Jump to content

Leaderboard


Popular Content

Showing content with the highest reputation since 10/25/2018 in all areas

  1. 3 points
    Simon

    G4x Base map

    For those looking for a base map to have a play with. Software can be found at http://www.linkecu.com/pclink/PCLink 6.15.15.exe Monsoon FW 6.15.6.pclx
  2. 3 points
    TechDave

    Share your Math Channel List

    So this afternoon I sat at the simulator and started mucking about with a few ideas. I wanted to calculate my average speed based on some event, I used a digital input latching on. In the UK we have average speed camera zones, you get snapped at the start and end and they do a distance over time calculation. Often you'll get slowed down by a Karen in an X5 not paying attention, I've always been curious how much I could safely speed up again to catch back up to 50 mph average over the trip. You can do this in log analysis with the statistics view, but this is obviously operating at runtime. So there's a few things here I've done that aren't documented yet, yes they will be. The first is that maths channels can do reassignment, x = x + 1 etc The next is that they're evaluated at 1000 Hz Lastly I know of a few built in functions too. I figured out two methods, of averaging, the first could be used not only for wheel speed, but for really any parameter, lambda, throttle position, you could use another maths function to find lambda, vvt, boost, etc error then feed that function into this one and have an average of error over some time period. This one uses this algorithm: newAverage = oldAverage + (newSample - oldAverage)/sampleNumber the cnt function is a real good cnt function, it counts 1 per cycle (so 1 per millisecond) while c is greater than or equal to 1, so if you feed it a status which evaluates to 0 OFF and 1 ON, then you're away. If you want to use this for Anti-Lag active we might have to get a bit imaginative about an offset, iirc Anti-Lag OFF = 0, ON = 1, Active = 2...etc I had to add a +1 to cnt because on the first cycle it's 0 and that's div by 0, plus the first cycle is the 1st sample too. The 2nd method I came up with was to calculate distance, which of itself is a useful function, then calculate distance/time, again using that good cnt function: Again using reassignment in the Distance function, multiply speed by 1/100th of a second, of a minute, of a hour to get distance travelled per millisecond (rate is 1kHz), I multipled by status to simply switch it on and off. This gives me km. distance a over time, divide by the same multi as above gets us back to an average. You don't need to use 8dp for this, I was just using it while getting my exponents sorted. My functions agree with each other and track to my expectations, but if you spot an error in my maths then shout out edit: unfortunately this is all in kph as I started running out of characters to add the mph multiplier, I've spoken with engineering and we'll see if we can get the length increased, 2.77e-7 helped a lot, but still lost a bit to parenthesis too.
  3. 3 points
    Ducie54

    G4x Base map

    As a feature request can we have the option on which Lambda sensor is used for CLL. As example I'm using 5 lambda sensors. One in the turbo dump pipe and 4 in the turbo manifold using a CAN module. The software uses a average for CLL but would be nice to use Lambda one so manifold pressure is not an influence. Currently to get around this from happening I've set the 4 CAN lambdas as CAN TC.
  4. 3 points
    Simon

    G4x Base map

    Yep Multi fuel and traction control will be coming
  5. 3 points
    Adamw

    What is wrong with my tune?

    I just had a quick look. There is some pretty odd stuff in there. It looks like it has been tuned for flex fuel but there is only one ignition map and no other corrections for ethanol content. Usually E85 can take a heap more advance than petrol so what fuel was that ignition table tuned for? VVT looks like the offsets are set incorrectly. The RH inlet cam is showing 10deg advance when it is in the fully home position. That means it will be getting 10Deg less than what is commanded. The VVT and ignition tables both have big "holes" in them around the cruise area that is going to make it drive pretty horrible. Boost control not working great at all. Fuel control is not great either. I would say this is due to there only being 1 VE table set up, which doesnt usually cover flex fuel well enough especially when injectors havent been flow tested on both fuels. I think the saving grace is you are running it on relatively high eth %, which is relatively forgiving. I wouldnt give it a hard time especially on petrol until it is looked at by someone more competent or interested. If you want to PM me the tuners name I will pass it on to the right people to consider.
  6. 3 points
    Adamw

