barge

Thank you for the clarification. The plan is to get all this installed and tested this Saturday. I'll follow-up here with results just in case someone else is trying to do the same.

That was my plan so I'll proceed with that route. If i have issues I can always change it. The benefit of doing everything yourself.

If that tach output wasn't required for the DCCD I'd relocate everything to 1-4 but since it'll allow me to set then I'll give 7/8 a shot. I will check resistance. I can't imagine they are much. It's the usdm 04 Subaru STI PnP in a 2004 Forester XT. I've remapped and moved a lot of IO around.

Bringing this back from the dead since I actually have a quad cam AVCS jdm motor in the car now and would like to get the exhaust AVCS connected and have a couple questions/clarifications. The software will let me set Aux Output 7&8 as VVT solenoid. I'm assuming if it will let me make this selection then it will work OK on those outputs? The power for the intake avcs solenoid is connected at the ECU. RH and LH each have their own +12v connection. Any comment on if i'll have an issue using this pin to supply power to the RH and LH exhaust solenoids? The exhaust cam sensors are a 2-wire VR type. is there a preferred sensor ground source? Will i have an issue if i bring both sensor, sensor ground connections, back to where the crank sensor connects? If that's not a good idea any recommendation on pin selection?

Thanks @Adamw that's the kind of information I'm looking for. I'm now understanding what you're saying about shifting during warmup. Not do to oiling or anything but due to the actual expansion of the block. I may be able to do some maths and get a good approximation of I could heat a block up and see how far the deck height moved. @Brad Burnett No it's not a "crazy" build... 400hp, 9k rpm ej22t with D25 heads. Stock cams to start since my initial power levels aren't crazy. However since it's my first real engine build and I have the winter I'm trying to make it "perfect". I'll be balancing the rods, getting the Pistons all matched, doing all the little stuff that people don't normally do. Having my cams where they should be is big. @Davidv Good idea on the MAF. I actually have a slot style MAF that I was going to use blowthrough before I went Link. Could be a useful tuning tool if nothing else if I slapped it in.... Or added some easily removable piece.

I guess that could make sense why people have measured cold motors stock and found significant error. As a DIY engine builder for this project I don't suppose you could point me in the right direction as to this fudge factor. I was planning on just ensuring the oil control gear/control inside the cam gear was pushed all the way to one edge and degreeing it from there. Basically it looks like you're answering my question with... degreeing the cams would be beneficial to ensure the cams are in the correct position. If nothing else ensuring both sides are matched while cold seems like it could be beneficial.

I guess my main question on the intake cams is how does "0" relate to the cam angle test value. If the cams are not both "0" at the same offsets (I know they won't be the same value but the same offset to TDC) does 0 put those cams at that offset value essentially or does it sync them to the motor at a "real" zero. Maybe the cam angle test is already taking care of this and I just am not understanding it. I've read that even factory cams have been measured 2-5 degrees off spec.

I've read through the manual and I don't see this laid out but maybe i'm missing it. We have cam angle offsets to define the actual position of the cam trigger. This tells us where the cam is but what if this is incorrect. If the cam is shifted a couple degrees due to poor quality control and we set the cam offset to this position couldn't that perpetuate the error across the control scheme? The tables say that 0 in the cam tables is the fully relaxed position. What if I want one came to be -2 at "Zero" and the other to be +2? Can this be factored into the cam offset if you degree the cams manually? I'm currently in process of doing a built subaru motor and plan on taking a lot of time with eccentric idlers and adjustable exhaust cam gears to degree the cam. However with stock ecus people say degreeing the intake cam is pointless because the ECU takes care of it. I'm not sure how true this really is... but i'm wondering if there is a set way to basically use the avcs to control individual bank cam timing. If so it may make more sense to go to a quad cam avcs head.

I had this same issue and pretty much asked the same question. The issue turned about to be my power wire routing which was set up just like you did. I wired from the trunk battery, to the starter, to the power distribution. I separated the feeds (you shouldn't need nearly as big a wire) over the winter when i did a bunch of wiring changes and it resolved any dropping issues during startup (I think I was also having starting issues in cold weather as well due to the same issue). FYI I'm running a tiny lithium battery in my car.

Pins 1 and 4 are the "Supply" Pins 6 and 3 are the Pump Motor Connection So essentially 2 "power" wires in, 2 "power" wires out. Could you bypass the controller controller with 12v directly to the pump. I'm not sure. I never tried that. I think I used 12 awg for those 4 wires. See if this drawing helps. This is how it's wired in my Subaru.

I run 2 outputs myself. 1 for a relay that I have connected to the battery to switch fuel pump power and then the fuel pump control/pwm signal that I use to send speed. And yes I think above should be everything you need for wiring and config. One oddity I noticed though was the the prime DC% doesn't seem to be scaled by the .5/1 scaling that i have set to convert 0-100% in the tables to 0-50% out. if I put 70% in the prime it won't prime. I have to use <50%.

@Serkan as a sanity check do you happen to have the TPS Main/Sub calibration values from the vehicles?

Did you log afterwards and see the signals tracking over each other. I believe I've resolved my 5V issue noted above. I've got some new connectors coming in just to make sure the Ethrottle connection is solid... but i'm still seeing some inconsistencies in the sub sensor. I test drove the car after and i'm not seeing any faulting. Some momentary error count blips.... but the issue is so random that it will work "fine" for a day... then the next day be nothing but fault fault fault. So i'm not very trusting in functionality right now. I did not that for some reason the main signal is shielded in the chassis harness for the vehicle but the sub is not. They are both shielded together in my engine harness. I'm going to put some miles on it tonight to see how things look and i'll get some logs posted as well.

