with. i never get audible pinging. I only get a cel blink when the car is really heat soaked and I hit 18-20psi but the light goes out pretty quick.

part of the problem is that he is taking alot of the 89 T2 SMEC stuff and adding it in there and I dont think all of it is good... like the knock table is less sensitve...

Brian

PS, can anyone take the stock 87 T2 calibration and add the CEL blink and launch limiter and thats it?

[Aries_Turbo]

im still confused at who actually has the part and full throttle timing tables correct. more than half the cals i look at seem to have them backwards... IE more timing in the table labled full throttle in the higher boost regions.

Brian

[sdac guy]

well, if i can ever get activated over at chem,

Even if you don't ever get the confirmation email, go back periodically and try your login there. For a lot of folks, the logins work and they have never received the email confirmation.

Barry

[MiniMopar]

PS, can anyone take the stock 87 T2 calibration and add the CEL blink and launch limiter and thats it?

It should be possible. The trick is to take the old 87 disassembly and locate everything that is so clearly marked in BB60. I find that BB42 is a good stepping stone, before the tables got really whacked. Then copy the corresponding tables from the stock disassembly into BB60.

If you don't want to mess with the assembly, take the old 87 T2 TBL file, hand-rename the tables and scales to match BB60 (or vice-versa). Then you can go through each and line-up BB60 to stock T2 in D-Cal. This is what I have been doing, but I've been using the CSX Stage II cal instead of the stock T2.

[MiniMopar]

im still confused at who actually has the part and full throttle timing tables correct. more than half the cals i look at seem to have them backwards... IE more timing in the table labled full throttle in the higher boost regions.

LOL. I know what you mean. The original TBL file for the 87 T2 is wrong and it proliferated to all the others. I looked at the BB60 assembly and it first grabs the full throttle table result, then checks the WOT flag. If set, it goes on. If not, it does a bunch of stuff involving part throttle. So as far as I can tell, BB60 is correct.

[Aries_Turbo]

yeah ive loaded the stock T2 as a comparison and with two d-cal windows open, ive renamed the majority of the tables so I can edit things graphically with the same scale. i have an old copy of 42 i think... if not, can you email it to me? what was the difference between 42, 43 and 60? is it mostly just table changes? constant changes? I might have to do all my tinkering all over again and post some cleaner cals. is 42 pretty much stock, scaled and relocatable?

Brian

[Aries_Turbo]

now do you know if the 88-9X cals follow this same convention... ie the full throttle higher boost timing being lower than the part throttle? I would assume that they do.

Brian

[MiniMopar]

yeah ive loaded the stock T2 as a comparison and with two d-cal windows open, ive renamed the majority of the tables so I can edit things graphically with the same scale. i have an old copy of 42 i think... if not, can you email it to me? what was the difference between 42, 43 and 60? is it mostly just table changes? constant changes? I might have to do all my tinkering all over again and post some cleaner cals. is 42 pretty much stock, scaled and relocatable?

IIRC 42 was the first "Blueberry" named cal from him, but not the first relocatable one. I'm not sure what the significance of the name is, but I'm sure it was some undisclosed milestone of his. BB came out after a long period of silence from the ChEM side (during which D-Cal picked up a lot of steam). Anyway, BB42 has some table names changed and also has table massagings. If you want it, PM me your email addy.

[MiniMopar]

now do you know if the 88-9X cals follow this same convention... ie the full throttle higher boost timing being lower than the part throttle? I would assume that they do.

The 88 code is very similar to the 87. As I understand it, little changed in 89 T2 code either. However, the 89 T1 code is very different and uses a 256kb ROM instead of a 128kb. Anyway, the answer is yes. The 89 SMEC cals have two warm advance vs MAP tables: part and full. There are other advance tables that the older cals don't have, though.

[Aries_Turbo]

yeah ive taken quite a long and hard look at the 89 stuff when making a 2.5L version of blueberry. i know about the timing differences. it didnt really affect me copying the 89 T1 timing curves into the 87 code. works just fine.

Brian

[amcpacer]

I love blueberry. I would recommend it to you first instead of megasquirt or standalone. If blueberry does not work well you will still have a Mopar Logic Module that you could try other cals with or even modify a stock cal. I spent a lot of time changing values in a modified version of blueberry from Aries_Turbo and have found it to offer excellent drivability. The only ongoing problem I experience is warm startup. It may be something with my car though. Perhaps a weak charge temp sensor or something.

Also it is way cheaper to use Chem.

[ottawa rogue]

well, it looks like there's a pretty good support system if nothing else.
i've got a spare LM laying around so i'll give chem a shot.
does it matter if it's a tbi LM?

[MiniMopar]

The only ongoing problem I experience is warm startup. It may be something with my car though. Perhaps a weak charge temp sensor or something.

That is a common problem. There are a bunch of things that are contributing to this.

[Aries_Turbo]

must be a turbo LM and PM.

whats your guess on warm start issues Russ? I would say fuel. Mine is pretty good on warm start. it start pretty quick overall. I did goof with the A,B,C tables and copied in the 89 T1 start fuel from engine temp table. that seemed to really help cause i have a 2.5L

Brian

[MiniMopar]

whats your guess on warm start issues Russ? I would say fuel. Mine is pretty good on warm start. it start pretty quick overall. I did goof with the A,B,C tables and copied in the 89 T1 start fuel from engine temp table. that seemed to really help cause i have a 2.5L

Yeah, it's not getting enough fuel during a "cold" start when it is above 70^ or so outside. Hot starts are fine. I made a lot of progress by undoing a lot of Geoff's changes to starting fuel related stuff. I have a detailed revision history of all I have done. It now starts most of the time, but on occasion I have to try a 2nd time. I am down to two tables and it gets a little better each time. For short pulse widths, I don't think injectors scale linearly. So scaling everything for a bigger injector is just step 1. Then you have to tweak all the edge cases where there is no feedback.

Pasting-in the 2.5L fueling stuff probably fixed a lot of the non-stock type issues. I am now concentrating on curve A in that temperature region (point 4) and StartPrimeFuelShot. I'm trying not to mess with StartingFuelFromEngineTemp too much.

[Aries_Turbo]

gotcha. I kinda run a hybrid 2.5L T1/2.2L A,B,C curve cause it didnt like the 2.5L one plopped in there directly. yeah I bumped up one of the curves in the ~70degF region to provide better starts and to keep it from stumbling there as it warmed up.

Brian

[ottawa rogue]

i finally got activated yesterday, and downloaded chem, the manual and a copy of blueberry.
this doesn't look like it's going to be too bad at all, the interface and programming menus remind me of some of the engine software i use at work:thumb:

once i get through this move, i'll have to hit up the junkyards for a '87 LM
and try it out for real:D
WHoops, got busted....i gotta go pack some more:(