    X-SERIES AEMNET CAN BUS GAUGE

    this is all you would need to send lambda from a fury internal wideband to the aemnet gauge. this would be channel o2-1 on the gauge.
  7. 3 points
    Thank for Realdash developing such a powerful app that I don't have to buy expensive AIM racing instruments. My car is Subaru GRB, ej25 with link G4 + thunder. I developed a protocol conversion module using STM32. Then can convert the Can signal of the link G4+ into a Realdash protocol, and then connect to the mobile phone through a TTL to Bluetooth module, so as to display almost all the data on the mobile phone (as long as the PClink can display). Welcome to exchange SM-A6050_20181227180733.mp4 https://www.youtube.com/watch?v=fNeBGatmBSc https://www.youtube.com/watch?v=eKJzVRUp09c
  8. 3 points
    mapper

    Nos control

    I think its much better to have all universal function like virtual aux, timers, 4D, 5D, fual fuel table etc. than a dedicated NOS function. This way you can build your own NOS function excatly how you want it. You just habe to be a bit creative with the universal functions.
  9. 2 points
    You can connect both 5v and 12v sensors (and VR sensors etc that have varying voltage output) all to the same sensor ground. The reason to chose 1 type of ground over another is not input voltage, but more: 1) how stable does the signal need to be for the engine to run well (crank trigger = really important, any of the temp sensors = not a big issue if the signal bounces around a few percent). And as part of this, will a ground offset cause an incorrect reading that still looks legit. (eg voltage offset on any analog sensor) 2) How much load is expected to go through the ground circuit. eg do not use a sensor ground as the ground side for a vtec solenoid or anything like that. As long as everything on the sensor grounds is low *current* then adding a flex fuel sensor to sensor ground wouldnt hurt, and generally anything that is an ECU *input* you would lean towards putting on the sensor ground circuits. But if you do see issues that seem like sensor noise once you add it, you can probably move the flex sensor to use a direct ground and not see too many issues. These dont report an analog output anyway so should work just fine with slightly noisy voltage/ground.
  10. 2 points
    Brad Burnett

    AEM Water/Meth Boost Safe

    Can wire that wire into a digital input on the ecu. Then could configure the ecu to change boost, timing, and fuel values when that is active.
  11. 2 points
    TechDave

    Fuel film modelling in G4X line

    Accel fuel has been changed to a simplified X-Tau model
  12. 2 points
    Ducie54

    G4x Base map

    Is it possible to have a Math block drop down box with the maths all ready worked out?
  13. 2 points
    Simon

    MR2/ST205 and Knock

    Highly likely.
  14. 2 points
    I think with engine temps of 10°C and 94% ethanol you are going to struggle with cold start and there will be little you can do from a tune perspective to help. The vapour pressure of straight ethanol is very low so there will barely be any vapor in the port at that temp no matter how much fuel you dump in, it will just remain as liquid. Somewhere below around 10°C there is not enough vapor produced to support combustion. This is the reason why pump E85 gets reduced to E60-E70 in winter so there is a little more of the volitile solvents available to create some vapor in the port. In regions like Brazil etc where they predominantly use straight ethanol it is common to use heated injectors to get around the problem.
  15. 2 points
    Yep sounds like you got it. Here's a quick demo: https://1drv.ms/u/s!AiYbYlZQuRHPqylX67m5Oo57-JVv?e=t3SK3f
  16. 2 points
    Ok, all the CAN is done, I havent added any of the new channels to displays or whatever but they should now be available. I have kept all the channel names pretty much as per the BPM datasheet. The only one that is a bit ambiguous is Fuel level. He has used 2 bytes like it is expected to cover a large range of numbers but then says the range is only 0-53L. Just one byte can do up to 255 and 2 Bytes is capable of 65000. So I suspect he might be sending it as "0.1L" rather than whole Litres but that is not clear. If your fuel level says "500 Litres" when you have it all working then set the scalar to 0.1. Marvin-LinkG4UpdatedV2Port2CAN Adam.aemcd7
  17. 2 points
    hexdmy

