1987 944 Turbo --> LS
#201
Well, the 5.3 is out and the 6.0 LS2 long block is in. The MEFI ECU went to the builder to be reprogrammed for the LS2 and was provided back with the long block. Still lots to do, but nothing in terms of “building”, mostly just assembling. When I installed the oil pan I used an Improved Racing oil pickup tube brace, so that’s different. I also rerouted the oil line to the remote filter housing which shortened it considerably and took out a bend. I also picked up an LS6 intake from a friend to use instead of the LS1 intake, that will hold me over until I figure out something better (Dorman LS2? FAST?).
I’ll either sell the 5.3 as a dead unit or crack it open and rebuild and use in another project, haven’t decided yet.
.
.
.
.
.
.
.
I’ll either sell the 5.3 as a dead unit or crack it open and rebuild and use in another project, haven’t decided yet.
.
.
.
.
.
.
.
#202
TECH Resident
Glad to see this moving forward again!
I noticed what I think is the knock sensor on the side of the block with a large nut extending it further from the block behind the motor mount.
Is that right? Curious, if so, why?
I noticed what I think is the knock sensor on the side of the block with a large nut extending it further from the block behind the motor mount.
Is that right? Curious, if so, why?
#203
Update: Builder emailed back that the spacers are needed between the engine and knock sensors. Didn't explain why, I'd assume it is a MEFI thing.
Last edited by Chrenan; 02-26-2018 at 02:22 PM.
#207
LS2 in and started! Around 4 bar cold idle oil pressure (58 psi) dropping to around 3 bar (43 psi) once the oil warmed a bit. No bad noises. Still lots of little things to do, but a big step.
#208
So I’ve got a brutally hard cold start condition. Takes a ton of cranking before it catches. Once warm engine will restart fine.
I figured maybe it was the long coil extensions and long plug wires weakening spark, so I went with the stock wires and coil location, no luck. Checked fuel pressure, solid 58 at rail before and during starting and while running.
Then I pulled out the multimeter. I think my MEFI 5a ECM may be the issue. My signal voltage at the coolant temperature and intake temperature sensors were less than 1v (should be 5v). Voltage at MAP and TPS was good at 5v. Grounds at all these sensors were strong, over 12v.
I worked my way back from the sensors to the ECM, at the ECM the signal voltage being sent out to sensors were:
Pin J2-39 (cam) 5v - good
Pin J2-43 (crank) 5v - good
Pin J2-44 (TPS) 5v - good
Pin J2-62 (IAT) 0.2v - bad
Pin J3-39 (MAP) 5v - good
Pin J3-55 (coolant temp) 0.89v - bad
So something is up with signal voltage for both my intake temp sensor and my coolant temp sensor at the ECU, I was expecting 5v and got less than 1v. This is an ECM that came with the original used engine I bought which turned out to have a knock, so no surprise really the ECM may also be junk.
Quickest thing to get me up and running would be a replacement MEFI 5a ECM or possibly repair the current one. But if I have to change is it worth sticking with MEFI? I’d rather have an OEM computer or Holley HP Tuners. Those would require running a new harness, but that’s a weekend of work, not too bad at all. Any opinions?
I figured maybe it was the long coil extensions and long plug wires weakening spark, so I went with the stock wires and coil location, no luck. Checked fuel pressure, solid 58 at rail before and during starting and while running.
Then I pulled out the multimeter. I think my MEFI 5a ECM may be the issue. My signal voltage at the coolant temperature and intake temperature sensors were less than 1v (should be 5v). Voltage at MAP and TPS was good at 5v. Grounds at all these sensors were strong, over 12v.
I worked my way back from the sensors to the ECM, at the ECM the signal voltage being sent out to sensors were:
Pin J2-39 (cam) 5v - good
Pin J2-43 (crank) 5v - good
Pin J2-44 (TPS) 5v - good
Pin J2-62 (IAT) 0.2v - bad
Pin J3-39 (MAP) 5v - good
Pin J3-55 (coolant temp) 0.89v - bad
So something is up with signal voltage for both my intake temp sensor and my coolant temp sensor at the ECU, I was expecting 5v and got less than 1v. This is an ECM that came with the original used engine I bought which turned out to have a knock, so no surprise really the ECM may also be junk.
Quickest thing to get me up and running would be a replacement MEFI 5a ECM or possibly repair the current one. But if I have to change is it worth sticking with MEFI? I’d rather have an OEM computer or Holley HP Tuners. Those would require running a new harness, but that’s a weekend of work, not too bad at all. Any opinions?
