Rozsko

Good to know you figured it all out. I have a little bit of misfire at idle, but I just asked only because I accidentally tripped into you topic and was quite an interesting read without a final solution.

Hey guys, I am revisiting my idle misfire situation and have a few questions on coils and HT leads. 1) In the help file there are two Subaru COP coils mentioned, FK0186 and FK0140 with 5ms and 3ms Dwell times. Are these numbers meant to be set for the normal operating system voltage and increased/decreased if voltage is lower/higher? 2) In the same setup diagram, there is a suppressor capacitor specified. What could be the effect if this capacitor is not added? 3) I just figured one of my coils is an FK0334, and the rest are FK0186. I found conflicting information about the Dwell time for FK0334, so wanted to ask if anyone knows whether this is 3ms or 5ms? 4) As I was looking into alternate coil options, I saw Dwell times specified for different voltages, but not considering RPM. In the V10 Subaru basemap though, the control table is Battery voltage and RPM based and the value tapers off as RPM increases. Is that to prevent overheating? Maybe I am not thinking this trough properly, but at 7k RPM, 1 cycle takes about 18ms, so the time itself should be plenty to use higher Dwell. 5) For IGN1A coils (I seems to like the one MSD makes), is there any special considerations or settings that need to be implemented or any major reasons against this option at all? 6) Regarding HT Leads, in PCLink help, it is mentioned it should not be spiral would nor inductive. Tried to do some research, but pretty much everything I found was spiral wound (unless solid core), with varying resistance. So is there any recommended brand or specific type that should be used? So far I found Magnecor R-100 or KV85 to be the closest to your recommendations, but even these are spiral would as per my understanding (unless spiral woulding means something different). Thanks a lot

Just out of curiosity, did you manage to bring her back to "normal" life?

After upgrading PCLink and connecting to ECU, this is the error message that came: https://1drv.ms/u/s!Ale4oyMCOgLThOcc9JetBYGfE4wwpw Don't think it is a big deal, still could upgrade the firmware (though the not responding message still came) and after that the error was gone.

Anyone keen on the new version, it is available for download since yesterday.

ok, will try that way and report back.

Understood. I can understand that, though in my case not only EGT sensors, but air temp sensors are also configured to be CAN TC, since there is not enough AN channels for all the pressure and temp sensors in my case. As I said, I understand I am the odd man out, yet, it would be nice either to have CAN TCs available as source in GP inputs, or to have more CAN AN channels.

Not sure if this is actually a problem or not, but ran into this situation the other day when I realized that I left one of the virtual aux's function on Test. Then I changed it to on/off and it worked the same way. Then I set it to off, and still worked the same way. In my case a GP output's output is a virtual aux which then is the source of the NO Brake switch as that seems not to be wired to my car to the ECU. Again, not a big deal I think, but wanted to highlight.

new record from my side: 1.8Hz https://1drv.ms/u/s!Ale4oyMCOgLThOZm9NouFVi9VnGT0w Since my last post, I fully disabled the Intel graphics card and only the nVidia is enabled. Nothing seems to really help and it's quite annoying that you (me) have a powerplant really, yet the software is not utilizing its capabilities.

What I was referring to more is his "In general expanded ability to map inputs and outputs to CAN would be nice. Radiator fan control, PWM, etc etc. " section. With regards to pressure and temp sensors I mean that in the GP Temp inputs, CAN TC inputs are not available as source, so I have to go trough a Math block as Math blocks have CAN TC available as source. Pressure sensor wise I was wrong, sorry, I did not remember well, but even in that section, some more CAN AN inputs would be nice to have. I know I am the odd man out, but I am a sensor freak, if that makes sense, and I am running out of inputs (and CAN streams) as I am using all sorts of pressure and temp sensors, including EGT, eMAP in multiple locations of the exhaust stream or multiple air temp sensors pre turbo, post turbo and post IC, just to name a few.

+1 Were running into some similar limitations with pressure and temp sensors (TCs) Plus more then 10 custom streams.

+1

This seems to be a long waited feature. Any actual plans to incorporate this into a new release? +1 from me along with some more log manipulation functions like "save as" only a section of the log. (marked in the log markers or only displayed on the screen) cut out/delete a section of a log and save it like that (to remove transient sections maybe, so mixture map could be used a little more targeted) easier retrospective log marker or log section creation. shortcut key B does not place a marker in the log if it is used during log playback. Maybe make the log marker window not a popup, rather an individual section like ECU Settings, Log manager, event log and parameters. This would help creating markers and sections much easier when pinned to the left of the screen as it would allow navigating through the log file and creating markers in two clicks pretty much. customizable number of entries (maybe with a sampling rate as well) to use in Mixture map to calculate the actual lambda as right now it is only the first 1000 data points that are used which is not too useful for long duration logs like cold start and warm up

+1 (the parameter list seems to be the only "gauge" that changes during log playback)