    tps and MAP accel fueling together

    Bumping this topic... I'm surprised no one else has interest in this. OEM systems I have worked with do this, and it makes sense. At low throttle openings, a change of throttle angle is a better trigger for transient fueling than map for engines that have low / noisy idle vacuum. However, at steady throttle angles, on a turbo engine, as the boost builds, there can be a need for additional fueling, otherwise you get a lean spike. I've noticed this when tuning steady state and then doing transient sweeps, if you tune to the sweep, you get a inverse V in your fuel map. Certainly this is not reinventing the wheel, just doing what the OEM's have already been doing for many years, and for good reason.
  18. 2 points
    Yes, Spartan has a termination resistor built in and it is enabled by default. I would use the Spartan in its default mode, use 1Mb/s and 1024 as ID. Use the attached stream definition. From the testing I have performed, the Spartan doesn't actually work in Link mode, the stream appears to be incorrect. (you would have needed to log in via serial bluetooth to change the mode/ID to Link CAN Lambda mode) HTH, Richard. Spartan3ADVDefaultMode.lcs
  19. 2 points
    I wanted to share my success using a bluetooth adapter since I have not seen any other posts about using one on this forum. So far I have only tested it with an Android tablet and RealDash. It has been working perfectly. Connection is fast and very reliable. There is no noticeable lag on the dash interface. There are a few advantages of using Bluetooth over USB for a tablet dash. 1) The usb port on the tablet is freed up to be used for charging (this can also be used to automatically turn the tablet on/off with key power) 2) The usb port on the ECU can be dedicated to the Laptop usb cable for tuning (dash cannot be used simultaneously with laptop, but less plugging/unplugging) This is assuming that your Link ECU has separate tuning and CANBUS ports. The Bluetooth Adapter I have been using is one made for Megasquirt MS2/MS3 – PerfectTuning is the manufacturer. https://perfecttuning.net/en/accessories/26-bluetooth-adapter-for-ms2-and-ms3.html There are a few other BT adapters for megasquirt, they are all likely the same Bluetooth chip and work the same. The Megasquirt uses a DB9 connector for serial connection, so that is what comes with the BT adapter. In order to connect to the Link you will need a Link CAN connector http://dealers.linkecu.com/CANF_2 These connectors need to be soldered and can be somewhat difficult due to their small size. Cable pigtail is also available but fairly expensive. The following is the pinout on CAN connector, the BT DB9 connector and where those wires need to end up. This can be wired a few different ways. Either the DB9 connector on the BT adapter can be mated to with a female DB9, or it can be cut off and hardwired (or use a different style connector). The two serial wires will go to the CAN connector on the ECU (TX to RX, and vice versa). The 5v and GND wires will need to go to a 5v power source. Do not use the 5v supply from the ECU, the BT adapter can draw up to 30mA. 12V DC to 5V power supplies are very common, it is the same voltage as USB chargers. What I did was buy a USB charger and hardwire a USB cable pigtail to make the connection. I used and like this model: I recommend powering the 5v power supply from the ACC terminal (radio power) on the ignition switch. This way there is still power to the Bluetooth with the key in position 1 (ignition off). This will allow you to power cycle the ECU without killing power to the Bluetooth adapter – leaving it wirelessly connected to the tablet. This means that the dash will be live in about a second instead of 5-10sec. In order to use the tuning port to connect a laptop, either the CAN connector must be removed or a 2 pole switch/relay must be put on the serial wires coming from the CAN connector. The switch must be within about 20cm of the ECU, otherwise the serial stream may become active and not allow the laptop to connect. Once the laptop has been connected, the ECU must be restarted to allow the serial stream to become active again. Hopefully in the future this can be changed by Link so that the serial stream is reactivated upon laptop disconnect. If there is interest and the forum allows, I could possibly make and sell a plug-and-play Bluetooth solution for the Link G4+. Cost would likely be just a little more than the Perfect Tuning adapter. Here is a video of my dash on a “cold start” – Powered up and connected in less than 10 seconds. Realdash can be set to allow the tablet to sleep when charge power is killed from the tablet. It works well as a OEM style dash and I don’t need to manually touch the tablet at all. This is what the inside of the PT BT adapter looks like. Blue = 5v Orange = GND Grey = RX Yellow = TX
  20. 2 points
    Richard Hill

