koracing

Very weird looking. Can you attach the log file as well as the tune file so someone trying to help may be able to check how your logging is set up?

Well just did a remote session with customer using the new USB direct beta firmware. It worked better for sure - maybe a little better than 50/50 it would stay on and connected when starting the motor. A very nice improvement for sure, but not perfect yet either.

It might help if you attach a datalog from the software from when key is off, then while cranking, as well as the file loaded into the ecu currently.

You could certainly set the activation control to a virtual aux and then use whatever logic you wanted for activation parameters beyond the basic Launch settings.

It should be doable with the Link. It looks like the setup information is in the help menu for the MoTeC at least. So given that, it seems like it should doable to get the CAN dual wideband from Haltech to work as well.

Yes I would definitely verify your I/O. You can change the values in the file you have or start with the ST205X base file. There's a few differences from what I can see when comparing.

Setting semantics aside for a moment: I don't believe this user ever said that he had a ST185 ecu. He said he used the ST185 version base map which theoretically should be closer to what his engine is than the ST205 base map. He did, however, say he has a 1993 USDM Gen2 3SGTE. 1993 would be the first year of the V2 aka late Gen2 in the MR2. I believe his ecu must be a V2-3 due to the fact that his car would not even try to start if he had the wrong ecu for his pinout. If we had the part number of his stock ecu that would clear it up definitively. The bottom line though: If it was able to run fine on an OEM ecu without a bunch of modifications to the harness, it would have to fall in line with one of the 3 versions stated above, and it would be extremely unlikely that it was some odd or unique pinout. As far as what you've found in the past, I can only speculate. One theory would be a mismatch of engine and body harnesses. The engine harness has 2 of the ECU connectors and the body harness has the 3rd. So it is feasibly possible to get some odd mismatch of the right connector from a V1 with a left 2 connectors of V2-3 or vice versa. This would power up fine with any ecu, and if the ecu matched the engine side harness it would see proper trigger signals, but it would not start up on an OEM ecu. In this scenario there are other harness issues that would arise (chief among them being a lack of any fuel pump turning on due to no output on the FC pin).

Definitely strange. I have to believe there is a software or driver issue on that laptop. I just tried plugging in a USB to serial adapter I have into my laptop and it immediately showed up in my device manager as COM6 under Ports (COM & LPT) as well as USB to serial converter under the USB controllers section. I don't have a Link ecu with me at the moment, but I'm certain it would do the same thing on this laptop anyway.

It is pretty well known on the gen2 3sgte there are two pinouts. Early gen2 and late gen2. The late gen2 is about 2 wires different than the gen3 (st205). The early is 14 wires different. At least it is pretty well known in the 3SGTE community. If your ecu is the later one it is possible the base map's default map sensor pin location is incorrect as that does chang e between the gen2 and gen3. Notice the pin locations for the PIM (Pressure: Intake Manifold) aka MAP sensor. "VS" aka Vane Sensor is the Airflow meter signal. So just changing the Analog 4 to the MAP sensor should solve the MAP issue. I should have recalled this previously. Yellow boxed pins are the trigger signals. Blue is the MAP signal. Green is the AFM aka MAF. Grey shows the other primary differences on the early gen 2 versus the later pinouts. From the Link G4x Help menu pinout for the ST205X: I'm attaching a revised version of your file once again with this change made.

Do you have a boost gauge? Can you tell me what it shows when key on engine off, versus when it is idling at 500rpm? Rechecking your datalog the map input voltage is 3.92v with key on and engine off. That would make a 0-5v range (assuming it is linear) of 0kap to 125kpa at 5v. I'm certain your map sensor is not reporting a correct voltage to the ecu. You should also double check the primary ground wires on the bottom of the intake manifold as well as the main ground from the transmission to chassis. By check I mean remove, clean all contact points with scotch-brite or other abrasive to have clean bare metal contact and reattach.

Your map sensor is showing 169kpa when the engine is off - this is very wrong -- approximately the equivalent of about 10psi of positive manifold pressure. It should be near to identical to the BAP reading of 98.5. It does drop when you start the motor so it's *moving* with engine vacuum meaning it's working and sending out some type of signal - but the calibration is obviously wrong. The early and late gen2 did use two different map sensors, however. You may want to see if the calibration for the ST205 map sensor is closer to 98.5 when key on engine off - this could make a huge difference in how the car runs at the moment. Here's a modified version of your tune with idle timing control turned on, and a custom guesstimated MAP calibration based on your logs. Also with closed loop fuel trimming turned on via the oem narrow band oxygen sensor. If you update your ecu firmware, you can also turn on long term trimming to help with that roughing in of the fuel table. tuning185 v3.pclx

I've never had an issue connecting with both sofwares open at the same time... Both set to autosearch...

Typically on these engines if the distributor is perfectly centered in the slot, 95% of the time the trigger setting default will be within 1-2 deg, but it should always be checked and corrected with a timing light. The RPM is very low which is probably why the MAP reading is hovering around 75kpa, but I would check that your MAP reading is showing a similar value to the BAP when the key is on and engine is off - and if not performing a MAP calibration. If you try to give it any throttle does the car actually rev or does it die? Try taking a log with starting the log before the engine starts, and then showing the cranking and startup as well.

Personally I don't like to run the supra fuel pumps in the MR2 because as they easily overheat and I've had many instances in the past where after about 20-30min of back to back dyno runs we start losing fuel pressure. Changing the pump to a walbro 255 and no more issues. So while they work, for a while, I certainly wouldn't recommend running one in any type of endurance type of use. I think it largely stems from how Toyota provided variable voltage or power to the pump in the MK4 supra, and the fact that we are tyically just providing straight battery power when on as typically installed in the MR2.

