Jump to content

Davidv

Members
  • Posts

    323
  • Joined

  • Last visited

  • Days Won

    32

Everything posted by Davidv

  1. I've been in two minds about this, currently with a MAF based tune on my engine (NA Toyota) I've got TPS as the load axis for VVTI. But I think with a MAP based tune it makes more sense to have it on same load axis as your fuel, because unlike MAF, MAP has no way to tell how the VE of the engine changes with cam angle. But the catch 22 here is that changing the cam angle can change the map reading at part throttle so it's a bit of a circular reference. Realistically I doubt you'll have much trouble using either.
  2. 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.
  3. Just been having this same issue. I have been using a disgustingly slow netbook for ages, Win7 with Pclink running awesome. Shiny new Win10 laptop and its as good as unusable. Just onboard Intel graphics card. Will see if there's an older driver.
  4. Yep that was my original theory too which is what instigated the testing with myself and Mike. I could end up pulling out a massive amount of fuel in idle conditions and car ran quieter and smoother. And that was just with petrol then Mike tested with E85 to good results as well. I'd like to have a play with E85 but my fuel system would shit a brick haha. Ive found this works awesomely to make my car quieter and a lot nicer when cold as well, but not sure how well it translates to a map based setup (that load axis only available with maf) There's no science to the numbers though was just a first guess to see how it pans out, and was an improvement.
  5. I have found some extra ign timing when cold helps too. Overlay ign table with ect and load as axes. Thats with petrol only but would assume similar for e85. Notice any difference with that for you, Mike?
  6. Thats awesome! Is the head reversed on that? Am I right in thinking its a gen 3 3SGE based engine. Pity it's not a beams or you could use my same roller setup. And re: fuel pooling, a friend said that when he tried it (with inboard injectors only) you'd just get fuel vapour/mist partially sucked into that inner part of the barrel because it holds a partial vacuum. Then it's got nowhere to go because of the barrel angle. So then when you go WOT it all dumps into the intake. Not sure how much of an issue it really is though.
  7. Oh my - Your brother has a BTCC car? Details please! And yes that 4AGE / starlet is the business! Also, with regard to linearity of throttle actuation. I ran this through the flow bench calculator thingy at different throttle opening angles, to measure the restriction in 5 degree iterations. Doesnt look nearly as bad as I thought, but this is hoofing an absolutely tremendous amount of air through a runner in steady state. But I'm thinking part of the reason people talk about drivability issues at part throttle could be on account of the roller barrel being so close to the head, and negatively impacting air/fuel mixing & airspeed. It's also possible that when the airflow demand is much smaller it will reach peak required amount of airflow at a very low throttle angle. Something else that I've noticed about the Titan rollers, they actually have a trimmed out part on the front face of the roller, so the inside of the barrel doesnt fill with partial vacuum. At idle conditions there's a 6mm gap between the front roller edge and the housing, while the back is shut. That probably solves the fuel pooling issues and weird issues that arise from having two "cut off points" in the flow path at low throttle angles.
  8. Hi, I have been giving some thought to making some Roller Barrel throttle bodies. So firstly if you havent seen them, the idea is that they are a rotating tube (barrel) with a hole in it. Rather than a butterfly valve which always has something in the flow path, even at full throttle. There are only a few cars that run these, or have run these - For a few good reasons. But the BTCC Vauxhall 2 litre engine ran these with a cool looking set. And made somewhere in the order of 320hp NA: https://racecarsdirect.com/Advert/Details/82002/swindon-rollerbarrel-throttle-body-kit-vauxha One of the Caterhams runs roller barrel throttles from factory, with a set made by Titan/Cosworth. Apart from that, and a few references to them previously being used in F1 cars at some point, there seems to be very little information about them. So the benefits: 1. Gimmick factor 2. Low restriction 3. Can keep the entire intake length, the same profile as the port easily. Rather than needing to go to a round shape to suit a throttle plate 4. Very compact 5. Very easy to fabricate compared to butterflies 6. Apparently generates very little vacuum at part throttle The downsides: 1. They have a tendency to jam 2. Not very resilient to fuel sludge residue or small particles coming into the barrel / housing clearance. 3. Apparently the opening rate is very very very non linear. Like, quarter throttle feels like full throttle. 4. Apparently generates very little vacuum at part throttle (Yes this is a benefit and a downside) 5. If the roller is very close to the head you can have poor quality of airflow/airspeed and so poor air/fuel mixing. 6. The inner part of the barrel pulls some amount of vaccum, and you can have fuel pooling issues inside the barrel with extended periods of part throttle. So here's why I'm going to make some anyway even though they are pretty much unreliable garbage: 1. Gimmick factor 2. They are very light 3. Learn how to make something new 4. I can use e-throttle to overcome the traditional disadvantage of a non linear opening rate 5. Gimmick factor 6. Possibly slightly higher horsepower potential 7. Gimmick factor So the first thing was to draw up a model. Pretty blatantly copying the Vauxhal and Cosworth design here. Two cylinders per barrel. But with dimensions to suit the ports of my engine. No point in reinventing the wheel. (Well, lets ignore the fact that roller barrels are shitty so I'm essentially reinventing a triangle shaped wheel) Then made some molds for the trumpets, for making some molds for the trumpets, to make the trumpets. Without the return edge on the bellmouths for starters while I try figure out how to make molds and use carbon fiber and so on. Gotta love 3D printing. Then after about 500 iterations and failures, making one that turned out alright. Not quite there but learning something from each iteration. Nearly good enough! I might add some draft angle / taper to the length of it and mmaaaayyybbeeee think about the allowance of a secondary fuel rail at some point. Seeing how I've got enough injector drivers and so on for staged injection with the Xtreme. I've 3D printed a test pair of the throttles, that I've test fitted to a dummy engine. But need to pull my car's actual intake off at some point and test for clearance to see how long the trumpets can be. I'm anticipating that they will need a curve in them for bonnet clearance, for the length I'd want. But that's not an issue. From here I need to finalise a few bits such as, decide what sort of bearings to use to support the rollers. Decide how to fit return springs to the roller assemblies. Decide how to balance the two halves together. Decide how to actuate the throttle, either cable pull for starters or e-throttle. The outer housings themselves though, and the flange to the engine are as good as finalized though. So just need to find a few chunks of alloy and start machining! Or possibly get the housings cast instead. I am anticipating that MAP readings will be next to useless. So it will likely be Alpha N based tune, or depending on how the airbox and intake arrangement works out. Possibly MAF based so long as the pipe isnt too big. Thinking that 4" pipe might be about the limit before you dont get much resolution from a MAF. But could perhaps use an Alpha N / MAF cross over. Will see how it goes. First thing is to start/finish making them haha. Dont hold your breath for updates this is going to take me a few months of mucking around to get sorted from here. But a brain dump of what I've been doing so far will help keep me motivated! And some peer pressure haha.