    Thunder lambda wiring - 12v

    Here are some measurements from a single 4.9 Lambda sensor running on a Thunder on the bench. The maximum the sensor drew was 1.43A at the end of the heating phase
  21. 2 points
    Adamw

    RPM 65535

    No, the dwell scatter is a symptom of a trigger error. It means the ECU isnt receiving the regular tooth pattern that it expects from the trigger. As I said earlier, the quickest way to find your problem is to get an oscilloscope capture of the triggers - preferably when it is playing up. The Oscilloscope will show us what is wrong with the waveform and this will usually give you some strong clues what is causing it. Do not waste time changing random parts - there could be many hundreds of different scenarios. Your logs show the ECU is not happy with the trigger so you need to start by diagnosing what it is about the trigger signal that the ECU it is not happy about.
  22. 2 points
    Sorry I haven't logged in for ages. Have just seen your comments are re-uploaded photos, should all be good now.
  23. 2 points
    I see, I wish that it would always ask if I wanted to load write the changes to the ecu, no matter how large. Because when it just loads the config out of the ecu and doesn't notify about anything, it's easy to think there were not any changes in the tune file. Then you carry on, tuning the config loaded from the ecu, until you realize there is no filename associated with the open tune file. Then you have to open the old file and see what's different, plus the changes you've recently made to the config file that was automatically loaded.... I've lost a good bit of time this way. If it's not possible for PCLink to ask if you want to load a tune with large differences, there should at least be a dialog that notifies what's going on (tune uploaded from ECU is replacing the open tune).
  24. 2 points
    Simon

    S15 plug in

    Jump on and download the PC link tuning software. The help file will have 99% of your questions covered.
  25. 2 points
    Adamw

    Link monsoon wiring

    I think I would avoid that, it looks very suspicious. Look at the main connector, it is missing the housing and looks like the wires are attached directly to the pins??
  26. 2 points
    Adamw

    ECU Unlock fails

    Someone has done a typo by the looks, The last two characters should be b7, not 67.
  27. 2 points
    Would be nice to be able to see at a glance if edits were additions or subtractions since last marker reset. Say on fuel table you add to a cell than the cell turns blue and if you subtracted the cell turns red.
  28. 2 points
    On my setup I use a 3d table to PWM an engine water pump. After a lot of testing I found using the temp output from the radiator and ECT the best way to control the pump. Takes into account out side temp and vechile speed. Would recommend you try that way to control the thermostat. Ive also connected a 2nd Digital output to the same soild state relay as an override. So if I'm at a set TPS or boost range the pump goes to 100%. After a bit of testing during different conditions I can keep the ECT within 2deg in all situations.
  29. 2 points
    Adamw

    MINI LINK G4+ Borg Warner Speed sensor

    I drew this circuit for someone the other day. I havent tested it myself with a turbospeed sensor so no guarantees but I have used a similar circuit to divide a wheel speed frequency down. Over here you can buy the 4024 binary counter for about $1.50 from Jaycar so I assume they should be easy to find anywhere.
  30. 2 points
    Richard Hill

    programmable logic

    I moved away from using generic dash as it employs a compound CAN message, which can be a bit of a mouthful for both the ECU sending and the device receiving. I prefer to use messages which are split over several CAN IDs such as transmit generic dash 2. The only problem with Generic Dash 2 is there's appears to be a bug in the non driven wheel speed (1 byte assigned, but shows as 0-1000kph, guess the text was copied from generic dash). Not a problem though if you aren't planning on going more than 158 MPH...
  31. 2 points
    Im not sure if you are asking a question or making a statement? Of course the fuel table and ignition tables (and all other tables) will need to be tuned to your specific engine before it will run well. It sounds like you dont even have a good grasp of the basic fundamentals yet so Im reluctant to advise you touch it yet. I would not attempt to tune a turbo car with this level of understanding. I would suggest you start by doing something like the low cost EFI tuning fundamentals course here to get a better understanding of the theory behind it: https://www.hpacademy.com/courses/efi-tuning/ If you try to tune without a good understanding of the basic theories then it will just end in disappointment.
  32. 2 points
    Simon

    [UPDATE] PCLink 5.6.6.3564 (on hold)

    Yes it is now ok to run the current release.
  33. 2 points
    Ducie54

    What is wrong with my tune?