G4+ Plugin as well as G4x - removed the control board from the adapter board on an MR2/ST185 unit - so just using the control board on a bench. I inserted header strip into the socket on the bottom of the board to bench test and attach leads to. I was using test leads/patch cord to connect to the control board pin positions. Full functionality on the first 6 injector outputs as well as all other outputs. The test leads to the header strip look like the following: Power supplied by variable voltage bench top power supply (set to approximately 13.5V DC). Ground to ground plane pins opposite the injector output (just after injector ouptut pin 1) pins (which all show zero resistance to one another).

I am well aware of the fact that the plug in ecus only have hardware to drive injectors on the first six injectors - but in doing bench testing it appears there is no logic at all on 7 and 8? I tried connecting a relay trigger to these output pins, and then turning them off and on as well as assigning a pwm output to make the relay click at low frequency. I also tried a low wattage 12v light bulb. No signs of life. I was told there is at least logic level (i.e. I read this as less than 0.5A) switching ability on these outputs. Can someone shed some light on these outputs? I tried asking through my local Link (North America) contact and have never received a definitive response other than "it should be able to switch a relay". What I would really like to know is if it is feasibly possible with the addition of some hardware mounted separately external to the ecu to provide the hardware required to run injectors off of these two outputs making the plug-in fully sequential 8cyl capable?

Not sure if you're still chasing this issue, but the advice here is mostly sounding pretty good. Additional info: the wires that come from the main-efi relay into the trunk to actually power +B and +B1 (black with yellow stripe wires usually on the bottom right corner of the right most ecu connector pins 71/72) have a splice where they go from 1 to two wires. I have seen this develop a drop and corrosion inside the harness. Replacing these two wires all the way from the bottom of the rear fuse box to the ecu pins solved it on a couple cars. The main power wire that provides the switched 12v power to the rear fuse box comes from a single wire in the left side kick panel that uses a large spade connection - I would look at that connection as well as test voltage at the kick panel as well as at the rear efi fuse to see if your drop is in the run of the body harness from there to the rear fuse box. This main power wire also has a splice just before the fuse box where the power splits to several locations in the engine bay which I disremember right now. Alternatively, while less than ideal, you could add a relay in the trunk with voltage supplied by the alternator positive (add a fuse and source wire from the large + terminals in the rear fuse box), and triggered by the original power wires to the ecu to provide power to the ecu. As someone else stated, however, it is best if the ecu is seeing the same voltage as the injectors, ignition, etc. to ensure you're running on the proper part of a response curve and ignition dwell for a given voltage, as well as other corrections that are dictated by the ecu voltage. What voltage do you see on the injector pins and coil positive with key on/engine off? Same as ECU or not?

What exact model is your ecu and how is it adapted to the car if it is not a plug-in?

If your oem narrow band is still installed it can be used to at least get a ballpark setting for a base to move a car around by watching when it changes from 1V (rich) to 0v (lean) - but like @adamw says - attach your base map and a pc log of startup (Hit F8 to start logging, stop logging, and the logging menu at the top to save the log file). With the Narrow band o2 it is theoretically possible that you can enable closed loop stoich trimming as well as long term learning based on the narrow band which may make it run well enough to move the car.

Why are you getting 820cc injectors and only planning to run 1 bar of boost on an upgraded stock turbo? Are you considering upgrading to a larger turbo later on? I've seen some signal issues with your chosen cam and crank sensors - so I would make sure it's all working well with the OEM sensors first before attempting to install and then trouble shooting issues with these pieces, or give yourself a reversible path to do so -- maybe make the cam and crank sensor signals plug into a sub-harness that can be changed depending on the trigger types and connection points.

If I could add a softkey toggle button to the screen or assign a hot key to this it would certainly be a little more elegant than what I'm doing now, but as it is, it doesn't sound much simpler than what I'm already doing. Perhaps the ability to assign a Fn key or software toggle button to inputs/outputs may be a useful item for other purposes as well?

I often seem to (when tuning startup and idle) have a need to kill the motor without killing the ignition power. Hydra EMS had a stop engine button that was very useful. So far the trick I've used in link is to highlight the the 9-12 cells right around where the engine is idling and type 0 enter, followed by a bunch of Ctrl+Zs to restore those values. It would be nice if there was a button to do this in the software with a dedicated button. In the Hydra the way it worked is it would only be active below a certain RPM, and not if the wheel speed was above zero, but i'm sure there are several ways this could be handled. What I'm doing works ok for now, but this *is* a wish list.

It should be able to start the engine if you input the proper calibration for the OEM map sensor - if memory serves me there is no table specified for the MAF signal in the base map (and who would want to use that anyway?). The OEM gen2 map sensor changed a little from early to late versions, but a little googling and you can find the voltage versus map signal in some old MR2OC threads. That being said: what are you thinking you will need to recalibrate and start up the motor for? I mean without installing an aftermarket larger range map sensor (3 or 4 bar) I'm not sure what you are hoping to achieve by starting it up on all the stock stuff. A wideband is pretty much required to make any type of fueling changes which a base map will very likely need, and should be installed at a minimum. I wouldn't recommend starting the motor until you at least have a wideband and map sensor installed and calibrated within the link. It would be possible glean the necessary information to get a reasonable idle if the wideband is installed and the readings visible via a gauge, but not wired into the link as well, but, again, why would someone do that?

Perhaps you need to "run as administrator"? i.e. right click on the Software and then click "run as administrator" from the options. I have had some weird issues with connecting before when remote accessing customer laptops and if I remember correctly this helped.

Here is the knock settings I used on one E36X equipped car that may give you a reasonable starting point: Use at your own risk of course.