Proposed hardware method of fitting GM (Hitachi) 3" or 3.5" MAF to 1G DSM
06-10-2005, 01:09 AM
#1
0.0 BAR
Thread Starter
Join Date: Feb 2003
Posts: 0
Proposed hardware method of fitting GM (Hitachi) 3" or 3.5" MAF to 1G DSM
I could post this to a DSM site, but sadly I'm a filthy ------- Honduhboy and they don't like me in such places. ---- 'em. They can line up single file to munch on my ****.
Stock 1G DSM ECUs "don't read past 15 psig," for common setups, with nothing molesting the MAS (DSM jargon for MAF) signal going to the ECU. They simply run out of mappings for the higher MAS readings, and the ECU stops reacting. No more fuel is injected, no more ignition timing is pulled, nada, zip, inert. Very similar to exceeding a MAP sensor in a speed density system. Anyone who has a chipped ECU w/ fuel cut removed is in trouble when/if this happens. OH NOES!!1
There are methods of extending the stock DSM's fuel and ignition tables for higher MAS readings (THANK YOU KYLE TARRY, YOU SIR ARE NOT A DSM *** LIKE SOME!), which are all well and good, but at this point we're approaching a limit for anyone who wants to extract half of the capability of the wondrous 4G63. Most people into DSMs don't notice a problem because they ---- drivetrain components faster than they can buy go fast goodies, but the fact everything attached to a 4G63 is a complete piece of poop is subject for another discussion.
The solution is to run one of the larger Hitachi MAF meters, as employed by GM. Like the OEM MAS, they send out a digital frequency signal that is very compatible - albeit four times higher than the DSM ECU is set up to read. There has been some discussion on the dsm-ecu mailing list about doing this on the code end of things, but fudging about this code-wise is a bit more complex than it sounds. The proposed code solution is to divide the incoming GM MAF signal by 4 before the ECU deals with it in any other form. My proposed solution is to use an electronic widget to do the same thing. This allows us to refer to already published GM MAF frequency-to-airmass charts, and already published DSM MAS frequency-to-airmass charts, and so construct pretty accurate MAF tables which in turn are verified by the good old wideband.
So.
http://www.ee.mut.ac.th/datasheet/sn74ls390.pdf
The HC74LS393N is an inexpensive and robust piece of logic. Please peruse the
datasheet, you'll it's basically two chips in one. Each "half" has a /2 and a
/5 function. What we have to do is feed the GM MAF signal in one side of the
chip's /2 function, out of that side and into the other side's /2 function, and
the resulting output is fed to the ECU. Babycakes.
(1) means the "first" side of the chip, (2) means the "second."
GM MAF --> CP1(1)
Q3(1) --> CP2(1)
Q0(1) --> 50% freq cut
... so...
Q0(1) --> CP1(2)
Q3(2) --> CP2(2)
Q0(2) --> DSM ECU!!!111
Many thanks to williB of electro-tech-online for his assistance in locating/recommending the 74xx393 chip!
Stock 1G DSM ECUs "don't read past 15 psig," for common setups, with nothing molesting the MAS (DSM jargon for MAF) signal going to the ECU. They simply run out of mappings for the higher MAS readings, and the ECU stops reacting. No more fuel is injected, no more ignition timing is pulled, nada, zip, inert. Very similar to exceeding a MAP sensor in a speed density system. Anyone who has a chipped ECU w/ fuel cut removed is in trouble when/if this happens. OH NOES!!1
There are methods of extending the stock DSM's fuel and ignition tables for higher MAS readings (THANK YOU KYLE TARRY, YOU SIR ARE NOT A DSM *** LIKE SOME!), which are all well and good, but at this point we're approaching a limit for anyone who wants to extract half of the capability of the wondrous 4G63. Most people into DSMs don't notice a problem because they ---- drivetrain components faster than they can buy go fast goodies, but the fact everything attached to a 4G63 is a complete piece of poop is subject for another discussion.
The solution is to run one of the larger Hitachi MAF meters, as employed by GM. Like the OEM MAS, they send out a digital frequency signal that is very compatible - albeit four times higher than the DSM ECU is set up to read. There has been some discussion on the dsm-ecu mailing list about doing this on the code end of things, but fudging about this code-wise is a bit more complex than it sounds. The proposed code solution is to divide the incoming GM MAF signal by 4 before the ECU deals with it in any other form. My proposed solution is to use an electronic widget to do the same thing. This allows us to refer to already published GM MAF frequency-to-airmass charts, and already published DSM MAS frequency-to-airmass charts, and so construct pretty accurate MAF tables which in turn are verified by the good old wideband.