    Another look show it's reducing boost target in higher gears but gear detection is not set up correctly. Your vechile speed is showing when in 0 gear.
  34. 2 points
    I would say the timing doesnt look unusally high for 8.2CR with E85, but it also doesnt look tuned at all to me, it looks like someone has just made up the numbers. Why have so many rows with exactly the same numbers on them? You would be better to just delete them all. Seems pretty odd to have the same timing at 207Kpa than you have at 138Kpa. And as JMP says, that big step between 1500 & 2000 is gnarly.
  35. 2 points
    Adamw

    G4 link new E-throttle module wiring

    Attached below. 1&2 are sensor ground. 3&5 are supply voltage (I assume 5V) 4&6 are AP sub and main. E-throttle Module Instructions.pdf
  36. 2 points
    TechDave

    ARF Choice

    http://dealers.linkecu.com/can-lambda CAN Lambda
  37. 2 points
    We see lots of wideband controllers here, the worst for sensor failure are AEM, Innovative seem to be better, there is a noticeable difference as you really don't see sensor failing that often on the LINK CAN Lambda, we put the sensor failures down 100% to the heating strategy not shocking the sensor and allowing time for residual water / condensation to be cleared. The other thing is that the AEM sensor supplied with the kit does not appear to be bosch, but although you could argue that is the reason they fail, replacement genuine bosch dont seem to last that long either. That is the main reason we try and get people to buy the Link sensor and we always ask about the placement of their Lambda bung. PLUS you get EGT with the Link CAN Lambda an invaluable device that has saved my customers $1000's in turbo rebuilds.
  38. 2 points
    Just wanted to clear the VAG coil wiring diagram. These coils have two ground connections when in the OEM installation . One is for the secondary coil - wired to the cyl head. The second is for the coil driver - wired to the chassis/ecu ground.
  39. 2 points
    There we go. Granted the RPM line itself is smoother than what i suspected. Could be because of slow logging rate. What gives it away to me though is looking at the RPM ROC parameter. Thats RPM Rate Of Change. How much the RPM changes over a second. Anything over 0 meens RPM is rising at a certain rate and below 0 means its falling. We bot KNOW you are accelerating so the CRANK is certainly not slowing down, right? However you can see i both my screenshots that the ECU is logging a very erratic RPM ROC. It should ALWAYS be on the positive side when the engine is accelerating. That IS the CAS flapping around. There is also some knock registered in one of the logs. I bet because of unstable timing. Never mind me saying it looked smoother than expected. I just wasnt looking close enough at it. Its actually one of the worst i have ever seen You sir, are going to install a cranktrigger before it breaks down completely
  40. 2 points
    Adamw

    Closed Loop fuel trims

    Yes, in the next 5.6.6 firmware the CLL trim has 2 fully configurable 3D tables, so you can have different limits for adding fuel than for subtracting fuel and have those limits mapped against MAP, RPM, or almost any other parameter you like. I dont think the release is far away.
  41. 2 points
    Adamw

