Jump to content

ashesman

Members
  • Posts

    1,061
  • Joined

  • Last visited

Everything posted by ashesman

  1. Hi Andy, Thanks for taking the time to look into that. I managed to reproduce the same thing and was very surprised to find that result! I have added an item to the firmware issues list but can not give any answer as to exactly when this issue will be resolved and tested. Sorry for the inconvenience.
  2. Hi Andy, Final ignition interpolation is to 0.1 degrees. While some tables may only allow entry of 0.5 degree steps, they are interpolated between cells to 0.1 degree internal numbers that are used for final ignition calculation. If you know of any tables that are not interpolated correctly then please let us know.
  3. ashesman

    MAF usage

    As previously mentioned, MAF support is limited. The reason for this is that while MAF is a very good indication of engine air flow and hence fuel required, it requires its own complex modelling, often engine and intake system specific. The MAF reading is accurate at steady state but less accurate under transient due to manifold and intake track volumes. It is possible to get a very accurate and repeatable tune without using a MAF sensor. Many modern vehicles do not use MAF sensors any more and almost no aftermarket installations do. My opinion is that you will get a better tune from a MAP based install than a MAF install on our ECUs and most other aftermarket brands. While they may support MAF sensor just as we do, check they offer all the correct modelling and compensations to get a reasonable tune first before buying other wise you may end up switching to MAP in the long run. The use of a permanently fitted oxygen sensor will be required to get a good tune for minor variations whether using MAF or MAP so ultimately the engine will probably end up running the same AFR but at least with MAP there is good transient performance. Good injector characterisation is essential especially if you plan to use large aftermarket injectors, this can make or break having a big injector engine that can also idle and cruise nicely.
  4. OK. Here are some answers to some of the issues raised on these posts: Injection timing can be set to use a single number for all RPM/load or a 3D table. If you want to use a 2D table, then using the axis setup form (press X while on the injection timing table), set the Y axis to None. 1 - Cal Tables 7-10. According to our testing logs these work correctly but I have added your comments to have them checked again. 2 - Short PW adder table axis is set to match the pulse width numbers provided with most injector characteristics (GM style). They have reasonably high resolution so should be adequate for most application. I have added a feature request for configurable axis. 3 - The Injector deadtime table can be either a 2D table or 3D table. Select under Fuel->Fuel Setup->Injector Setup->Injector Setup 4 - Can you please give examples so we can reproduce and find this problem. 5 - I will have the help guys look into this and see what is going on. 7 - Good suggestion, added to the feature request list. 8 - You can switch between modelled and modelled-multi fuel modes at will. We recommend tuning on just modelled mode first (just fuel table 1), then switching multi fuel on to enable both tables. 9 - Unfortunately for now there is no intention to add these buttons, in fact they are to be removed from Link software as well soon in favour of more screen space. I will note that you would prefer to keep these buttons. Regards Ashley
  5. The idea of the air temperature sensor is to help determine the temperature (and hence density) of the air charge entering the engine. So, if the manifold is heat soaked then the air entering the cylinder will be significantly heated by the manifold itself and so much less dense. The sensor does not get "heat soaked". It is the manifold that gets hot and heats the air inside it. It is entirely up to you where you place the sensor. If you do not want compensation when the manifold is heating the air temperature above that before the throttle plate, then place the sensor before the throttle plate. You will however notice that all OEM installs place the sensor in the manifold. The fuel equation (both modelled and traditional with charge temp estimation turned on) adjusts the air density exactly as per the ideal gas law. There is nothing stopping you carrying on with doing things exactly as you have in the past. The traditional fuel mode can be used to tune as per older firmware. In traditional mode charge temp estimation can be turned off and the IAT table is applied as a simple percentage correction.
  6. Generally it is considered desirable to have a constant manifold pressure. A constant manifold gauge pressure would suggest less cylinder filling pressure at higher altitudes. Whereas a constant absolute pressure implies there is a constant cylinder filling pressure as altitude increases. If boost target was based on MGP: If you have 200 kPa gauge pressure boost, then at low altitude (100kPa baro) you would have 300kPa cylinder filling pressure. If you had the same boost at 70 kPa baro then you would only have 270kPa cylinder filling pressure. This is overly simplistic and ignores all other dynamic effect, but in general would cause significantly reduced power at high altitude.