#209
9 Second Club
Not sure your mode of testing or where you're getting info from...but the ecu does not send out 5v to sensors as such.
3 wire sensors will typically have a 5v regulated supply, a clean signal ground and then a return voltage to the ecu which will be variable 0-5v.
2 wire temperature sensors would typically have one side to the clean signal ground and the other will be a signal return to the ecu. The sensor is basically a variable resistor.
Now that signal return will often have an internal pullup to 5v, but it is not technically a 5v supply but it may measure close to 5v if you place a digital voltmeter across that and ground when the sensor is NOT plugged in.
If the sensor is plugged in, then you would expect to see a voltage between 0-5v, depending on the actual temperature.
So what you're seeing above sounds quite reasonable, depending on the specifics of the wires you are testing.
but really, it would be much easier to plug in a diagnostic scanner and view live data to see if any sensors are unhappy or odd and check for any fault codes indicating the ecu is not happy.
3 wire sensors will typically have a 5v regulated supply, a clean signal ground and then a return voltage to the ecu which will be variable 0-5v.
2 wire temperature sensors would typically have one side to the clean signal ground and the other will be a signal return to the ecu. The sensor is basically a variable resistor.
Now that signal return will often have an internal pullup to 5v, but it is not technically a 5v supply but it may measure close to 5v if you place a digital voltmeter across that and ground when the sensor is NOT plugged in.
If the sensor is plugged in, then you would expect to see a voltage between 0-5v, depending on the actual temperature.
So what you're seeing above sounds quite reasonable, depending on the specifics of the wires you are testing.
but really, it would be much easier to plug in a diagnostic scanner and view live data to see if any sensors are unhappy or odd and check for any fault codes indicating the ecu is not happy.
#210
That’s good info, thanks for that! Makes sense then where I’m seeing 5v, since those are multi wire sensors, whereas as the two temperature ones are 2 wire. So if that’s not the issue I’ll have to look at something else. But I’m not sure what at this point, the basic stuff, fuel/air/spark I’ve checked. Once started everything seems normal. It’s just cold starts. That is a problem with MEFI 5, there’s no easy scanner that I know of, MEFI 4 there was.
#212
9 Second Club
A quick google of MEFI5 ? I've never heard of it before.
http://mefi5.com/
Do you have a wiring diagram of your setup or whaetever ecu you have ?
Has it ever been tuned correctly for your motor etc ?
http://mefi5.com/
Do you have a wiring diagram of your setup or whaetever ecu you have ?
Has it ever been tuned correctly for your motor etc ?
#214
Yes I have a wiring diagram for my harness and manual for MEFI 5 ECU that shows all the pins. MEFI was reprogrammed for LS2 by the company that built the long block.
That $350 scanner option would be good I might pick that up. The other option was only for MEFI 4.
That $350 scanner option would be good I might pick that up. The other option was only for MEFI 4.
#215
9 Second Club
Is this software for that setup ?
http://www.mefipro.com/index-m9.html
$350 is a huge amount for a code reader or scanner for a single application.
#217
9 Second Club
You have to pay for the software as well as the ecu ? It looks like some sort of hacked OEM ecu then ?
How much is their tuning software ? I don't see any prices on their site.
If it was based around an OEM ecu, does it give any standard OBD info via the DL16 connector on the harness ? Or is that not even present ?
ie, so you could just use any generic OBD code and data scanner ?
Anything that can give you data like that rather than trying to mess about reading voltages etc with a voltmeter, would be useful.
How much is their tuning software ? I don't see any prices on their site.
If it was based around an OEM ecu, does it give any standard OBD info via the DL16 connector on the harness ? Or is that not even present ?
ie, so you could just use any generic OBD code and data scanner ?
Anything that can give you data like that rather than trying to mess about reading voltages etc with a voltmeter, would be useful.
#219
TECH Junkie
iTrader: (1)
Seems like, in the long run, you'd be better off switching over to something that is more user friendly.....mine's just a street car. But I intentionally went with a bone stock crate engine and stock OEM ECU. And I've got bone stock drivability, emissions and fuel mileage. Just what I wanted.
#220
I’m leaning that way. Not sure what the “best” would be for me, OEM or something like HP Tuners. Megasquirt is a little too do it yourself for me, and MOTEC is too expensive.