koracing

Not really a slick professional way to tap those pins that I can think of short of rewiring on the back side of the fuse box to an external 2 wire connector maybe. The solid state relay option - if you can find one that drops into the AC relay location - would be a cleaner install.

The boost solenoid will be operting via a PWM funciton - pulse width modulation. That means it will be adjusting how long it is off/on at a given frequency - usually 20-32 Hz 0r 20-30 times a second. Most relays won't be able to repeatably and precisely control that level of precise movement as they are mechanical devices. I wouldn't try to switch a relay faster than about 1-4 times per second and even then the mechanical armature will likely fail in short order. A solid state relay without mechanical moving parts and with sufficient output response time could be used for this. Alternatively, just remove the AC relay and use the pins 85 and 86 to run to your boost solenoid. That would provide your switched 12v power via pin 86 via the 5A F16 fuse, and then the pin 85 would go to ecu pin 48 (IGN7 output). Also - Merry Christmas!

How did it show 1.06V before and now shows 0.02v? Irregardless, hopefully a new sensor solves it.

Do you have the AEM sensors working anywhere else on the car? It seems like it's pinned ground and signal backward or something?

Location should be no issue. I've done the same thing many times. I have my chassis dyno extra sensors (all the same AEM stainless sensors for 9+ years now) set up with these type of in-line port adapters for input to the dyno analog inputs for those (many) vehicles that don't come equipped with their own sensors. I use them to monitor differential fuel pressure (dyno software calculated difference of dyno MAP and FP sensors) most often, and sometimes oil pressure.

If you have an air compressor you can try testing the pressure sensor with compressed air also to see if the voltage changes in the ecu with that. I've only had 1 AEM stainless sensor failure in the past 10+ years of using them.

Couple things I've seen: don't have multiple copies of the software open at the same time. Set the Connection Port to Auto instead of picking a paricular COM port, or vice versa select the com port that shows up in the device manager if it's already set up for Connection Port -> Auto. If you post some information about your issue connecting like the images requested above and a brief description of what's happening, that my also help. Is this a new install or was connecting before and now it won't?

The way an analog input works is the wideband controller (or other gauge with analog output) outputs a 0-5v signal to the ecu and then the ecu reads in that voltage on an analog input and then uses a calibration table to read that input and interpret the voltage to a corresponding value. Due to the nature of the gauge using electrical components with tolerance to generate that 0-5v signal, and then the recieving ecu using electrical components with their own tolerances to receive the signal, there can be differences in both the slope and offset of the signal received versus the manufacturer's claimed calibration. The calibration *must* be checked and corrected such that the ecu reading matches the gauge, and in many cases these values will change over time as well due to various reasons. The beauty of CAN is that there is none of this interpretation at either end and you get a clean signal reporting what Lambda/AFR is without and electrical components doing an analog interpretation. I'm not sure what you mean by a "value of 0.5v on ecu". I would ground all sensors to the ECU ground or the same spot as the ecu grounds, but power for each of the gauges I would run to the same spot on the chassis/engine the ecu grounds, but not go *throught* the ecu epansion pin as there is a limit on current capacity there. The wideband itself may pull 3-5A at warmup so I'm not sure a single 5A fuse for all of that will work ok. I would run the ecu on it's own dedictated fused circuit separate from the gauges so if there was ever a short in the prosport gauges (kind of cheaper gauges) it would not take the ecu out of operation. Also a lot of what your gauges are monitoring should be already monitored by the ecu or could be if you wired the sensors direct to the ecu, and then you could display all those things in one gauge (Link CAN gauge) instead of having a bouquet of gauges all over, plus the ability to set up warnings and have failsafes set up in the ecu based on those things (low oil pressure for example).

Analog inputs have a calibration interpreting the 0-5v signal which works and has been done for a generation, but is prone to offset and drift of the calibration particularly if you don't have a wideband with a ground output along witht he analog signal output. This can lead to chasing the calibration to get it to match the gauge. The biggest advantage of CAN is that you get the actual lambda value reported by the controller directly to the ecu, without any interpretation. Using the CAN JST harness on the plug-in and then running at least the 2 wires (CANH and CANL) to the controller should be all that is needed other than switched power and ground to get the wideband installed. The CAN setup in the ecu has great device specific information for both the AEM and Link lambda options with built in streams that can be loaded without entering everything manually. The Spartan3 comes with pretty decent documentation instructions if memory serves also. If you still needed help setting it up, lots of support on this forum can help get you set up. There is certainly the ability in the ecu to do Coil on plug - looking at the MX5 NB1 pinout - there are two additional igntion outputs on the main header (ignition 5 and 6) that could be repurposed for coils, but you may need to rewire those functions to other spots. Alternatively the Link could be modified to provide Ignition 3 and 4 outputs to allow for COP by Link before shipping to you. Fuel pressure compensation in the fuel model when in FP Sensor mode is not the same as adjusting dead times. It means that given all the parameters of what the engine needs for fuel at a given point in the fuel table, if the fuel pressure were to drop from 60psi to 43psi, the ecu would automatically compensate for the pressure differential and theoretically provide the correct adustment to the commanded injector pulse width to provide the same amount of fuel. I think this is what you were asking? The AEM CAN and Spartan CAN are less expensive than the Link. AEM is $185-$200 depending on if you want a gauge or not. (actually AEM 30-0310 in line controller is on sale for $166.46 direct right now and the 30-0300 gauge type controller is $180 - leave it to Holley to sell direct for under MAP pricing). There is no need for a knock block as the NB1 has a knock sensor and the Link can monitor it. I would stay away from Innovate wideband controllers in particular.

