RyanG

Exactly what I was looking for, thanks! I was just researching low pass RC filters then. I also should've mentioned I had tried various DI freq filter values, I'd just happened to leave it on "none" after giving up. Looks like in the latest update the max has increased from 7 to 10. Regardless, it didn't help the huge noise spikes, but it did significantly reduce the amount of averaging required so I'll take another look at that once I wire in a filter. At max speed of ~250km/h my frequency should be 180hz, so I was looking at an RC filter to suit. I downloaded your log from the post and your frequency values aren't too dissimilar, so I'll replicate what you've got and see how it goes. Hopefully I can sort it out as the car desperately needs boost by gear now haha.

Hi, Trying to sort out a reliable VSS source on my 1988 Honda Civic ED which has a B18C turbo swap. It uses a cable VSS drive as the speedometer is cable driven. There is an electronic variant dash in a JDM model, but unfortunately they don't quite fit my pre-facelift surround. I do have an electronic VSS which I tried to test, but had no luck getting that to read. Ultimately I need a solution which leaves my factory cable driven speedo working so didn't spend too much time on it. I originally tapped into the VSS output from the reed switch in the back of the dash. With appropriate filtering or averaging, it would read great until 80+km/h. After that it would get large noise spikes, meaning I couldn't use it to calibrate gear position for boost by gear. I pulled the dash out and replaced the reed switch, as well as tested it using a drill. Unfortunately, still the same issue. Does anyone have any advice for things I can try? I'm going to modify the reed switch mount and experiment with adjusting the distance between the switch and the magnet. I'm not very electronically experienced, but have seen things like Zener diodes or low pass filters being suggested - anything like that worth trying? I can also look at rewiring the switch to go to an ecu signal ground, if that would help? Or is something like this just never going to work? My last resort will be to wire in aftermarket hall effect wheel speed sensors which is doable, but would require some custom work. I've seen some pass through cable VSS boxes which have a hall effect sensor, but they're not cheap and would require work to get them to fit with my VSS and cable. I've attached some images, tune file and 2 logs that were taken directly after each other. The first the VSS was working fine, then on the way home it had noise spikes (and reads 6000+km/h). I'd think these large spikes could be filtered out with a median filter, but I don't think that's possible within the Link itself. https://drive.google.com/drive/folders/1r0GfJJ8OPbbY3yK17nv4TpFiGwVZ8N20?usp=sharing

Just to close this one out, ended up changing the distributor coil to a known OEM one and issue is fixed.

Given only a single value is able to be specified for injector flow (per fuel type), I assume there is a background correction for differing fuel pressure from rated pressure? How well does this work across completely different setups? Is it able to maintain constant AFR during mild fuel surge? Or minimise required VE changes if base fuel pressure is increased, but flow at rated pressure left the same? I don't currently have an adjustable pressure reg, but need to get one shortly as I'm up around 90% DC on 1000cc and don't really want to upgrade so looking to just up the base pressure for head room. I was tuning a friends car (different ecu) and it had a 3D user editable table for injector flow. Setting up inj flow against both differential pressure and fuel composition % seemed neat. I could see it helping maintain consistent AFR across fuel comp with a single VE table if required on lower end ecu's. Perhaps it's unnecessary if Link's background calculation "just works"? At least a second value is able to be set for the multi fuel. I believe other platforms have a single value then a correction factor.

I had the same issue when grabbing the speed signal from the back of my dash (cable driven driven). Worked okay, but quite noisy and random 400kph spikes from memory. Filter didn't really fix it. Can you share how your average math channel works? I had put this issue to the side until I got time to actually fixed what I assume is a hardware issue in the dash (30yo electronics).

Then what's the purpose of the current interpolate function? In what situation would it be preferable for it to interpolate, but not consider the breakpoints and create a jagged line in a map? If you've tuned 50kpa and 80kpa, and want to interpolate because you think it will be close to linear between those regions, why should the interpolation be non-linear between those points if you have uneven axis breakpoints? Yes this has the same effect as deleting the row/column, but it's just to get it in the ballpark until you tune those sites directly. I guess I'm the opposite, I don't understand the point of it working the way it does now. It is probably useful for Alpha-N tuning like Electredge mentioned, I haven't got much experience with TPS as load. In saying that, every other ECU software I've looked at uses standard linear interpolation. An option for both would be neat, or an option in the settings to choose "axis scaled interpolation - yes/no".

Any thoughts on this one @Adamw @Vaughan?

Would definitely be useful. Scatter plots aren't particularly helpful without filtering. I posted the same here:

I think absolute values for the transient lockouts would be very helpful. Seperate minimum lockouts would be even better.

No, I want to interpolate between cells on the fuel/ignition table, but I want it to do it linearly and with respect to the break-points in the axis between them. I wish I could post images, but I have no kB left. With a MoTeC M1 if you interpolate between 50kpa and 80kpa, with 60kpa between: 50kpa: 50% VE 60kpa: X% VE 80kpa: 60% VE You will get X = 53.3% VE. If you do the same on Link G4X, you get 55% VE. It doesn't make any sense?

