Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


Everything posted by paulr33

  1. paulr33

    programmable logic

    Good work. I ended up going for a plex usdm but making the arduino CAN display did help me learn it all and how it works which was good fun
  2. I still think you’ll find the tablet a half bake setup where to have to press buttons every time or wake it up each time or load software. Hope it all goes well but from making my own CAN display to using an as-built one (the plex uSDM) the commercial product is a million times better. Means I can focus more on tuning the car and not troubleshooting display issues
  3. The problem you’ll find is that it will eventually cost you more in connectors and other bits and you’ll have to solve a heap of problems such as standby, powering on etc and customisation. i made my own CANBUS display thinking it was going to be cheap ended up about $300+ and it still had issues and the plex uSDM 102 is $750 AUD (non pro version) so stuffing around with a tablet or display is pointless. Have you tried reading data from a tablet at an angle in your car while driving? I bet you can’t even read it based on the angle and display type. just trying to save you from re-inventing the wheel
  4. Plex usdm 102 and be done with it, no need for any displays or tricks. Just raw CAN and fully customisable. I did a thread on it the Vipec section
  5. Is there a recommended rpm I should be travelling at when I do gear ratio calc in each gear? Evo 8 runs 6 speed close ratio and currently I get quiet a bit of “gear 0” in my logs. I’ll re do all calcs in every gear from scratch, I was thinking 3500rpm in every gear when I do calibrate ? also for the onboard boost control is there a specific solenoid you recommend? I have an eBoost2 and want to use per-gear control so can I re-use that solenoid or should I get a new Vipec compatible one? The wire to the ecu “aux 2” on my v44 evolution 9 is this a Switched ground to the solenoid?
  6. ok ive found the issue, its how its calculating per cyclinder and the overall aggregate, see below. it says global level is 0 but it shows numbers per cyclinder still ill do some work on this tonight and see what the issue is, thanks for your help
  7. see attach 40hz log from my car tonight. the log data pattern for knock follows engine load so i think the sensor is working, but again it still shows the data pattern where in one cell its 0 knock level then the next cell its almost at the max range, i.e Section time: 98.225 - knock level global: 0 Section time: 98.25 - knock level global: 620 Thanks for your help Log 22-03-19 9;03;34 pm.zip
  8. thanks , i think i have a better picture of it now, so basically log at the highest possible rate (40hz) then use that to build a knock map of what the engine looks like under good health without detonation
  9. Out of curiosity I tried gain level 2 but with noise cancellation and the level appeared the same so I’ll just disable knock control
  10. from the manual Later model vehicles (2000 onwards) may have difficultly operating with the Internal knock detection circuitry. This is due to the use of "2nd Harmonic" knock sensors that some manufactures are factory fitting. There are two solutions to this problem: 1. Contact your nearest Vi-PEC dealer for the purchase of an after-market knock sensor. I am trying to use the factory 4g63 sensor but happy to replace with an aftermarket unit if someone can tell me what sensor will work with the link knockblock and what sensor settings to select on the box, i.e FREQ setting
  11. see attached tune and .log file. see these section times section time 694.9 in the log section time 922.8 in the log tune.zip
  12. before changing anything i would look at the idle control setup and idle functions. the online help covers how to set up "closed loop" idle control, do this first and then when its stable check out open loop idle control. from memory on a turbo engine, map is the recommended axis (again check the help online in VTS or link software)
  13. is it the knock block noise cancelling mode that is causing that see-saw affect of the knock "level"? my box is set to FREQ level C and gain is 2 which im certains is for evo lancer 13khz with noise cancelling
  14. Hi all, Can anyone show me what their knock level global values look like? i am curious what other data logs look like for this type of data set as mine doesnt make sense. the values are all over the place. i know my engine is noisey as its a 4g63 stroker with balance shaft delete but the values go from 0 to 800 to 400 to low numbers etc. i checked the values in the log and the values on the chart are all in the "knock window " range. See below graph Left is RPM in blue Right is knock level global in red The concern i have is that the data pattern looks like knock, i.e going from low levels of engine noise to excessively high noise quickly which is what knock patterns look like. so i use the data set to fine tune the knock targets but i just end continually increasing the knock target values each time higher and higher. This is 4g63 stroker kit, balance shaft delete, external "link knockblock" with factory knock sensor and Vipec v44 evo 9 model. i can attach the tune and logs if this helps but more curious on what do other knock level patterns look like. why isnt it linear with engine load / peak torque instead of see saw'ing between values?
  15. Ok thanks, maybe I read it wrong and saw the other values to be multiplied by 5 in the help. All good thanks
  16. noticed a strange issue where in the CAN channel config if i enable knock level global, knock threshold and knock counter cyl #1 all in the same stream the knock level global is no longer required to be multiplied by 5. that is, the number that comes out in the knock level global packet is the actual correct value and doesnt need * 5 to display the correct value, but only after i enabled the other two CAN channel items. Vipec 9 PnP, Firmware 4.10.2 Not urgent just an odd issue i found
  17. took a while sorry, but finally have video
  18. i run innovate LC2 and it works great for wideband input and wideband closed loop on v44
  19. made some good progress tonight. a few values to out how they are displayed, but working with the vipec v44 correctly.
  20. The Vipec v44 and possibly other models require math computation on some of the values so they are displayed properly, so for example temperature is offset by +50 so you need to tell the display this logic otherwise your airtemp will show as 80deg when it’s actually 30. the plex uSDM 102 Pro does this via a feature “math channels” but the function is used for more advanced logic such as showing average of two sensors, using cosine, log etc I.e averages of two EGT probes to be shown as a single value or used for alarms on the display. The software for the plex uSDM 100 and 102 Pro both allow the user to configure an “offset” for every sensor value in the CAN packet so you can say airtemp is airtemp with an offset of -50 thereby solving the wrong values issue. the offset feature is different to the “math channels” so it looks like on the 100 Pro you can run this on a Vipec v44 and use the offsets to do all the calculations and therefore you don’t HAVE to buy the 102 pro version purely for the “math channels” function. in summary you can go for the cheaper option if required. Wish I had of known that when I ordered mine
  21. Hi All, Thread for setup guide of the plex uSDM 102 Pro CAN display on the Vipec v44. My display should arrive this week so I’ll post up info and details on getting this working on a Vipec v44 (my model is Evo 9 topboard PnP).
  22. I’ve purchased a plex usdm 102 pro CAN display so I will post up some info once I have it working etc
  23. sorry its locked, there is no place in 2018 for locked tunes. make sure you tell all your friends the workshop does this. its a beginners trap if you arent advised upfront and find its locked later. can you still read logs off the ECU when its locked? you could reverse look at the logs to see how its tuned etc.....
  • Create New...