  9. Yeah mine is really snappy. If you look at the logs of requested throttle angle vs actual, it's very fast. Not instantaneous obviously but its not half a second delay or whatever that OEMs feel like. In my opinion the OEM mushy pedal is likely for sake of reducing the amount of harsh transient conditions where fuel is least best controlled. For emission / economy reasons.
  10. E-throttle is just awesomely awesomely awesome. Firstly - Cruise control Secondly - Much easier to set a non linear opening rate to give you more control over the engine thirdly - Context sensitive pedal control. Make the pedal less sensitive when it's raining, or give the engine less total throttle opening at low rpm, or whatever you need to do to make the car the most drivable in those conditions. Heard a really good example about a guy who raced in a series with a breakout laptime. He would tune the max throttle opening down, so that he would drive full throttle each lap and stay just over that time. But then if he was in traffic he could switch it off, zing past someone then resume the previous throttle mode. In my opinion (for combined road/track car that I use) even just having the cruise control makes it worth while. EDIT: Using a G4+ though so I'm not sure which of the above things are available on G4.
  11. In that case a helmholtz resonator wont really help. You would end up with some rpms being dampened, and some with the noise increased. Also worth noting - Apparently the AMR500 is the little brother is the little brother of the Toyota SC12 and SC14 superchargers. With AMR500 apparently being the TX10 listed below: So you might find that the clutched pulley from an SC12 or SC14 will fit to your setup? Then you can just have the supercharger turn itself off when not at full throttle or whatever. Also - I assume you're running V belts? If you've got Gilmer drive belts, that could be where the noise is coming from. Not the supercharger itself. And it's also a really small supercharger, do you know what your pulley ratio is? You might be overspinning it in order to make boost... Even with an SC12 or SC14 people dont make that much power (like 220ish hp?) before it just starts becoming a heat pump.
  12. If you're based in Auckland flick me a PM if you like as I've got some good recording gear and so on. And this sounds like a fun challenge haha.
  13. A Helmholtz resonator only really works at one frequency, which in your case, means 1 particular rpm range. The first thing to do though is quantify the frequency that you are trying to cancel out. So record the noise as best you can, isolated from other noises. So maybe try hold the engine at the rpm where your car normally cruises at. Then save it as a .wav file and load it into this program http://www.artalabs.hr/download.htm Then you can load one of the analyzers which will show you what the main frequency of the noise is. Then you can calculate dimensions for a helmholtz resonator that will be able to cancel out this frequency. An easy way might be to get one 3d printed so you can experiment. Generally the higher the frequency the smaller the resonator needs to be.
  14. Hahahaha no trust me, once you're used to it you'll not want to go back. It's a lot better than tuner studio. And as mentioned the fuel equation in megasquirt is very simplistic by comparison. And just take a browse through the help file, possibly the best help file I have ever seen for any product. I will say this though - I still use Megalog viewer sometimes because PClink doesnt currently support custom/maths channels which is a pity. It's the only reason I stray outside of the Link system though, and only occasionally.
  15. Here's a more detailed post about how I tested deadtimes: https://www.hpacademy.com/forum/general-tuning-discussion/show/calculating-deadtimes-with-a-link-g4-and-injector-test-bench I made an extension lead for injector 1 so I could use the ECU in my car, the fuel pump thats in my car, the FPR that's in my car, etc etc so every variable is as close to actual use as possible. Then I just made the ECU vary supplied voltage by using a battery charger connected to battery, and used different combination of accessories on and off to get the voltage to where I wanted to test it.
  16. My arduino based canbus box thingy has escalated into a digidash haha. So 3d printing a housing that will mount up to standard points and accept the standard loom plugs for power supply and so on. Lots of work left to do, but will be cool.
  17. +1 to this, a non linear scale is especially useful for MAF.
  18. I've had a bit of a side project going on, I've wanted to learn about canbus so I bought an arduino and a canbus shield. Using some of the examples of code found on the net, managed to get frames both sending and recieving. I've never coded anything before so I had a few blunders like using = where I needed == and so on haha. I'm using another shield on top of the canbus one that runs a screen, and have put together a box for it all. The goal is to have the 4 pots control trim tables for goal AFR, ign timing, cam timing, and injector timing. So when I have cruise control turned on, I can quickly go through a lot of variations and have a graph showing fuel economy on the screen. I can also write a fuel economy value back to the ECU as well, and then have that log into the ECU's logs so that's cool. Currently my car gets around the mid 7 litres per 100km mark, and I'd like to see if I can drop it into the sixes. I'm most of the way there to finishing this thing so hopefully not too long till I can have a play around. I am thinking that for version 2.0 instead of pots I will have rotary encoders and then build an array which will act like a load vs rpm based trim table for each of the different areas. So it will hold the values in that area without affecting the others. It's a bit more complicated though so I'll just get this running as is for now. Since I run 98 octane and MBT is quite far away from knocking I have also been thinking about using a feedback loop where the arduino adds a few degrees timing, see if this positively or negatively impacts economy, and then either reverts or keeps change. Then repeats, and maybe hones in by a smaller change each time. But I think that might work best once I've got some arrays setup, rather than just blanket applying a trim table to everything.
  19. Davidv