So.
http://www.ee.mut.ac.th/datasheet/sn74ls390.pdf
The HC74LS393N is an inexpensive and robust piece of logic. Please peruse the
datasheet, you'll it's basically two chips in one. Each "half" has a /2 and a
/5 function. What we have to do is feed the GM MAF signal in one side of the
chip's /2 function, out of that side and into the other side's /2 function, and
the resulting output is fed to the ECU. Babycakes.
(1) means the "first" side of the chip, (2) means the "second."
GM MAF --> CP1(1)
Q3(1) --> CP2(1)
Q0(1) --> 50% freq cut
... so...
Q0(1) --> CP1(2)
Q3(2) --> CP2(2)
Q0(2) --> DSM ECU!!!111
Many thanks to williB of electro-tech-online for his assistance in locating/recommending the 74xx393 chip!
06-12-2005, 01:20 PM
#2
1.5 BAR
Join Date: Feb 2004
Posts: 1,039
Re: Proposed hardware method of fitting GM (Hitachi) 3" or 3.5" MAF to 1G DSM
so its a one chip solution for the expensive pos GM MAFT?
thats pretty neat; i have a 2 friends with DSMs; one fwd auto one gsx awd. both are down atm; LOL go figure. My friends AWD is getting a massive turbo; fmic; huge injectors; etc etc and a megasquirt 2 standalone.
My friends fwd is getting a 16g turbo fmic and an eprom ecu. I've been looking into tuning on the eprom but there isn't much out there for the cheapskate diy guys like me.
What is a good editor for a dsm rom? besides a hex editor?
thats pretty neat; i have a 2 friends with DSMs; one fwd auto one gsx awd. both are down atm; LOL go figure. My friends AWD is getting a massive turbo; fmic; huge injectors; etc etc and a megasquirt 2 standalone.
My friends fwd is getting a 16g turbo fmic and an eprom ecu. I've been looking into tuning on the eprom but there isn't much out there for the cheapskate diy guys like me.
What is a good editor for a dsm rom? besides a hex editor?
06-14-2005, 02:01 AM
#3
0.0 BAR
Thread Starter
Join Date: Feb 2003
Posts: 0
Re: Proposed hardware method of fitting GM (Hitachi) 3" or 3.5" MAF to 1G DSM
No, not an alternative. I don't think of the MAFT as much of an option.
See, the stock mappings run out before you run out of stock MAS - fiddlyfudging the stock mappings for the Hitachi meters so you can read more power than the stock MAS is like ------- a $2 syphilized ----- because your girlfriend had to go home for the weekend for her sister's graduation. Doesn't make much sense.
The guys on the dsm-ecu list are wierd - stock MAS Hz to g/sec airflow is charted, Hitachi (eight different flavors of) MAF Hz to g/sec airflow is charted, and a little creative math + trial an error is all that is required to nail down the MAF/MAS calibration table.... yet they want to be all whiny and homo. It's less trouble (if you own a ROM emulator) to make load point adjustments in the MAF calibration table of the stock ROM this way than to have to continually poke around at the MAFT adjustment box between pulls. Seeing as how you can set your fuel maps for a flat AFR and log it easily, you can play hi/low for all given load points on the calibration table, as long as you are using injectors of a known cc/min and known deadtime. Pfft.
But, that's how we make a short story real long...
See, the stock mappings run out before you run out of stock MAS - fiddlyfudging the stock mappings for the Hitachi meters so you can read more power than the stock MAS is like ------- a $2 syphilized ----- because your girlfriend had to go home for the weekend for her sister's graduation. Doesn't make much sense.
The guys on the dsm-ecu list are wierd - stock MAS Hz to g/sec airflow is charted, Hitachi (eight different flavors of) MAF Hz to g/sec airflow is charted, and a little creative math + trial an error is all that is required to nail down the MAF/MAS calibration table.... yet they want to be all whiny and homo. It's less trouble (if you own a ROM emulator) to make load point adjustments in the MAF calibration table of the stock ROM this way than to have to continually poke around at the MAFT adjustment box between pulls. Seeing as how you can set your fuel maps for a flat AFR and log it easily, you can play hi/low for all given load points on the calibration table, as long as you are using injectors of a known cc/min and known deadtime. Pfft.
But, that's how we make a short story real long...
Thread
Thread Starter
Forum
Replies
Last Post
stuckeypcs
Forced Induction
7
01-10-2006 11:10 AM