    Bosch DBW 0280750101 / 82mm

    The Dodge Ram settings from the help file are usually quite a good starting point for most of the Bosch throttles.
  42. 2 points
    Here you go: https://1drv.ms/u/s!AiYbYlZQuRHPn3zpsxhdV-DI1p5D
  43. 2 points
    Stevieturbo

    water meth injection

    You will never see 300psi, nor will 2x1000cc nozzles be as efficient as smaller nozzles. Well other than when the pump is pumping against the closed valve and if the pump has been adjusted to see a higher pressure. In fact, with nozzles that large I'd be surprised if you ever see much over 100psi. The pumps simply do not flow what the suppliers claim, and any pressure claims are against a dead head. I've tested the AEM big pump under load and in the real world at pressures required....realistically they're only around a 1200-1300cc pump, and that was with a quoted 1600cc worth of nozzle...and tested not fighting against boost pressure which would reduce flow further. Even with only a 600cc nozzle...my factory set 200psi pump, could never create 200psi. With 1600cc worth of nozzle, in the car, in real world use max pressure I see is around 140psi....so if using say 40psi boost, that's now only 100psi of effective pressure to inject, so even less flow. But stick a pressure sensor on the line and log it for yourself and use this to control any fuel/spark adjustments, rather than what you're currently doing which offers no failsafe. It's a cheap and easy way to monitor what the system is doing, and could be an eye opener. It's a pity nobody seems to have done much testing on the various systems and pumps out there to prove or disprove the suppliers claims.
  44. 2 points
    Adamw

    12v PWM Output

    The aux outputs have a pull up resistor to 12V, so they are 12V when "off" and ground when on. These can be directly connected to an aux out with nothing extra. As Richard says though, the logic may be inverted ie 0% = max flow 100% = min flow, I have set one up before but cant remember if they work that way or not. What is important though is they need a 3ms "wake up" pulse, so to do this you have a cell in one corner of your PWM table that it will always pass through during at startup (say 500RPM), then you can drop back to normal numbers after that. Use 150Hz frequency and 50%DC to give the 3ms. Read this for the min/max DC: https://www.tecomotive.com/download/PWMinfo_EN.pdf
  45. 2 points
    ayjayef

    Trigger Errors - troubleshooting?

    Good to hear I'm not the only one ever to be confused with that. Even knowing to imagine another wave in the missing teeth gap makes my brain hurt. I can see it sometimes. That fixed it! Was 245deg and now -119deg offset (after a 4deg timing light adjustment). Trigger scope during idle looks good and NO TRIGGER ERRORS! Starts easy now, timing is steady, idle is good. Now we can actually start tuning. Thanks for your patience and odd hours troubleshooting! Log 2018-11-12 9;11;38 pm-idle.llg Trigger Scope Log 2018-11-12 9;08;33 pm-idle.llg
  46. 2 points
    cj

