Adamw

61 is not even connected in our ecu, that is just the constant 12V wire that would normally be used to keep the original ECU volatile memory powered up. Pin 71 & 72 are the main power supplies from the EFI relay.

I thought on the M50B25Tu the vanos was just an on/off type, not fully variable? Can you confirm. The wideband can be connected to the expansion loom (easiest option) or you can connect it to one of the unused analog inputs on the main connector. Pin 41 (AN Volt 6) is one option, this was for the MAF originally. Pin 13 (AN Volt 3) is another, this was originally for the narrowband oxy sensor.

Can you backprobe accross pins 1 & 71, these are the main 12V & ground.

I will give a simplified explanation to keep it short. Volumetric efficiency as the name suggests refers to the volume of air in the cylinder, this volume is not directly related to achievable torque or power output. The measurement that is more related to the torque output is the mass of air in the cylinder. To get from volume to mass you need to multiply by density. If we want to compare the example numbers that you gave above then to keep it simple we could assume charge temperature, fuel charge cooling etc and other things that affect density are constant, then we only need to consider boost. Assuming baro was 100Kpa, then we could say a MAP of 220Kpa gives 2.2 X the density and a MAP of 200Kpa gives 2.0X. So; 90%VE x 2.0 = 180 mass/power/torque units. 85%VE x 2.2 = 187 mass/power/torque units.

Yes this would be the best option

Yes, it does sound like it is quite possibly retarded. The base timing should be near the first thing you do on a new install or if its been apart.

You cant remove the connector off the sensor, it has a laser trimmed calibration resistor inside that is unique to the sensor. You could extend the ECU side of the wiring, or something like one of these: http://dtec.net.au/Lambda Extension Cable - LSU4.9.htm https://www.ebay.com/itm/LSU4-9-WBO2-Wideband-Sensor-Extension-Cable-5ft-fiberglass-loom-/262585943294

Yes you can just remove one tooth from our 24 tooth wheel to make it into a 24-1. I have done that before and it still provides a good waveform. The main advantage of using a missing tooth crank wheel in your situation is the position of the cam tooth is much less critical than it is when you use a non-missing tooth crank.

Yes, any aux can be assigned to "E-throttle relay". The power output from that relay then just goes to the pin on the ECU B connector labelled "+14V Aux9/10". This is done so that if there are any safety issues or loss of E-throttle control detected by the ECU it can shut off the power supply.

You are fine. For those that do have problems it is very obvious, ecu will be disabled/locked and engine wont run.

It is in 5.6.6. As per Simons post above we have pulled it off the website temporarily as we have found a problem when connecting to old firmwares. I expect it will be fixed and re-released in the next few days.

Check that you dont have Lambda 1 assigned to any of the analog inputs also.

It swaps the trigger 1 and trigger 2 signals. The dip switches change which pins the CAS trigger 1/trigger 2 signal comes into the ECU on. In "R34" position, CAS 1° signal must come in on ECU pin 41 or 51. CAS 120° signal must come in on pin 42 or 52. In "Other Models" position, CAS 1° signal must come in on ECU pin 42 or 52. CAS 120° signal must come in on pin 41 or 51. Do you have a contact for Kenji @ Link Japan? He may be able to help you in your native language if my info above does not translate well? I believe the reason for the difference is the pinout of the CAS is different:

What firmware was in your ECU prior to the change? There have been no known changes to rotary oil pump control for many releases but it is possible if you were on something older than 5.6.5 you might have got hit by the corruption thing that Simon mentions above. Try again once we release the fixed version (hopefully in a few days).

Inside the airbox or air filter, preferably with some airflow moving across it.