Thanks... mine is looking to be a wiring problem as well. I'll know more tonight. However last night test driving my car I lost 5V to every analog sensor. It's all supplied through 1 pin on the stock subaru harness. When i built my new harness I didn't like the look of that connector... just a weak looking pin design and such. Checked for shorts, none. Verified ECU output supply value and it was 4.97 if I recall correctly. No voltage when I checked at the EThrottle Connector. 5V at the bulkhead connector, chassis side. I'll be replacing all the stock bulkhead connectors with deutsch solid pin connectors tonight then i'll relog. It appears that something in that pin connection gradually became worse and worse until it decided to fail terribly last night. If nothing else it was a giant billboard saying "Issue Here". Seeing as how i'm a controls engineer that troubleshoots electrical systems every day it was a rather frustrating weekend trying to track down what was going on (I reseated this connector several times to ensure i wasn't getting bad connections :/) I'm still a bit lost as to why the error only seemed to show itself on the sub signal but I guess if both sensors were shifting a bit, with slightly different scaling, it could show some odd numbers. Sometimes you learn the easy way... sometimes the hard way.

I don't have a log handy but over the weekend I was logging and watching a lot because I would get repeated tracking errors. I can run a TPS calibration. Have the signals overlay perfectly. Cycle power to the ECU and be off 2%. This would almost always manifest as the Sub being at 0% while the main was at 2ish%. (as a note: The error count/accumulation does not increase while it's sitting like this) They use the same 5V, the same signal ground, and they're a shielded pair inside the same shield. My first thought was a wiring issue but I see no evidence of that. Granted I was potentially stranded hours from home so now that I made it home i'll be able to step back and go through things step by step. One thing I didn't like to see is that almost every time I calibrated the TPS the settings would change slightly. I will be doing some more testing tonight and I will try to get some logs but it's seemingly "fine" right now even though the TPS signals are not perfect. I'm beginning to wonder if it's just that this throttle body doesn't like being held at 2-3% throttle opening. (Default opening percentage with motors unpowered is about 8%).

Was there ever a better solution or resolution to this? I'm running the same throttle body that ran for over a month although I did get one tracking error that I was able to reset and didn't come back until it reared it's ugly head this weekend almost stranding me and ruining a track day. Tried 2 new throttle bodies with the same issues (well one of the throttle bodies looked really bad). I'm thinking i'm going to dissect the one throttle body that I have just to see how this is set up internally to see if I see anything odd. I was able to get home after no issues though basically tuning the PID loop to the APS sub vs the Main. I'm idling at the same type of throttle positions as the original poster 2-3% so i'm wondering if it just don't like being at that low of a position. Also brings up another question I have. How are the TPS Main and Sub signals used? Does the control loop only look at Main and use sub for error checking? is it an average of the 2 signals?

If I had to guess there's an output CAN msg from the ECU that it's expecting when it's in motion. Engine running, RPM, something. Do you have capability of datalogging the CANbus with the stock ecu in place?

Just searching through the FSM for your car it looks like that PWM controller is looking for a 0-5V PWM signal. It doesn't show a frequency though. 100Hz seems to be a common automotive control frequency though so I'd start there unless someone has more concrete info.

Have you verified base timing?

Cut from a Jaguar Service Manual that I have. PWM is Switched Low, 100Hz, 0-50% DC for 0-100% pump. I did not connect the monitor in my vehicle. Also of note i'm currently supplying it with 18V via a voltage booster.

I had V1 AIM MXS and you can send GPS speed over the CAN output on CAN1. I'll see if I can get some screenshots of my old setup tonight as I've now switched over to ECUMaster's ADU5.

I'm using a Siemens/Jaguar unit. That's what the NZEFI is/was assuming you're referring to this out of stock item. https://www.nzefi.com/product/fuel-pump-pwm-speed-controller/ You can find them easily on ebay for under $100. Look for a Jaguar XF fuel pump controller. I've got a supply of the connectors and can get you one for a couple bucks or help you with some harnessing. https://www.jaguartroy.com/order-parts-and-accessories.htm?path=%2Foem-parts%2Fjaguar-fuel-pump-controller-c2c35672%3Fc%3DZz1mdWVsLXN5c3RlbSZzPWZ1ZWwtc3lzdGVtLWNvbXBvbmVudHMmbD0yNCZuPUFzc2VtYmxpZXMgUGFnZSZhPWphZ3VhciZvPXhmJnk9MjAxMSZ0PXN1cGVyY2hhcmdlZCZlPTUtMGwtdjgtZ2Fz I believe that would be a new one. I have a wiring diagram too that I can dig up.

AUX PWM outputs over CAN is currently a requested feature I believe. Myself and a couple others have requested it I believe.

I'm actually in process of switching to the ECUmaster ADU5 because the screens are very customizable. I had the v1 AIM MXS and was very frustrated because they did not support multiplexed custom CAN configuration. Took them over a year to fix the issue (while it was already functional with the v1.2 that Link distributes) I will say the ECUmaster does have some of its own issues though. The pages are so customizable that it requires you to spend a decent amount of time setting them up. They have some templates but I didn't love them. They also require a CAN-USB adapter and those usually run about $250.

I'm not sure if the Link software supports zero'ing out gains but you'd typically tune a PID loop in order. Tune the P gain until it overshoots, then back it off a touch. Then tune the I gain to handle that last 10% of motion to get it to follow nice. Then play with the D gain and tweak it. Based on those charts I'd say you need more P gain. You can do all this testing with the motor off so you're not rev'ing to the moon or crashing into things (mentioning it because you say you have a drive home log).