    2JZ Bosch 74mm DBW idle position table

    Did you have a mechanical throttle tuned with reasonable idle before you swapped to the DBW kit? There is no magic number for ethrottle idle and it depends what RPM you want to idle at, and how much ignition advance you have. More ignition = less throttle needed for the same RPM (and vice versa). Assuming roughly stock internals and ~15* timing, you probably want the throttle angle to be about 3-5% once warmed up, and up to 5% higher at 0 Deg. Before you even worry about fine tuning the idle base table, warm up the engine then zero out the idle table, start with your ethrottle target table: the top "0" row - you want somewhere between 2 and 4 in the columns around idle (500-1500rpm?), and either tapering out to 0 above 2-2.5k, or holding around 2% in that whole top row (depends on how much engine braking you want - zero's = more engine braking). The idea here is that once warmed up, the engine should be close to but just under normal idle rpm with the idle table zero'd out (much like setting idle screw on a mechanical throttle). You can only go in 0.5% increments in the throttle target table. you also want to turn off idle ignition control for now as it will confuse things, and make sure ethrottle idle is set to open loop. Now go to your idle base table that you zero'd out earlier, and set the value for the warmed up temp (80 deg C ish) to 0.3 or 0.5 or whatever gets you the idle you want when warm. set this same number for every temperature above 80 as well. Now add 0.5 for each 10 deg cooler than 80. so if your 80 deg idle number was 0.2, 70 would be 0.7, 60 would be 1.2 etc. This should put you in the ballpark but probably slightly high while warming up. Now turn it off and leave it overnight to cool. Once really cooled down connect a laptop, make it so you can see the rpm, target rpm, ECT, and the idle base table. Start the engine and wait ~5 seconds for post start enrich to drop off, then look at which temp you are at in the base idle table. wait until you warm up enough to be dead center of a cell then quickly change it up or down until the idle RPM is correct. Wait a minute or so until your temp is dead center of the next cell, then adjust that cell so idle matches what you want. Just sit there watching it warm up and adjusting the cells as you hit them. Once its fully warmed up, you can look at tuning ignition idle to make it a bit more stable. I wouldnt turn on Closed loop ethrottle idle though until the next firmware as there are currently some bugs in it where it will add in 0.4% throttle angle every time you get on the gas pedal at idle and can work its way up to being too high. You now have a pretty well tuned throttle angle for warm up. extrapolate out from your final values to anything colder than you actually tuned to complete the idle base table. Remember, if you change the ign angle at idle you may have to re-tune this.
  47. 2 points
    Another way to work around the missing inputs on the Thunder (or to divide an input frequency down on other ECUs) is to use a CMOS 4018 IC. No external components required and can be powered from the ECUs +5 Volt supply. To divide by 10, the inverted Q5 output is fed back into the data input pin. I used this last week to divide by 8 on a fury (I fed back the inverted Q4 pin to data) Hope this helps, Richard.
  48. 2 points
    Hi, Currently with mixture map you set a threshold so that samples within say 25% of the centre of a cell vertically and horizontally. This pool of results are used to contribute towards an average value in the centre of the closest cell. However this means that you've got 25% variation of rpm and load, contributing to a static value in the centre - and you need to throw away 75% (?) of recorded values. I have another idea that can let you use all of the data instead, and improve the results. For simplicity's sake imagine a 4x4 grid, and our current load and rpm point is 25% of the way towards the lower RPM value and 25% of the way towards the lower Load value. If we interpolate these values, as per what the ECU does. Note: I have just titled the columns and rows with percentages to show what percentage of the each cell we are interpolating from. We get a value of (25% * 25* 10) + ( 25% * 75% * 30) + (25% * 75% * 20) + (75% * 75 * 40) = 0.625 + 6.075 + 3.75 + 22.5 = 32.95 is the table value that interpolation produces. Now lets say that you wanted to add 10% to this value. If we just adjust the closest cell by 10%, as per current Mixture Map strategy. Then our bottom left cell changes to 44 so our table now looks like this: If we do the interpolation again, but with the new value to represent running the car again after the update: We get a value of (25% * 25* 10) + ( 25% * 75% * 30) + (25% * 75% * 20) + (75% * 75 * 44) = 0.625 + 6.075 + 3.75 + 24.75 = 35.2 as the new overall value. Which is only makes 6.8% difference to the interpolated value, rather than the 10% we wanted. On the other hand... If PCLink De-interpolated the 10% that it wants to add. Instead of adding 10% to the one cell, we split the 10% addition across the 4 cells based on the same percentage that the value was interpolated from initially. So: Top left cell: (10 * 1.1 * .25 * .25) = 0.6875 Bottom left cell: (30 * 1.1 * .25 * .75) = 6.1875 Top Right Cell: (20 * 1.1 * .25 * .75) = 4.125 Bottom Right Cell: (40 * 1.1 * .75 * .75) = 24.75 = 35.75 is the table value that de-interpolation produces. We were trying to add 10% and this new value produced is 10.5%. So that's pretty good! (The 0.5% error comes from rounding to 3 decimal places in my example) So it's accurate to the provided data in every instance. Which is especially relevant when it's applied 1000s of times across all of the cells. You dont need to throw away any of your recorded data, it all contributes to the cell values. Mixture map is pretty good for roughing out a map initially but because of the inaccuracies of the "nearest cell" method I don't really use it that much anymore when trying to dial in a fuel map. You always overshoot or undershoot unless you set your cell tolerances impossibly tight and have millions of samples. And, since this is all only done in PCLINK rather than the ECU, there's not really any worry about the overheads of the extra maths involved. It's worth having it chug away for a few minutes longer if you can get an awesome result on first or second iteration of Mixture map logging. So - that's my Friday night suggestion. Thanks for reading if you got this far, haha.
  49. 2 points
    mapper