Updated both the NVidia and the Intel driver to the latest, I disabled Windwows virtual memory (I guess 32GB physical mem should be more then enough) and I am starting PCLink forcing it to start using the Quadro card. Tried to disable the Intel card as well, but since the Optimus architecture, disabling the Intel card pretty much reverted back to Microsoft base driver with 1Hz refresh rate (LOL), so that is not a way to go. here is a screenshot of the page struggling: https://1drv.ms/u/s!Ale4oyMCOgLThOZRUxP4dEc54gg5gQ One more thing that struck me, is that PCLink is still 32bit? Any plans to introduce a 64bit version? I guess that should help quite a bit with overall perfomance of the GUI. Anyhow, any other suggestions is appreciated. Thanks a lot

Thanks Adam, good catch. I was on the initial 6.20.40 released in April. Is there a way to get notified every time a new version is available? I am subscribed to the newsletter, and I come back to the forum time to time, but I did not see any announcement for this June release. Thanks

Well, then I am not chasing my tail. As with the timeplot, it drops down to less then 10Hz in about 1-2 minutes of logging. Will continue to experiment then. Thanks.

Ah, ok. That is set to Fast. Just so I am not chasing my tail here, what would be an expected sample rate when logging is on and you have a timeplot running on the page with quite a few parameters displayed?

Thanks Adam. Connection settings are on USB, so connection sped can't be adjusted this way. In the power management I set the Intel display chip to maximum performance, and also tried to set the nVidia card (that is also set to max performance power setting) to be used by PCLink, but didn't really help. PCLink still does not utilize any GPU cores meaning 0% load on GPU, while COU and memory continues to be negligible (15% and 80MB). Any other ideas? Also what is exactly effected by this sample rate? Tried to see in help, but did not find anything really. Thanks

Recently when I was doing some logging, I realized that the screen refreshes quite slowly and the crosshair in the tables move significantly slower compared to no logging. Initially I thought it is the usb connection, so I changed the connection settings from auto to USB, but that did not really help. Then I started to play a bit and realized that without logging started, the Runtime sample rate is about 40Hz and when I start logging it pretty much drops to half of it to be around 20-25Hz. Now if I am on a page with time plot, then the sample rate drops to about 5-10Hz. Now on the face of it, I wouldn't really care as logging is not to be watched realtime, but it seems this effects the "playback" of the logs as well. Meaning, if I open a logfile, and start walking though it, the graphs and parameter values are refreshed only a few times a second resulting in quite big jumps of the crosshair in the tables, despite the graphs being quite nice and smooth. So, the first question is, can you explain how these things really relate to or depend on each other? Question two is why is the Runtime sample rate dropping so drastically when none of the laptop components are overloaded (CPU~15%, mem~30%, GPU~10%)? Final question, is there anything that can be done to improve this situation? Thanks Béla

Sorry to hear the lockdown. I've been through that (just like many others), I know it is though. Glad to hear the timing though. Thanks.

Sorry to get back to the quick tune popup window problem, but I tried that on a different install and it is the same, and the Tune Cell button never becomes active in manual mode, although maybe this is due to some setting, but I can't access the settings checkbox either due to the sizing of the window. Did you guys experience the same?

Any news on the next firmware release? When can it be expected? Thanks

Hi there, Here are couple of issues I wanted to share with you. Not sure if these are actually PCLink or Windows related, but anyhow. - The quick tune popup window is cropped, so not all the controls can be seen. (https://1drv.ms/u/s!Ale4oyMCOgLThOYMqZfor8-VKWFMbQ) Unfortunately the pop-up can't be resized. - Sometimes (I actually wasn't able to determine what triggers this actually) when I want to save the tune, when I click the save button in the dialog pop-up, nothing happens. It acts as if I never pressed the button. PCLink is not frozen, as if I press the cancel button, the dialog pop-up disappears and the system continues to work. And then if I disconnect the ECU USB cable, PCLink suddenly displays a File saved success message. - If I have a tune open and I connect to the ECU, there used to be message saying, the tune I have open is different from the one in the ECU and asks me if I want to copy open tune to the ECU. Now this message does not appear in about 80% of the times. Thanks, Béla

Quick update - PROBLEM SOLVED! First I tried the engine capacity increase, and that worked well. If I increased the capacity to 4700cc, lambda was bang on. This obviously confirms that no mechanical issue is present, and yes it is a fuel related problem. Now back to your MAP sensor comment Adam, what stick out to me, but previously I did not pay attention to, is that idling this time happened around 30kpa, while in the past, that was more around 47kpa. So, I replaced the 4bar OMNI sensor with a 3 bar OE sensor I used in the past, changed the calibration, and voila. Engine runs perfectly. Couple of more things I checked/did. I put back the OMNI, calibration (liner)was good, but baro pressure was shown as ~70kpa, so the MAP sensor calibration function did not work due to the 5kpa max difference allowed. By changing the calibration to a 6bar sensor, I could get the baro pressure read 0 and do a calibration, but I guess this clearly confirms it is a faulty MAP sensor. I also looked back the log with the brake cleaner and actually I should have paid more attention there, since when the engine stalled, the MAP sensor was reading ~70kpa. One more thing I had in mind is if this is a wiring problem, but since the 3 bar OE sensor with the Subaru V9 calibration works just perfectly, I think we can rule that out. Long story short, big thanks again for your help. Beyond the G4X being an awesome product, this is why I love Link! Awesome support!!!