MartinS Posted September 30, 2021 Report Share Posted September 30, 2021 Hello, have been working on getting my EZ36 running with my link G4 Xtreme. Some problems but everything going pretty well. Spent the last week or so sorting some issues with the cooling system and getting the settings so the I could start and run consistently. Couple of days ago, I got to the point where I could start increasing the RPM and setting up a basic tune. When getting a little over 2000 RPM, I noticed a slight miss. Investigating this, I can find no fuel or spark fault with the engine or ECU. What I did notice is that, as soon as the misfire starts, the 'Trigger Error' count goes from zero to 255 in a matter of seconds. No trigger errors at idle and up to 2000 and both trigger 1 and 2 have signal. Have searched topic and checked as much as I can. Triggers appear to be wired correctly and have shielded cable. Ive checked the timing with a timing light and it is set correctly at TDC #1 cylinder. The triggers are set as the defaults for EZ30/36 engine included in the software. I've looked to run a trigger scope but it looks like my G4 can't do that. Will attach setup file as separate post. Cheers, Martin Log 30-09-21 1;08;29 pm.llg Have tried to upload .pcl file but the site says it's over 37.12 kB and cannot be uploaded. Quote Link to comment Share on other sites More sharing options...
Adamw Posted September 30, 2021 Report Share Posted September 30, 2021 Yeah you definitely have a trigger issue. The only way you will figure out the problem is with an oscilloscope so we can see what is wrong with the signal. Do you know anyone with a scope - or maybe a G4+ ecu you could borrow? Quote Link to comment Share on other sites More sharing options...
MartinS Posted September 30, 2021 Author Report Share Posted September 30, 2021 Ah, wondered. Nobody I know has a G4+ but a friend has an oscilloscope. It's basic but I could take a photo of the output screen. How would I hook it up to get the tracing needed? Any chance it's a trigger arming threshold issue? I could easily adjust the table.... Quote Link to comment Share on other sites More sharing options...
Adamw Posted October 1, 2021 Report Share Posted October 1, 2021 If its a two channel scope, connect 1 probe to trig 1 pin at the ecu and the other to trig 2, connect one of the ground clips to any ground. If it is just a single channel scope then you will have to get a capture of trigger 1 & 2 seperately. 1 hour ago, MartinS said: Any chance it's a trigger arming threshold issue? I could easily adjust the table.... That is one possibility, but really shooting in the dark without a scope. Normal arming threshold for a subaru 36T crank would be something like below. Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 1, 2021 Author Report Share Posted October 1, 2021 Ok, I'll borrow my friends scope over the weekend and send some screen shots... Thanks Adam Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 2, 2021 Author Report Share Posted October 2, 2021 Hi Adam, I have some photos of the oscilloscope screen tracings of my triggers. I've tried uploading them here but they are bigger than the allowable 37kB. I've tried re-sizing them but they still come up as too big. Can I e-mail them to you? Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 2, 2021 Author Report Share Posted October 2, 2021 https://drive.google.com/file/d/1FgKFFiOnMWci631VIqDE3CfxIVd2wc7X/view?usp=sharing, https://drive.google.com/file/d/1HWn9K2x3bjVWin_AVxVf8N4QnOxVxvZ4/view?usp=sharing, https://drive.google.com/file/d/1K9XzZUPN5QtClSY3S77Zlbuk073acUrF/view?usp=sharing, https://drive.google.com/file/d/1Vny38m-whvIKM1M_tE7VN8sZw1MsVGUs/view?usp=sharing Google Drive links to photos. Didn't think of this initially. Should work... Martin Running engine today to bleed cooling system. Thought I'd play with the arming thresholds. Tried the thresholds you sent me as they are a little higher than mine. No change, still miss at 2000 RPM. Played around with the thresholds and found by entering 0.5 in all of the windows, the engine will not rev up to 3200 RPM without the miss/instability occurring. Is there an explanation for this behaviour? Cheers, Martin Sorry, will now rev up to 3200 So, improved quite a bit... Quote Link to comment Share on other sites More sharing options...
Adamw Posted October 2, 2021 Report Share Posted October 2, 2021 The main problem will be these big spikes of noise coming through on the cam signal. It looks like there are 6 spikes every 360degs so if you are running wasted spark this would likely be the spark events. This could be something like a poor grounding scheme on the coils or maybe non-resistor spark plugs. Can you confirm spark plugs are resistor type and give a run down on the coil wiring. It is also a bit weird how the cam signal is not flat - the floor tapers up from left to right. I suspect this may just be a scope setting - maybe channel 2 is set to AC coupling or something. Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 2, 2021 Author Report Share Posted October 2, 2021 Will do Adam, thanks. I wasn't sure about that spike. I'm running direct spark. Hoped it was maybe something to do with cranking. Checked the spark plugs already, have 5k Ohm resistor so suppressed. I'll check how I wired the ground for the coils and look at the scope settings concerning the falling waveform. What do you think about the fact dropping the trigger threshold under 1 let's the engine run 1000 RPM faster before the fault occurs? I'm having trouble understanding that. Martin Quote Link to comment Share on other sites More sharing options...
