Eric S Posted September 21, 2020 Report Share Posted September 21, 2020 Hello, I have recently installed Link G4X for MX5 on a naturally aspirated 1996 MX-5 NA (with custom 3-plug harness adapter), but 50% of the time having issues starting the car potentially due to trigger sync loss (trigger 1 error counter going up during cranking); whether the engine is cold or hot. Sometimes the car starts fine normally, but sometimes I need to repeat the cranking few times which is frustrating in cases where the engine is stalled mid-traffic etc. Using the "Mazda MX5" trigger mode, with RPM filtering set to 1 and trigger offset is set to 0. The car only has camshaft angle sensor (CAS) on the exhaust camshaft (it spits out 2 signals), it does not have any sensors at the crank. I've checked the timing is correct when the engine is running. Checked with trigger calibration (locked to 10 deg BTDC) along with a timing light. Also tried changing the trigger edge type to both falling and rising, it only appears to start when trigger edge type is set to ALL. I haven't touched the wiring as it uses factory harness. Attached trigger scope log when the car is not starting, and the other one when the car is idling. Thank you! Eric TriggerScopeLog not starting 1.llgx TriggerScopeLog running idle.llgx Quote Link to comment Share on other sites More sharing options...
Adamw Posted September 21, 2020 Report Share Posted September 21, 2020 Can you tell me what firmware version you are using. >help>ecu information. It looks like there was a change to MX5 triggering in 6.17.8 so if your firmware is older than that then that would be the first thing to try. Quote Link to comment Share on other sites More sharing options...
Eric S Posted September 21, 2020 Author Report Share Posted September 21, 2020 Hi Adam, I'm using 6.18.8. One thing I noticed is that the voltage on trigger scope only reaches ~3.8V, I wonder if voltage threshold is not high enough to register as "high". I can have a look with a oscilloscope and see if there are any noises, or signal offsets etc. Other than above, I'm running out of ideas apart from giving a go at wiring the CAS directly to ECU instead of using the factory wiring. Quote Link to comment Share on other sites More sharing options...
Adamw Posted September 21, 2020 Report Share Posted September 21, 2020 Can you attach a copy of your tune please. Your voltages and waveform looks fine to me but the "not starting" scope capture shows the ECU is not happy so I want to pass this on to engineering to take a look. Quote Link to comment Share on other sites More sharing options...
Eric S Posted September 21, 2020 Author Report Share Posted September 21, 2020 (edited) Map file attached! Edited October 4, 2021 by Eric S File deleted Quote Link to comment Share on other sites More sharing options...
Adamw Posted September 22, 2020 Report Share Posted September 22, 2020 I have passed some questions on to engineering, I will hopefully feedback tomorrow what they think. Quote Link to comment Share on other sites More sharing options...
Eric S Posted September 23, 2020 Author Report Share Posted September 23, 2020 Thank you, shall wait for your feedback Quote Link to comment Share on other sites More sharing options...
Adamw Posted September 23, 2020 Report Share Posted September 23, 2020 Ok, I have talked to the firmware engineer and he cant see what is causing the problem either. It appears to be a firmware issue but we cant reproduce what you are seeing on the bench. There were some recent changes made to the MX5 trigger mode so the engineer asked if you can roll back to 6.16 firmware and try it to see if it is related to these recent changes. I have attached the 6.16 firmware file. To use this perform a firmware update process as normal, but use the browse button to navigate to and pick the attached file. If this test doesnt give us any clues then I think the next step will be to give you a special firmware version with some extra debugging tools added to give us more info. G4xApplication-6.16.27.img Quote Link to comment Share on other sites More sharing options...
Eric S Posted September 23, 2020 Author Report Share Posted September 23, 2020 I only had a chance to start the car once this morning before driving to work with 6.16.27 firmware. The car started OK with no issues, I'll give it more start-ups after work today and see if the trigger sync looses randomly again. Attached trigger scope traces with 6.16.27 for your reference. Would you mind sharing what the changes with MX5 trigger were between 6.16 and later version of firmware? Coming from electronics/electrical background makes me very curious. TriggerScopeLog.llgx 2020-09-24 fw v6.16.27 run 1 cold start OK.llgx Quote Link to comment Share on other sites More sharing options...
Adamw Posted September 23, 2020 Report Share Posted September 23, 2020 I dont have all the details on exactly what was changed, but it was something to do with the early 6.16 only worked fine with wasted spark ignition but didnt always sync correctly with direct spark ign, so it was a fix to correct that. But the weird thing is the Evo 1-6 trigger mode uses exactly the same code/trigger pattern as MX5 so it seems there are plenty of users out there running the new version successfully. Not sure why it is picking on you... Quote Link to comment Share on other sites More sharing options...
Eric S Posted September 24, 2020 Author Report Share Posted September 24, 2020 (edited) I see, thanks for sharing. So I started the engine few times in both cold and hot conditions, and it starts up well just like a factory ECU with 6.16.27. Started 7 times in total; 4 when cold and 3 when hot. Attached few more trigger traces, including the sync loss during cranking. Feel free to let me know if you'd like me to do some tests. Edit: I recall a jumper on PCB with silkscreen etched "MX5 1800", should that be jumpered? TriggerScopeLog.llgx 2020-09-21 1800 fw v6.18.8 not starting.llgx TriggerScopeLog.llgx 2020-09-24 1616 fw v6.16.27 run 2 cold start and idle.llgx TriggerScopeLog.llgx 2020-09-24 1616 fw v6.16.27 run 2 cold start OK.llgx TriggerScopeLog.llgx 2020-09-24 1617 fw v6.16.27 run 3 start OK.llgx TriggerScopeLog.llgx 2020-09-24 1654 fw v6.16.27 run 4 hot start OK.llgx Edited September 24, 2020 by Eric S Quote Link to comment Share on other sites More sharing options...
