Adamw

Please do a triggerscope and save as log then attach here

It looks like most of it is working and it is not far off running. Have you checked coils are connected to the correct cylinders using the ignition test function? Have you check base timing is roughly in the ball park while cranking? Is there fuel pressure/fuel pump working? Also the log shows quite a high inj PW while cranking so that may flood it quite easy, you might need to reduce your master fuel.

So how is the RX8 ECU getting an RPM signal? Do you have an RX8 triggerwheel on the 1J as well? Is the tacho on the dash working?

Since it is an adjacent pair of cylinders there are many possibilities. I have seen all sorts of fails over the years, so you cant assume anything, you need to eliminate the most obvious/logical possibilities as completely as you can first then move on to the next. Just some random examples of what I come across before that could potentially affect compression on 1 bank; Cyl head bolts/studs bottomed out on the threads before the head clamps tight. Head gaskets on upside down or back to front. Small fire ring gasket on oversize bore causing leak between cylinders. Valve lash too tight or lifters not bled. Cam timing. I have even come across things like rags left inside the ports/manifold...

Does it still have the factory ECU in place or is the Thunder taking care of the CAN bus messages for the dash and steering etc? How did it work before - did it still have an RX8 ecu in there too?

Im not sure how you have been testing it, but be aware the CAN Lambda wont work until the engine is running. It wont even start heating until it gets an RPM signal from the ECU.

You will need a MAP sensor for either ECU. I dont generally bother with Lambda as a protection strategy, I have never found it to be fool proof enough even with ECU's that have much more complete lambda based protection strategies built-in. My preference is usually to monitor differential fuel pressure as the main fuel related engine protection - this wont catch things like a bad connection on an injector wire however. It is still really nice to have Lambda for logging and tweaking the tune however and if you plan to do any tuning yourself you will of course need Lambda. If budget is the priority another option is to go for the Storm and later when you can afford it get a CAN bus wideband, either the Link one or there are some decent lowcost options such as Ecotrons ALM or 14point7 Spartan 3.

Since there is some evidence of a compression issue based on your starter current, the first thing I would do is buy a cheap compression gauge and do a basic warm WOT compression test. Get that possibility eliminated first.

Although that "nc" you have circled above is for pin 5, I think that pinout possibly has an error for pin 55. I only have an outdated schematic so nothing concrete, but from reading between the lines, It looks to me like pin55 is probably controlled by Aux 12 on the current boards. They would have done that because Aux 4 cant be used with E-throttle. Try aux 12 in test mode and see what happens. Maybe @Simon will confirm tomorrow.

Note HP Academy is nothing to do with Link, you should ask them.

With the engine running can you give us some photos of the CAN tab and the ECU Status tab of the runtimes screen (F12).

Actually, try something more like this. It still wont be foolproof but it should work more like what you want. Having the close breakpoints will effectively remove most of the interpolation, You will still need to be careful to press the button somewhere near a 1000RPM increament as if you pressed it at say 4400, RPM it would only need to climb a little and it will jump up to the next limiting range.

Can you do another triggerscope now it is running. Your camshaft signal in the last captures didnt look good so it needs to be checked again.

I dont think it would be possible with an Atom. The problem is as soon as you turn on the limiter, your RPM drops, so does your target limit. All the other Link ECU's have latched launch mode (sometimes called rolling antilag) to achieve this but it was left out of the Atom firmware. I think the closest you could do would be to have a potentiometer or multi-position trim switch connected to an analog input so you can vary the limit.

The 5 wire cable is the problem, this means the serial pins are connected, it should only have 2 pins (CAN H/CAN L) connected. The easiest fix would be to either cut the grey and yellow wires near the ecu or pull the terminals out of the plug.

Download the Link plug & play data for your model injector from here: http://help.injectordynamics.com/support/solutions/articles/4000074340-link-engine-management Set the fuel pressure you intend to run in the spreadsheet, then copy/paste the data straight into PC Link.

Yes. Just copy the settings out of this file into your map. You can open two instances of PC Link side by side to make it easy.

This file is not for any engine in particular, I think from memory it was based on the Link RX7 basemap.

How is the ECU connected to the CAN lambda? Just two twisted wires or with the CANDASH cable that has 5 wires inside?

The limiter will start to come in at 6700 minus the limit control range, so I would set the lockout to more like 6400RPM. I have little experience with these engines sorry. Spark duration will have no effect with a direct spark system. It really only comes into play on V8's or high RPM 6 cyl engines still using a distributor. With a distributor you have a very limited time to charge 1 coil 6 or 8 times every cycle so the ecu needs to know the spark duration so that it can reduce dwell if it is going to overlap with the previous spark.

I think the "running rich" around idle and warm up could well be explained by the new injectors operating probably well into their non-linear region. However I dont have much to offer as to why it would be running significantly leaner at higher loads unless they dont actually flow what they are supposed to. I dont see any obvious mistakes in your fuel settings. Im not a fan of the CNG injectors as a lot of people see large variations in flow with fuel/injector temp and also a decrease in flow as they age, but dont have much personal experience. Check the basics such as the MAP sensor is still reading correctly, but otherwise I would be doing a check injector test in a jug or similar to confirm flow rate.

Note, it sounds like you have some other (possibly wiring) problem from what you describe, but just FYI, you do not need a custom cal table for that sensor. Just set the calibration to "1000Kpa (150psi) TI sensor".

You will need to move the wire, Pin 106 is not connected to anything on the 300Z ecu. The injectors being "hot fed" is not a problem, only the genuine aux outputs can backfeed if hot fed, ign and injectors drives are ok to be live all the time (at least on ecus that dont have P&H drivers).

DI1-8 are all fine for speed inputs. Where does it say only 1-6 can be used? It is possibly a typo or an old note that someone missed left over from the older G4 ecus.

As per JMP's suggestion, some ECUs also have serial comms pins in the same 6 pin comms connetor. So if you use the old CANDASH (not recommended) cable which had serial wires connected inside it will interrupt PC comms. Follow the CAN wiring instructions in the CAN Lambda manual and it will work as expected.