Adamw Posted October 3, 2021 Report Share Posted October 3, 2021 Sorry, it is 3 spikes every 360deg, so that fits with spark events still. 7 hours ago, MartinS said: What do you think about the fact dropping the trigger threshold under 1 let's the engine run 1000 RPM faster before the fault occurs? I'm having trouble understanding that. Likely not relevant. It should rev all the way to the limiter if the trigger was acceptable. Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 3, 2021 Author Report Share Posted October 3, 2021 8 hours ago, Adamw said: Sorry, it is 3 spikes every 360deg, so that fits with spark events still. Likely not relevant. It should rev all the way to the limiter if the trigger was acceptable. Ah ok. Could this be a faulty coil? I guess that would be one event every 360 though... Alternately, is there a spike buried in each of the cam signal waves meaning there are 6 spikes per 360deg? The spacing of the waves and spikes argues this could be happening. I've checked my wiring diagrams and, while the ECU is grounded to the engine block, the ignition coils are grounded separately to the frame. The 6 individual grounds run to a common wire on the stock engine wiring loom which is then taken to the frame. I'll make sure that ground spot is secure and tight today. The coils also have a separate relayed/switched power supply running from relay to common coil wire on engine loom. Ok, one obvious thing just slapped me in the face: the ECU is disconnected for these oscilloscope screens. I'm taking the tracings directly from plug A on the wiring loom and I also unplugged B as I couldn't recall which was which and have them marked. Thus there is no spark event possible. Does this mean faulty sensor then? I would have thought that would not create a discreet regular spike. I'm going to try and tap into the signal lines from the triggers today so I can take some scope readings with the engine running. Hopefully that will sort this out... Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 3, 2021 Author Report Share Posted October 3, 2021 Hi Adam, Took a bit of doing (not an electronics guru here) but I can get scope tracings with the engine running now. I've linked to three photos of some of the scopes. One is idle, one is 2000 rpm no misfire, one is 3000 rpm with misfire and escalated 'trigger error' count. The two above idle scopes are the same scales; I've had trouble getting the scope consistent, it's adjustments are very twitchy. To be honest, they look good to me! I hope you can see something which explains the misfire and trigger error issue. Let me know if a scope with different parameters would help. Thanks, Martin https://drive.google.com/file/d/13SdZl8_UqZL5EqgWOJks0nk6EcBLvOC8/view?usp=sharing, https://drive.google.com/file/d/1BPXrM0IX0DY4N7p32Puq2Cw8cmkUEYYG/view?usp=sharing, https://drive.google.com/file/d/1tY5LdW48R9kWaewGyAbFGCvF6QUIdyQ6/view?usp=sharing Quote Link to comment Share on other sites More sharing options...
Adamw Posted October 3, 2021 Report Share Posted October 3, 2021 Those files dont have the right share permissions. Go to sharing options and set them to something like "anyone with link" Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 3, 2021 Author Report Share Posted October 3, 2021 Ooops! Sorry, corrected... Quote Link to comment Share on other sites More sharing options...