I had exactly the same problem so I know your pain! I used 3D CLL trim limit tables referencing TPS delta to try and minimise it until I got a chance to correct injector data and retune.

Would definitely be nice. I have saved layouts for whether I'm on my laptop or looking at logs at my desk with a 27" monitor, but then if I make an update I need to do it to two layouts.

I haven't used mixture map all that much, I still make changes manually by going through logs which is getting a bit time consuming. I felt it was a bit limited in its filtering, but I'm going to have another look at it this weekend using GP outputs and math blocks as filters to see if I can get good results.

Thanks guys, I've updated my formula to match! For reference; at 0.8La, 2% ST and 2% LT you should get 0.832, but I was previously getting 0.84 in this example. This error also translated into my "VE% Required Corrected" math channel. Which is: ("Corrected La" - "La Target" + 1) * "Fuel Table 1"

Just opened up my old Hondata software and tune. It's in camshaft degrees with a couple of compensations, but looks like it was around 3.3ms at 1500rpm and around 1.7ms up at 9000rpm. Haltech's basemap for a B series is just a 2D table against battery voltage, but has a lot higher numbers.. 3.2ms at 14v. Not sure if there is a minimum spark duration setting which would be bringing that value down in the top end.

I haven't touched dwell settings at all, I've got no experience with that. I'll do some reading and have a play to see if any improvements can be made. Yeah the signal is quite noisy due to only having a cam mounted sensor. The plan is to go a crank trigger and COP kit down the track. The limiters at 8k rpm are correct, it's just the last couple of drives I've rarely hit them without 'hitting a wall' at 7.5k or thereabouts. While troubleshooting initial trigger issues and the car not starting over a year ago, Adam helped me out trying to get it to work. It gets noise which was causing a false trig2 signal from memory. I needed to up the trig filter otherwise it has issues. I cut a tooth off the 12 or 24 tooth (can't remember) sensor to run it as a -1 setup so trig2 isn't used at all. I'll go for a drive after work and try fresh plugs at 18 thou and try sneaking in some more dwell.

Hi all, Having some issues with what feels like an ignition rev limit at 7.5k on occasion, but I can't see it triggering any limits on the log. If anyone else has a chance to take a quick look it would be much appreciated. When initially turning up the boost it had some minor ignition breakup, but was fixed after gapping the plugs down to 20 thou. It feels far too much like a rev limit to be ignition breakup, but not seeing any indication in the log to show its software/tune related so I'm assuming it's engine related. Car is a Honda civic with a B18 turbo and still uses a factory distributor. Definitely plan to go to COP soon however. Log file: https://drive.google.com/file/d/1eHA1MwVbkOm0eNCPTh9MeXSIolICWAvb/view?usp=sharing Log file: https://drive.google.com/file/d/18NBftpL6wPd4QCnljgXy8lTVY2Ayc7dt/view?usp=sharing Tune: https://drive.google.com/file/d/1Jm2KlLEpLDhb1HuVFRMIPLwFOeZ2O1Xv/view?usp=sharing Cheers, Ryan

Seems to. That's the whole point of using Lambda right? Percent change in fuel results in a equal percent change in Lambda (if injector data is correct). Unless there's something else I'm missing?

a = Lambda 1 b = LTFT c = STFT a+((b/100)+(c/100)) Obviously need to be mindful that it wont be accurate during transients due to transport delay.

Would be cool to have a little more flexibility for sure. Anything which has the ability to be enriched in open loop should really have the ability to also adjust the target lambda so that closed loop doesn't just over-ride it. Overlay tables can get you part of the way there, but you quickly run out of axis' and need to carefully consider the compounding effect like Mapper mentioned.

Any chance the interpolation feature could take into account the axis breakpoint so that it's linear? Given the lower table resolution compared to some other ecu's (arguments to whether needed or not, I'm sure) often axis breakpoints aren't evenly spaced. I've attached an exaggerated example. Please let me know of any downsides with what I've suggested? Ryan

Awesome! I eagerly await the next update. As you may remember I'm on unreleased firmware due to a couple of issues I was experiencing, be nice to not have that error ID pop up when I connect (and every time the ecu/laptop drops out).

Yep, that's basically all I'm doing with a math block (and then using that as the lambda source). Hadn't thought of that actually. Will have a play and see what I can come up with, thanks. Also I agree with Mapper, a reactivation delay for transient filters would be awesome. Perhaps CLL status as a conditional filter is sufficient to achieve this.

Thanks Adam, I'll try those values and recommendations this week. If they do only flow ~800cc on E85, that probably explains why I got to ~90% Inj DC quicker than I expected.

Yeah I've purchased an ethanol sensor and will be fitting shortly. I'm aware of the ability to adjust values between each fuels and I've done this, just separately in different maps and only at straight 98 or straight E85. I haven't really gone down the path of flex fuel yet because it just adds another complexity that's not needed until I fully sort it on a single fuel. Yes manufacturer data is where I've started and tweaked from there as stated. It's the how to go about "tweaking" it I'm really interested in..