Adamw Posted September 24, 2020 Report Share Posted September 24, 2020 Can you update me tomorrow if 6.16 is still working reliably, then can you update to 6.18 and confirm that the no sync issue comes back. I will then feed that back to engineering and see how to diagnose from there. 1 hour ago, Eric S said: Edit: I recall a jumper on PCB with silkscreen etched "MX5 1800", should that be jumpered? This just connects a 12V pull-up to a circuit on the ignitor, this is needed to make the tacho work on some 1800 models. Quote Link to comment Share on other sites More sharing options...
Eric S Posted September 25, 2020 Author Report Share Posted September 25, 2020 (edited) Yep will give you an update later today. Also my car doesn't have trigger wheel at the crank, only CAS on the exhaust camshaft and it made me wonder if 36-2 trigger wheel can be used in conjunction with existing CAS? For instance, setting up trigger 1 on PCLink to use "multi tooth / missing" and trigger 2 set to "Cam Pulse Window" to sync? Not sure if the waveform of two uneven pulses (see attached), can be used to sync. Edited October 4, 2021 by Eric S deleted cas waveform photo Quote Link to comment Share on other sites More sharing options...
Adamw Posted September 25, 2020 Report Share Posted September 25, 2020 Yes, you could use a missing tooth crank wheel with a two tooth cam using the "cam pulse window" feature. But it will be nice to get this OEM trigger working correctly first. Quote Link to comment Share on other sites More sharing options...
Eric S Posted September 25, 2020 Author Report Share Posted September 25, 2020 Good to hear that, most forum searches (miata.net etc) I came across said that I need to remove the one of two rotor "blade" in order to generate only one pulse. But as you said, I'd like to see what the issue with trigger. So, good news is that 6.16 appears to be solid, 1-2 seconds of cranking and car starts right up. But bad news is after updating the firmware to latest version 6.18.11, starting the car became a hit-and-miss again. It's very strange since the trigger scope looks identical between "not starting" and "starting". For time being I'll use the 6.16 firmware. What do you suggest next? Quote Link to comment Share on other sites More sharing options...
Adamw Posted September 25, 2020 Report Share Posted September 25, 2020 Ok, so now we have confirmed it is related to that recent firmware change I will feed that info back to engineering and get some advice. They were talking about possibly having to give you a special firmware with some extra debug variables added to give us more info as it syncs reliably everytime when testing on a simulated engine. Unfortunately the main Firmware engineer is away most of next week so there will be a bit of a delay before I get back to you. You will have to stick with 6.16 for a while longer sorry. Quote Link to comment Share on other sites More sharing options...
Eric S Posted September 26, 2020 Author Report Share Posted September 26, 2020 No problem, will wait for an update from your side. Quote Link to comment Share on other sites More sharing options...
Eric S Posted September 30, 2020 Author Report Share Posted September 30, 2020 Just want to bump if your firmware engineer had a chance at looking into the matter? Quote Link to comment Share on other sites More sharing options...
Adamw Posted October 1, 2020 Report Share Posted October 1, 2020 As I previously said the firmware engineer that does most of the trigger code is away most of this week. You wont hear anything until probably this time next week. Quote Link to comment Share on other sites More sharing options...
Adamw Posted October 7, 2020 Report Share Posted October 7, 2020 @Eric S Engineering believe they have found the problem and resolved it. I have PM'd you a beta firmware to test. Quote Link to comment Share on other sites More sharing options...
Eric S Posted October 7, 2020 Author Report Share Posted October 7, 2020 Awesome! PM received. Will give the beta version a crack tonight and report back. Fingers crossed Update: So far beta firmware looks promising and the cranking sounds better than the older 6.16 firmware (sounds more smooth?). I encoutered only one trigger error count. Log attached for reference. I'll drive around next few days and see if I experience any start up issue. Cheers PC Datalog - 2020-10-7 11;25;42 pm Beta FW start OK, one trig error.llgx Quote Link to comment Share on other sites More sharing options...
Eric S Posted October 13, 2020 Author Report Share Posted October 13, 2020 @Adamw Had zero issues so far with beta firmware (apart from one trigger error count log posted above), whatever your engineers did appears to have fixed it May I ask what the changes were internally? Quote Link to comment Share on other sites More sharing options...
Adamw Posted October 13, 2020 Report Share Posted October 13, 2020 29 minutes ago, Eric S said: May I ask what the changes were internally? I suspect even if I got a basic explanation from the firmware engineer I wouldnt have enough understanding of the system to pass that info on accurately. Quote Link to comment Share on other sites More sharing options...
Eric S Posted October 18, 2020 Author Report Share Posted October 18, 2020 No worries. I'm still yet to have any starting/sync loss issues with beta firmware When the official version gets released, I'll update it and carry on. Thanks for your help! Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.