I also find it hard to believe someone in New Zealand would design something new today that had SAE fine thread...

FYI M5x0.8 and SAE #10NF32 thread are usually interchangeable. (M5x0.8 is 31.75 thread per inch and 3/16" diameter - 4.76mm)

Obviously having them flow tested via a bench top flow testing unit would be one way. Emperically in the tune I'm sure there's a clever way, but it would probably require a bit of iteration to derive flow...

@Adamchong Yep all sorted once I changed the ID to the 100psi setup. Thanks!

AEM X-Series wideband connected via CAN or Link CAN Lambda are the two widebands I typically use. 14point7 Spartan 3 via CAN also looks decent. I also always recommend a fuel pressure sensor for the ecu fuel model to compensate for any fuel pressure fluctuations and/or possible future troubleshooting.

I think he said he pulled two bulbs out.

Having the wrong frequency on the solenoid can also change the response range of the solenoid into a very non-linear response. You should bench test or "ear" test the solenoid with different frequencies and should hear clear clicking with a good frequency from 10 or 15% duty up to 80-90% duty. Also there are a *ton* of knock off mac solenoids that have terrible response and are super sensitive, so where you get the solenoid from makes a big difference in how it responds as well.

If the voltage is 1.06 - that doesn't seem like a dead sensor. Does the voltage change on that screen if you unplug and plug the sensor back in? Does it change if you turn on and off the fuel pump output? If you can attach a copy of your tune file that may help to figure out why it isn't starting if it's related to the fuel pressure. Is there a mechanical pressure gauge on the car? What does it show doing the fuel pump on/off test? What exact sensor are you using as there are some sensors using the same connector with a different pinout.

My 2 cents: If it doesn't have an apparent difference while testing I wouldn't get too worried about it. If the default setting is working, I would go with that. Ultimately when mapping or tuning, you're going to set the ecu to what the engine responds to for values in the timing table, and that's what matters most. Having perfect timing offset means that the timing values in your timing table are accurate all the way to redline versus the real world mechanical crank position. Having it incorrect may mean there is some difference between what the ecu is commanding or "thinks" the timing is versus the actual real world. What is the practical difference in this potential couple degress of difference? In my opinion, not much since as stated previously you're going to tune to what the enigne wants in either case. This also only can be checked/set when the ecu has a fixed timing value. You'll need to have the ecu set so that the trigger calibration window is open while checking this to fix/hold ecu timing.

My feeling is the Link is a bit more advanced in capability versus ECU master - but I will admit my knowledge of ECU master is limited to messing with a few of the EMU Classic ecus. I haven't seen what their black version does differently. Fuel pressure compensation in the fuel model on Link is one of my favorite things about Link. Link has a lifetime warranty on their ecus, factory direct support via this forum, live chat support chat on their site, and via email with Link tech support also aid in selling the Link products.

If you are very far from compression in the cylinder there may not be enough current flow to actually get a pulse the timing light will pick up. I would try changing your offset by 30 degrees at a time and see if the timing light starts to flash.

Yeah you will change the trigger in the ecu as follows: Triggers -> Trigger Setup -> multi-tooth/missing Trigger 1 -> Trigger 1 -> Multi-tooth position -> Crank; Tooth Count -> 12; Missing teeth -> 1 Then re-verify your sync position/trigger offset of course.

So all you need is a second 12v battery in series to power that one solenoid, lol. J/K please don't do that.

So the Trac fuse didn't take care of it?

If you're not getting a reply on the email - try using the live chat on linkecu.com and get the email address of whoever is chatting with you to send the request to so it can be forwarded to the correct person. Any chance the necessary information could be provided here for a field fix in the case someone doesn't want to can't send in their ecu? Is this an issue for VVT on all similar solenoids across the G4x lineup or just plug-ins? Is a recirculation diode the same as a flyback diode? Perhaps like a 1N4007?

Can you provide a copy of your tune and the actual log file? Something is strange with AP main at 65% and E-throttle target at 89% still.