    Closed Loop fuel trims

    Regarding Lambda control error correction table. I spend alot of time to tune these. The base map is adjusted the wrong way around. Because the error correction tables is a % corretion of actual error, you want big corrections like 15% on small errors (0.03 lambda error) and small correction (like 5%) at the biggest error on the table. This is because a fuel film built up first in the ports when big correction are applied. This means it needs several burn cycles to get the whole change applied and measured. This means lambda control applies big changes two or three times for big corrections which leads to Lambda oscillation. On small changes fuel film built up is much less. Lambda change is done and measured much faster and within same burn cycle. This means the Lambda correction can be set much higher, because the change in AFR is measured instant. I have attached a tuned example.
  50. 2 points
    Okay so there are a few stages to setting it up. 1. Wiring Run one wire to the knock1 or knock2 wire on the link loom, and one to sensor earth, polarity unimportant. must must must must use shielded wire. The knock sensor outputs a very low voltage signal that's prone to interference. 2. Initial settings Since you are using the 'wideband' knock sensor and an engine with an ~86mm bore has a knock frequency in the ~6khz range select your Freq Channel as 4-10khz Wide Band. Set Ignition Retard limit to 0 degrees. Set the RPM high and low lockouts however you like. (500rpm likely not ideal for the low setting) 3. Cylinder balancing Your knock sensor is mounted closer to one cylinder than the others. It picks up vibrations, so the vibrations from that one cylinder will give a stronger signal than the others. So what you need to do, is hold the motor at say 4000rpm (no load) and check the signal strength of each cylinder. You can check the signal strength by pressing F12 to get to the runtime values screen and looking at these numbers, knock level cyl 1/2/34 See how in that example above, the numbers are 235 / 160 / 255 /145. You need to get these numbers as balanced / equal as possible. You can adjust the values up or down by tweaking the numbers up and down in Knock control > Cyl setup > Cyl 1/2/34 knk level gain Best to start with a value of 1 for the cylinder that's closest to the knock sensor, and increase the other values to suit. If one of the values reaches '2' (maximum) you can reduce some of the other numbers to less than 1. 4. Non knock noise levels Since the knock sensor picks up vibrations, there are of course vibrations happening even when there's no knock. As RPM increases, the amount of 'natural' background noise increases too. The ECU can tell that knock is happening, because there's an unexpected large spike in the 'noise' from the motor around the time of the iginition event. Soooo, you need to find out what the background noise level is for your engine. According to the manual, a 2 row table with full throttle and 0 throttle is sufficient but this is up to you and how long you want to spend on it haha. So head to Knock control > Knock target, right click on the table and select Axis setup to define your table similar to this (if you want) Then you need to run a datalog through the rpm range at full throttle to see what the values are for this table. (and coast back down off throttle for the zero TP target, although I'm guessing not much knock happens at 0% throttle) Open the datalog and bring up a screen to show engine rpm and the knock level global. Knock level global has a maximum value of '1000'. If you find that you are hitting 1000, you need to reduce the Gain Channel number on the main knock sensing setup page to something a bit lower and try again. Remember that the '1000' has to be the maximum even including allowance for knock which is much stronger signal than the background noise so you need to allow headroom for that too. Once you've established these background noise levels for the motor in your table, increase all of the numbers in the table by 20% to give it a bit of a margin against picking up normal engine noise as knock. At this point, because you've set the maximum ignition retard to 0 degrees in your first step, the ECU isnt taking any action against knock. Now that you've got everything setup though (unless I've missed a step here, haha) you can turn the knock sensing on by setting an ignition retard limit here, to say 3 degrees or 5 degrees or whatever you want: Then as per reccomendations from the manual, it's best to test that knock sensing is working under a scenario that minimises risk of damage to your engine. So you could drive along at low load / low rpm and induce knock by creeping the timing forward until it knocks and you can see from the runtime values table (F12) that it's working. From here, it should all be working awesomely. (No responsibility taken for blown up motor though! This is just what has worked for me) Hopefully it all makes sense though Where are you based / what is the car used for? Keen to hear how you get on.
×
×
  • Create New...