    Can protocol sample

    @CamB you can set the byte order to MS first or LS first to swap the order when setting up incoming frames (In G4+ at least)
  20. That's really interesting Mapper, I've been meaning to have a play with that CLL gain table but havent yet. I think you've just piqued my interest however! I'll try those other numbers for a starter and see how it differs.
  21. I've got an injector timing table set to 240deg BTDC end of injection all the way through rev range. With some changes to cam timing I decided to test 400deg btdc above 2200rpm / redo some testing at different angles. And as soon as I make this change, when my car gets to that rpm where the injector timing advances a lot the car bucks and lurches and stutters. If from here, I set all of the timing to either 240 degrees or 400 degrees, its fine again. But when it's on that transition point, things get weird. It seems more prominent at low load, I'm guessing because its an issue if start of injection is beyond a certain point? I suspect that an injector event gets skipped, the car hesitates and rpm drops and same thing happens again. (It's a really aggressive lurching of the car, and definitely not just caused by the effect of different timing) If I set the table to all 400 deg or all 240 deg the car drives smooth and fine. I can supply logs and a basefile as an example if needed. Thanks David
  22. Thanks for the detatiled answer Adam! Appreciate that you took the time to test this, thanks.
  23. It doesnt seem to, but it may be that the change is so small that it isnt able to be seen with a map sensor of this scale. And possibly because my copper line is too smaller diameter as well. I've parked this idea for now, I'll revisit it later on when I can redesign the pipes a bit better. But I'm just about there with my MAF based tune, might need a test driver at some point while I tinker if you're keen!
  24. Hey guys, I've recently gone back to using a MAF sensor, and I notice that with modelled fuel there is the option to have as a load axis "Air per Cylinder Measured" Which is awesome. Since this figure comes directly from the MAF sensor I am assuming that a lot of the calcs for modelled fuel get bypassed. I also notice there's a seperate category for "MAF IAT" Rather than regular IAT, I assume this does the IAT calculation differently too. However I have a few questions. 1. Options for load axis in the fuel calculation are either MAP, BAP, or Off. I've set this to Off. Would that be best option? 2. Which of the modelled fuel settings now get ignored, since you're measuring air directly? I get the feeling that it air per cyl measured is used as part of the calc which then still generates "air per cyl estimated" which is what the fuel calc runs from? I realise that MAF isnt the suggested option but it offers some benefits while I'm doing some experimenting with cam timing, so I'd just like to get it setup to "best practice". So any advice/info on how it calculates would be appreciated. Thanks David It's pretty cool having grams/cyl as a load axis because the values are quite linear... This was the MAF basemap that I put together based on observation of my MAP based tune, and car ran awesomely on it! It's now changed shape a little after the first iteration of tweaking it but it's still very linear.
×
×
  • Create New...