Adamw Posted October 3, 2021 Report Share Posted October 3, 2021 Crank signal is all messed up in those captures. Can you give us a photo of the knobs on the scope so I can see the settings. It looks like it is scaled in amps instead of volts. Can see now why the lower arming threshold works as some teeth are barely rising above zero volts. It looks like the crank sensor is missing its ground or something. Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 3, 2021 Author Report Share Posted October 3, 2021 Yes, scope is scaled in amps. Can't see how to get it into volts. I'd show you the knobs but they have no scales or settings on them. It's a cheap scope and getting a to give any kind of tracing is a challenge. Should the crank sensor be grounded to sensor ground? I will check it... Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 4, 2021 Author Report Share Posted October 4, 2021 The sensor is connected to a shielded cable with a black and white wire within. The black is signal and the white is the shielded ground. Connections look good. I will test continuity in the morning. Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 4, 2021 Author Report Share Posted October 4, 2021 18 hours ago, Adamw said: Crank sinal is all messed up in those captures. Can you give us a photo of the knobs on the scope so I can see the settings. It looks like it is scaled in amps instead of volts. Can see now why the lower arming threshold works as some teeth are barely rising above zero volts. It looks like the crank sensor is missing its ground or something. When I first set the system up, with cranking I was getting trigger errors. I went through the wiring of the crank trigger, convinced myself I'd wired it backwards and switched the wires. Your comment about sensor ground made me think about that. So, this morning I switched them back as an experiment. The cranking and running scopes now have more even amplitudes in the crank trigger and, to me, overall appears more consistent. I have included links to a scope of running and cranking. Now get crank errors but mostly with cranking, I don't think much with running but hard to tell (see below). However, only 39 crank errors although I don't know if that's the whole session of 8-10 start/run cycles or just one. Still a lot less than when misfire happening. The problem is the engine barely runs now and stalls after a few seconds (15 or less) with much misfiring. The AFR, which had plenty of time to warm and give a reasonable reading during the session read extremely rich (off the scale at 7 much of the time). I tried turning the Master Fuel down to 4 from the 20 I had it at which had the engine running very nicely (except when misfire occurred above 2000 RPM) as well as several settings in between and even up to 25 to make sure I wasn't going the wrong way. Only ran worse or didn't run no matter what I did. I have included a link to a log of the engine running (poorly) with the new trigger setup. Any insight into this would be greatly appreciated. Thanks again, Martin https://drive.google.com/file/d/1W9Sl7_hniaRQgmukMWU98sb7O1JNRTwQ/view?usp=sharing, https://drive.google.com/file/d/1mHZM4bObmvySBgUungWRJknkku8-ZZ5h/view?usp=sharing https://drive.google.com/file/d/1qnL23qk-UgeLxMkHj7x5XCkJxJjjDSEH/view?usp=sharing Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 4, 2021 Author Report Share Posted October 4, 2021 Sorry, the log file above is incorrect, link to correct file and current setup file below. https://drive.google.com/file/d/17KTFbQCYPVnhwdNWoTTlFp52rMdrZJXj/view?usp=sharing, https://drive.google.com/file/d/1wcv-TAwaGQEvs3xIhPK7kyCDo0AgfzL1/view?usp=sharing Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 4, 2021 Author Report Share Posted October 4, 2021 Better log of running with trigger polarity switched: https://drive.google.com/file/d/15MMunm5QZgENKYt_NDsvYIP4JioIzEKg/view?usp=sharing Post start enrichment set to 0, ran a little better... Quote Link to comment Share on other sites More sharing options...
Adamw Posted October 4, 2021 Report Share Posted October 4, 2021 Crank sensor polarity was already correct. It is now wrong. Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 4, 2021 Author Report Share Posted October 4, 2021 Ok! Good...sort of...still have misfire/crank errors at increased RPM. Quote Link to comment Share on other sites More sharing options...
Adamw Posted October 4, 2021 Report Share Posted October 4, 2021 With crank sensor unplugged, measure resistance between ground pin at the crank sensor plug back to ecu ground (any ecu ground will do) and also to engine block. It should be near 0 ohms for all. Also pull the crank sensor out, and check if there is any resistance between either of the pins and the metal body or the sleeve for the mounting bolt. It should be OL or many Kohms. Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 4, 2021 Author Report Share Posted October 4, 2021 Ok, will do in am Quote Link to comment Share on other sites More sharing options...
MartinS Posted October 6, 2021 Author Report Share Posted October 6, 2021 You nailed it Adam! Pulled the crank sensor, tested it, no continuity between pins and sensor body or mounting so no issues. Tested plug, continuity on signal wire from plug to ecu: continuity. Tested ground from plug to ecu: nothing. Tested from engine plug/socket to ecu: continuity. From engine socket to sensor plug: nothing. Wondering what the hell I was going to do next, I looked down the 18 inch hole behind the throttle body the crank sensor lives in, reached in to pick up the plug to look at it, and the ground wire fell off the plug. Cut the signal wire and pulled out the plug to look at it, and the cable had literally been hanging on by the insulation. Right at the plug itself. The wire was quite corroded, it's been failing for quite some time. So, like I said, you nailed it: sensor ground fault and I think I got extremely lucky to find it right away. Having said that, it was not an easy fix, mostly because the wiring to the plug lives under the intake manifold about 12 inches below the top of the intake manifold. Took me the better part of 2-3 hours to re-pin the plug and splice in repair wires. Almost took the intake manifold out but got it done without that hassle and a little quicker...I think. So, next time, could you find me a problem in a little bit more convenient location? Seriously though, thanks Adam, you got me on the right track immediately. I'd had enough for one day so haven't got the engine fired up again so I'll let you know if there's something else but I'm actually quite confident the misfire issue has been diagnosed and solved. Cheers, Martin 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.