  7. Here are a few things: - The CAN Id is correct 1000 (0x3E8) - The CAN Id is NOT extended - The data is sent in frames of 8 bytes. - The first byte (byte 0) indicates what data is in the frames. Â This is what is referred to as the compound Id. Â When byte 0 is zero then it indicates the frame contains RPM, MAP, MGP. Â When byte 0 is 1 the frame contains BAP, TP, Injector Duty Cycle - The multiplexer field specifies the compound Id. Â While technically it is 1 byte long, how you have it set up should be OK as byte 1 is always zero. Â Besides the extended box being checked, everything looks correct to me in your picture. - Some CAN setups change the byte numbering when you go from little to big endian. Â Eg a two byte variable at byte 0 in big endian would start at byte 1 when little endian. Are you able to receive anything? Â Or is the data just erratic?
  8. Ethrottle Signal 1 is the PWM control Ethrottle Signal 2 is the Direction control These signals together form a PWM control for an external H bridge driver. Â When Direction is low, PWM is pulsed high with short duty cycle to open the throttle a little and high duty cycle to open it further. Â When Direction is high, PWM is pulsed low for short duty cycle to close the throttle a little and high duty cycle to close it further. Â The Link ethrottle module provides the high power H bridge required to drive the motor. Â The ECUs control signals directly control this driver. Â The module also includes a power switch to disconnect power to the driver in case of faults. Â This is controlled by the ECUs ethrottle relay signal. An external driver module could be used if it can work with these signals. Â However, this is not supported or recommended and may compromise the safety systems. Regards Ashley
  9. Hi Steve, I have added an item to our todo list to change all 5 Bar calibrations to 7 Bar. The reason they say 5 Bar as that the original sensor they matched up with was sold as 5 Bar even though it could measure up to 7 Bar. I have also got someone to look into the Aux 9/10 supply warning. Someone will be in touch soon regarding testing once the FW is ready to go. Cheers Ashley
  10. Hi Steve, We are currently in the internal testing phase of firmware version 5.1.0. The next phase, due to start in the coming weeks is external testing. The release date is not set yet. This is not a "full re-write", just an update with many bug fixes and some new features. If there is a particular issue you are interested in seeing if is resolved then let us know and we can confirm if it is in this update. If you would like to help in the external testing program, then contact [email protected] and put your name down. As for the analog inputs having their own calibration each, this is on the feature request list but definitely not part of the next release. I thought we had cleared up that 5/7 bar thing. If you could let us know where it is still a problem then we will tidy it up. Regards Ashley
  11. Hi Jeremy, While this sounds like a useful suggestion, there is a slight technicality for most users: Do you know the actual delay from the exhaust port to the sensor plus the response time of the sensor and controller, plus the filtering of the analog input signal? Â Without this exact information, you would be risking making your tuning information less accurate by entering in an incorrect delay. Â Furthermore, this response time varies depending on the average exhaust velocity (ie engine RPM/load) and also on the rate of change of AFR. On a high quality wideband controller located close to the exhaust under high flow conditions, it is usually considered to have a response time that nears the typical sample rates used for AFR measurement (eg 20 to 50 Hz). Regards Ashley
  12. The problem here is that you are sending (in very quick succession) two frames with the same CAN ID. So the dash receives one with the light on, then one with it off, but the other light on etc... You need to put everything you wish to transmit on one particular CAN ID into a single frame so that the dash periodically receives only one CAN frame containing the state of all of its inputs. When trying to replicate OEM can modes, you will never (actually very rarely) have a stream with more than one frame in it. Multi-frame streams are only used by aftermarket devices such as loggers.
  13. The exact fuel equation in full detail is proprietary. However, it is not overly complex and is based on the ideal gas law http://en.wikipedia.org/wiki/Ideal_gas_law. Nearly all aftermarket ECUs use the same or a very similar fuel equation. It takes into consideration all of the measurements available to the ECU in the absence of a Mass Air Flow meter. No after market ECUs fully model the engine as OEM ECUs do as this process is extremely complex and not practical from a tuners point of view. OEMs spend millions characterising one engine exactly and mathematically (and empirically) model it in their ECUs. A good understanding of the exact maths f injector pulse width calculation is not critical for a tuner but a firm understanding of the requirements to change fuel delivered with the changes in measured variables is extremely useful. A good understanding of injector characterisation and the effects of incorrect injector information is also very important. Andre's course would be very valuable if you have not been exposed to the ideal gas law and exactly how injector pulse width can be estimated from it taking all the factors into consideration. Andre offers two courses, one on basic tuning and one on using PCLink software. The tuning one is what you want.
  14. For a fully scychronised (sequential injected) engine, the injectors are fired once per two crank revolutions (ie once per intake stroke). Â So the frequency is RPM dependent. Â Injector Frequency (Hz) = RPM / 120. The injector pulse width and duty cycle is determined by the size of the injector you use and the amount of fuel required. I would recommend using a GP PWM function on an aux output for your aquamist or looking into the ECUs Auxiliary Injection function.