My suggestion for the best balance of ease of setup, with good resolution, is a missing tooth crank wheel, 36-2 or 24-1 is my personal favourite. The extra resolution that is possible in theory from say a 60-2 is barely noticeable in most situations. For a 4 cylinder, aim to have the missing teeth pass the sensor around 90deg before or after TDC. Reluctor or hall is up to you, reluctor gives slightly better accuracy and much more favourable signal to noise ratio at higher RPM but can be a little more fussy in terms of tooth design. For a reluctor sensor the teeth need to be about the same size as the pole piece inside the sensor (you can use iron filings to see its size if it doesnt poke out the end of the sensor). Also with higher tooth count wheels, the design of the "gap" becomes important - not usually important with 36-2 but if you are drawing from scratch you might as well do the proper "half depth" design in the gap like the image below. For the sync in the distributor, almost any single tooth event will work, reluctor or hall is fine. Often you can just use the stock distributor internals and grind all except one tooth off. The only real rule is you dont want the cam tooth to occur close to the gap on the crank. However if you are starting from scratch a nice option to consider (when coupled with a missing tooth crank) is what we call a "cam level" sync. This needs a hall or optical sensor to work but is where you have a "half moon" type tooth/slot. The advantage of the "cam level" sync is the engine only need to turn a maximum of 1 revolution to achieve full 720 sync. Compared with a more typical single small tooth, the engine possibly has to turn 2 full revolutions before it can sync. Another important consideration for some users on the choice between a hall or reluctor sensor is you really need an oscilloscope (and know how to work it) to confirm the polarity of the reluctor sensor is correct. Obviously this is less of a consideration with our ECU's that have a built-in triggerscope but unfortunately the Atom doesn't.

Can you start by attaching a log and a copy of your map. Lambda 2 comes in on a completely different CAN ID so for that to happen I think someone would have had to use the re-programing tool on the "find devices" tab and reconfigured the CAN Lambda. Just a 1-minute long log of an idling engine would be good. https://www.youtube.com/watch?v=_P1LRANeO4A If we cant work it out from the log and map, I might have to log in to your PC with Teamviewer to take a look.

You can use 5.6.5.3338, this will connect to all versions. Perform an firmware update to 5.6.5, then you will be able to connect 5.6.6 and update to 5.6.6 if you want.

I dont have any maps to share, if yours is all running ok then it would be best to tune what you have now. HP Academy have a reasonable webinar on setting up and tuning the G4+ knock control, you can view these with just a one month $20 membership. The knock threshold is a full 3D table, not just a single number. It will also be unique to your engine and setup so it really needs to be tuned for your engine. I believe the stock subaru knock sensor is a narrowband type and overly sensitive, we generally recommend replacing it with a common Bosch wideband one, such as 026123006. For wideband lambda the easiest option is via the expansion loom but the other option is to repurpose one of the existing analog inputs on the main header that you may no longer require. The old narrowband oxy sensor input or MAF input are common options for instance.

Oh I see I got my wires crossed now... When you said "map is running V7-8", I was assuming you were using the V7-V8 base map. Not just the MAP sensor calibration... Report back if Simon's suggestion of the dip switch change doesnt help.

For the E-throttle, in your map that was attached to the original post, you dont have an E-throttle relay assigned in aux outputs. The E-throttle system will not work without this as its a safety feature. Do you have a E-throttle relay wired as per the help diagram?

G3 manuals are in the G3 forum here: Have a look for one named G3 LEM WIring and Installation Manual.

No, Im talking about the cable connected to your dash, not the USB PC cable. The USB PC cable is perfectly fine. You need to change the wiring to the dash.

Extract the brown, grey & yellow wires out of the white plug (or snip the wires). Only green and white are used for CAN. Our you shorten the CAN DASH cable so that the RS232 wires are less than about 200mm long then it wont interfere with PC Comms. Note this problem is because you are using the CANDASH cable which is not the correct cable for this (despite its poorly chosen name it also has the RS232 wires inside). For a CAN only device connection we recommend using just the CANF plug and your own twisted pair wires for CAN H/CAN L.

Is your car the older gen 1 shape or the "newage" shape? I thought a '99 should be what we call a V5 or V6. They use a different trigger pattern to a V7 so the V7 map wont work. Can you do a triggerscope to confirm: https://1drv.ms/v/s!AiYbYlZQuRHPmieMTkwQDCXEb2LY