Adamw

Not quite sure what signal wires you are talking about?

The trigger settings look ok, personally though I would go for the stock distributor over the EPM, they are horrible devices. But I dont think it is your problem in this case. A couple of things that I would look at first: You have almost no accel fuel. All the accel fuel settings look way off anything I have ever needed to use. So I would suggest you open our 96-99 Honda base map side by side with yours and copy the accel settings across as a better starting point. You only have 1.6-1.7ms dwell, that may be a bit on the low side for K20 coils. I would try it up around 2.0ms. Lock base timing to around 20deg and give the engine some good revs while watching with a timing light, make sure timing is reaonably stable and doesnt drift. I have seen some EPM's that produced a large timing drift with RPM.

I have never heard of Syltech, what does the datasheet give for calibration values?

If you set the CAN2 mode to anything but off, then it is "On".

No, with GDI and a dual clutch transmission on a relatively uncommon platform your aftermarket options will be very slim. You are probably better to investigate reflash options.

You should send it with a multiplier of 1000. You currently have a multiplier of 1, which means the parameter will lose the resolution when truncated to a integer for transmision. For example a lambda target of 1.050 will be transmitted as 1 or a lambda target of 0.875 will be transmitted as 0. If you use a multiplier of 1000, then a target of 0.875 will be transmitted as 875.

Since this post is 10 years old and possibly not relevant to you, can you give us some information about your ECU and what you want to do, maybe attach a copy of your tune also.

Share a link to it using google drive, dropbox, onedrive or similar.

I can assure you there is built in internal termination on CAN 2 but it is difficult to measure due to the hardware that is used to switch those pins functionality from DI to CAN. The termination is only connected when DI9/10 are switched off and CAN2 is switched on & ecu is powered up. When there is data on the bus you cant measure the resistance. I suspect if you have CAN 2 set to on but all of the channels set to off and all other devices unplugged from the CAN2 bus then you will be able to measure the termination resistance.

Just make sure the aux output is set to "ISC Solenoid" instead of "ISC Stepper" is the only set up needed. The "Stepper motor type" setting is always visable in the ISC settings but is not used when the aux is set to solenoid. You can test the wiring/valve by setting the aux output to "Test (PWM)" and set a low frequency like 10Hz and you should hear it clicking. Normal frequency to set it to after testing would be around 150-200Hz. If still no joy then attach a log of it idling and a copy of your tune.

RPM limit is set to zero so you have 100% ignition cut. Also fuel is turned off so you have no injection either.

Is the blue power light on on the ECU? Yeah I noticed that a while ago and asked if if can be changed. Not sure when that will happen. Note it is not a particularly useful method for checking deadtimes for injectors any bigger that about 500cc as when you halve the pulse width you will be going deep into the non-linear range. I find a better method is to adjust the lambda target up/down 5-10% and adjust deadtimes until measured lambda closely follows a target change.

Yes, Injector or ignition drives will not backfeed so are fine with a hot feed.

Post a log showing the issue and a copy of the tune.

Most likely, the waveform amplitude and the noise floor are effected by many things that the ecu has no control over, but usually the values above will be a good starting point that will work for most. The ECU has a trigger scope built in that you can use to confirm arming thresholds are accptable later if you have any doubts.

That typically means the exhaust manifold pressure is very high and exceeding the spring in the gate.

Turn off channel 2, you should only have one channel assigned to CAN lambda with the base ID 950. You can receive up to 8 CAN lambdas all on that one channel.

If its a G4+ then you have all of the missing features I mentioned. G4+ was our current generation up to around a year ago so it is a good capable ecu.

Can you attach the tune, a short log of some cranking, and a triggerscope.

Are you sure its a G4 and not a G4+? You say you bought it 2 years ago but we stopped making the G4 in 2013 so it would have already been well obsolete when you bought it in that case. If purchased new 2 Years ago it should have been a G4+. Having said that, it will still run the engine well, the main things missing compared to the newer ecus would be configurable CAN bus (can still run a dash etc tho), knock control, proper flex fuel support, more tables, some of the increased flexability in IO and a few inputs and outputs. Apart from updating to the latest firmware, there is no other hardware updates that can be performed.

Try this. Fury_FD2_K20A CAN AC.pclx

No, the MSD6A can only be used with a single coil and also cant be used with any coils that have internal ignitors.

That should be correct. I think their document is probably wrong. Im pretty sure I remember reading somthing years ago that the AEM load value was scaled in MBar based on a 2 Bar map sensor or something really odd. As a quick test, set up your stream like below. Parameter "5V Analog Supply (status)" will just send a zero out all the time, so you can then use the offset to send different numbers over the bus. An offset of 1 will send "1" over the bus, an offset of 1000 will send 1000 over the bus etc. So try an offset of 1, 100, 255, 1000, & 65535 and record what the gauge displays for each of those, that should give you a pretty good idea what is going on. Edit: found this on their forum too, so one of them is obviously wrong. Better check yourself:

Attach your map. What CAN port is connected to the car bus?

Post some actual info if you want help, my crystal ball isnt coming up with anything. Log, map, triggerscope, description of the problem would be a good start.