  15. The ViPEC ECU uses a full speed density calculation with the exception of intake temperature. The master fuel setting takes into consideration the factors for engine displacement, injector size, equations constants and so on. The fuel table is the VE number in the equation. MAP is the pressure used in the equation. The open loop AFR target table is used as the lambda element in the equation. If you do the maths you will find that it is not necessary to completely calculate and display the calculated air mass to determine the injector pulse width exactly. The same result can be achieved. However, the next release of firmware for iXX ECUs will include a charge temperature correction option to also consider an estimated port temperature in the fuel equation. A full non linear injector characterisation will aso be included in this firmware. If you are having trouble getting stable AFR on a day to day basis, the problem may be your setup or tune and may be able to be resolved easily.
  16. You can find complete data for those injectors here: http://www.injectordynamics.com/ApplicationData.html Currently there are no intentions to implement the small pulse width adder function into the VX firmware. Sorry.
  17. The number of frames you see will depend on the number of parameters you have selected. If you have heaps selected then each time an id 1000 is due (depends on the transmit rate) a series of sequential frames will be transmitted. Also the ratio of 500 to 1000 frames will depend on the rates selected. Typically in sequential mode you will never have more than four parameters per frame otherwise multi frames will be transmitted and it is impossible to distinguish them from each other (hence why compound mode exists).
  18. In compound frame mode the sequence is: 00 xx aL aH bL bH cL cH 01 xx dL dH eL eH fL fH 02 xx gL gH ... ... Where a is the first parameter selected in the list, then b, then c etc...
  19. OK. Â Always confusing to explain, but I will try put this one to rest... There are a couple of separate things to discuss here. Â The fuel equation setting, what type of load parameter you span your fuel table with and what happens to volumetric efficiency with altitude. Volumetric Efficiency and Altitude On non turbo engines, volumetric efficiency does not change with altitude. Â Volumetric efficiency is the ratio of the volume of gas the cylinder receives per charge to the volume of the cylinder (and chamber). Â The key point is volume. Â At higher altitudes, the air is less dense, and there is less oxygen molecules. Â The volume of air consumed remains the same, but the mass (or more importantly the amount of oxygen) is less. Â Seeing as there is less oxygen, you need less fuel. The story is slightly different for turbo charged engines where the efficiency of the turbo charger changes with air density and exhaust back pressure. Â Most people do not have this information and therefore roughly assume VE is remaining constant with altitude. It may seem that the intake manifold pressure is the only thing influencing how much air the engine breaths as it is what pushes the air into the cylinder, but various combinations of running conditions can give the same manifold pressure but require a different amount of fuel due to different volumetric efficiency. Â Imagine going up through RPM at WOT on a non turbo engine, manifold pressure is always 100 kPa, but at every RPM there is a slightly different fuel requirement due to the engines volumetric efficiency changing. The fuel equation setting. Â I am not going to cover all options, just the MAP setting. Â When the fuel equation mode is set to MAP, the intake manifold absolute pressure directly influences the amount of fuel delivered. Â Ignoring all other corrections, in MAP mode, if you double the intake manifold pressure, you will get double the fuel. Â Likewise if you reduce the intake pressure (by high altitude) you will get proportionally less fuel. The final amount of fuel delivered is a function of MAP and the number in the fuel table. Â Half the number in the fuel table, half the fuel! The Fuel Table Load Axis The fuel table load axis decides what parameter is used to select the fuel table number used for a given load. Â We will restrict ourselves to just the MAP and MGP options for this conversation. Â A reminder that MGP is the difference between atmospheric (barometric) and manifold absolute pressure. Â It is 0 when the manifold pressure is the same as atmospheric, negative when the manifold is in vacuum, and positive when the manifold is at a higher pressure than atmospheric (boost). It is important to note that the parameter selected to span the load axis does not directly influence the fuel equation, rather the fuel table number it helps to select influences the fuel equation. Putting it together I will base the examples around a NA engine as it is simpler to explain. Â Remembering that VE is constant with altitude, so if we want our fuel table to represent a VE curve, then the ECU must take the same number from the fuel table for a given RPM and load (eg WOT) at sea level (eg 100 kPa) and the top of a mountain (eg 80 kPa). If we had our fuel table spanned with MAP (rows at 80, 90 and 100). Â At sea level, WOT we would be in the 100 kPa row. Â Lets say that row has 30 as the fuel number. Â Lets say we are delivering 3.000 ms pulse width. Â We close the throttle until we get 80 kPa MAP. Â The fuel delivered is reduced by the use of MAP in the fuel equation to 80% of what we had at WOT (now 2.400 ms). Â But at the same time we have throttled the motor and altered its volumetric efficiency (maybe by altering the intake dynamics). Â This means to achieve the correct AFR we need a smaller fuel table number in the 80 kPa row (eg 25). Â So the actual fuel delivered is 2.000ms. Now we drive to the top of a big hill where the atmospheric pressure is 80 kPa. Â At WOT the MAP reading will be 80 kPa. Â So, we would want 80% of the fuel required at sea level (100 kPa). Â So that would be 2.400 ms fuel. Â But as WOT puts us in the 80 kPa row and the fuel table number is 25 there we only get 2.000 ms fuel. Â This is 66% of what we had at sea level, not 80%. Â Uh oh, we are a bit lean!!! If instead we had used MGP on the fuel table load axis, WOT would be in the 0 row at sea level and at the top of the hill. Â This is great because at WOT we have the same VE at the bottom and top of the hill, so we are in the same row in the VE table at the bottom and top of the hill. Â Our number 30 would be in the fuel table at our WOT row (0 kPa MGP) and the delivered fuel at sea level would be MAP = 100 kPa, table = 30 = 3.000 ms and our delivered fuel at high altitude would be MAP = 80 kPa, table = 30 = 2.400 ms fuel. Â Exactly 80% the fuel at sea level!
  20. Hi Zack, ECU design information is strictly confidential so unfortunately we can not share it with you. Â There are no methods that I am aware of that you can improve the ECUs logging ability by altering hardware as that would involve changing the firmware itself. Regards Ashley Duncan Link Engineering Team Leader
  21. The problem here is not to do with the performance of the CPU. Â They have been tested here in excess of 20000 RPM with no problems. Â Please be more clear about what the problem is here: Does RPM read correctly? Does the trigger error counter increase when the problem happens? If so then you have issues with your crank angle sensor, the trigger setup, trigger arming voltages or something else. Please send your pcl file to [email protected] (also include a link to this forum post) and one of the technicians will test it at high RPM and try determine what is going on.
  22. That piston has deinitly been hot! Â Looks like it has been molten! Â I can only think of a couple of causes: Â It went lean and got hot or it started seizing and got hot! Â Doesn't look like damage due to detonation. Â The stuff on the spark plug is melted piston. Very unlikely the ECU decided to run a cylinder lean. Â It is also very unlikely one injector driver has been damaged. Â More likely lack of fuel due to low fuel pressure (low pump voltage) or an injector not properly opening due to low voltage (pretty unlikely). Â The answer to if it is a voltage problem is in the log files. Â If it was lean enough to do that it should of been surging and mucking around as if it was fuel pressure related all cylinders should have been suffering. Â Any chance it picked up water or air in the fuel and just happened to be feeding that cylinder a water/fuel mix? Â Pretty unlikely though. Â It points to a blocked injector but that is rare and the driver should have noticed. The other possibility is that it was always tuned a little too lean and advanced and has been running hot on that cylinder the whole time. Â Might of just had enough. Best install some safety sensors (fuel pressure etc...). Â Can you put pyros into the exhaust by going through the flange rather than the middle of the pipes?
  23. Here are the release notes for the next version of PCLink.  You will notice that there are a number of fixes around the axis setup form.  There was a major change made to the axis setup form at build 4.x.x.1112.  A lot of new code was introduced and as you inferred was obviously not well enough tested.  I am confident these issues are now resolved.  The next release will be made public very soon. Release Version 4.x.x.1194 New Features · Option to disable the startup splash screen. Changes · The default display min, max, warning, error and color values for all runtime parameters have been updated. · A warning has been added to prevent the use of incorrect ignition modes for rotary engines. · Improvement to some default color theme colors. Fixes · Cell colors did not change in tables when the number was changed. · Copy and paste did not work correctly in tables if the block being pasted would go off the edge of the table. · An error occurred when doing a Quick View log download if the location where the files were to be saved did not exist. · Could not select 'None' as an option in the axis setup form. · When clicking apply in the axis setup form, not all axis numbers would be set. · Table axis did not work if the same parameter was selected for both X and Y axis. · Some axis options had a 'Must be between 0 and 0' error. · Q and A or (Page Up and Page Down) did not work correctly on all settings. · An error occurred when selecting slip as a table axis.
  24. You can choose to use the delta lockouts if you want. They are there to prevent CLL working during transient conditions (eg acceleration enrichment). If you set them to a really low number (eg 0) they will have no effect. The example number of 75% in the help file is a mistake. It has now been corrected. To select an appropriate number, you can watch the TPS Delta number in the analog runtime values. Observe its value while slowly or quickly opening the throttle to decide an appropriate number for the lockout.
  25. Obviously there are some issues regarding the splash screen. Personally I have never seen it be a problem, but there have been several reports of it freezing. It should fade in and out in about four seconds. I have added this to the list to be improved or removed. To filin: Please give some more information about your wideband problem. Perhaps post a map so we can install it in an ECU and test using your settings.
×
×
  • Create New...