Facking hell guys 
I thought the same!
Pretty basic stuff really (except for me, which is why I started the thread).
As ever, the community spirit shows itself when not in a moaning/cracking/hacking/"post your last..." thread. Many many thanks.
Racing on Sunday and Monday at Mallory Park. If anyone wants to come, feel free to do so.
As ever, the community spirit shows itself when not in a moaning/cracking/hacking/"post your last..." thread. Many many thanks.
Racing on Sunday and Monday at Mallory Park. If anyone wants to come, feel free to do so.
Good stuff. I did end up going out last night so couldn't get my little app polished off. You've saved me a bit of work so I can now watch FP from Spa in peace
Good luck.
Question, does everyone else put this much "theory" into racing at this level?
I'll be honest, I assumed it was a load of well-trodden upgrades limited by budget and then the odd bit of old school mechanics...!
I'll be honest, I assumed it was a load of well-trodden upgrades limited by budget and then the odd bit of old school mechanics...!
No, I don't believe most people do put this much level of theory into it - if I speak to most people about gear change points, ratio theory, engine mapping, tyre load sensitivity or anything to do with dynamics, then they really seem to struggle to comprehend what I'm on about.
They finish a race and start drinking. I finish a race and start looking at data.
However, I only trust the theory if the practical side agrees, and I only trust the practical way if the theory agrees. That way I don't end up making too many guesses!!!
I think I do a fairly professional job considering the level of racing and our very limited budget. And I'm still learning. In 2007 I barely knew a slick tyre from a gear ratio, and now I have people knocking on my door wanting to know how to go quicker. Sadly, my driving is quickly becoming the weak point in the team effort.
They finish a race and start drinking. I finish a race and start looking at data.
However, I only trust the theory if the practical side agrees, and I only trust the practical way if the theory agrees. That way I don't end up making too many guesses!!!
I think I do a fairly professional job considering the level of racing and our very limited budget. And I'm still learning. In 2007 I barely knew a slick tyre from a gear ratio, and now I have people knocking on my door wanting to know how to go quicker. Sadly, my driving is quickly becoming the weak point in the team effort.
From what I've seen of your finishing record, it obviously works 
I'll be billing Tristan Cliffe racing for the flight.. no?
I am quite surprised you have so poor mix when floored on low revs - I know it is a single seater with big power/mass ratio but
Doesn't every vehicle do that on low revs with aggressive throttle work?
Given it's a car designed for track, most of the time he isn't in that range of revs anyway.
Most of his worries are in mid/high revs and low/mid throttle position.
I'm no expert on this, just saying what seems logical to me, so I could be very wrong.
Given it's a car designed for track, most of the time he isn't in that range of revs anyway.
Most of his worries are in mid/high revs and low/mid throttle position.
I'm no expert on this, just saying what seems logical to me, so I could be very wrong.
Yup - there just isn't much data for low revs, high throttle - and then it's going lean. High AFR = lean fuelling.
I think I improved it a bit at Mallory, but it takes time, and I've got to gradually make improvements rather than try to cure it all at once and makes things worse for a race...
I think I improved it a bit at Mallory, but it takes time, and I've got to gradually make improvements rather than try to cure it all at once and makes things worse for a race...
theres hardly any data at any point
not sure how the others have created their surfaces but theres so little data anywhere in the middle that you can tell exactly nothing from it
not sure how the others have created their surfaces but theres so little data anywhere in the middle that you can tell exactly nothing from it
Indeed - I think the sample data I posted is just one laps worth. But when I have 15 or 20 laps worth of data, some more meaningful results appear - although admittedly there is less at partial throttle than full throttle because it is a racing car, and not much time is spent at partial throttle.
Do you have a programmable engine management system? Could you export the fueling table and have a look what it's supposed to be doing vs. what the datalogger gives you?
That's what I'm doing. I can see/edit the fuelling map (and the ignition timing, and other paramters), and I wanted an easy way to visualise the datalogger data - this excel thingy does that. Sure, I still have to interpret the data, and take into account when little or no data is available to make a valid judgement, but it's better than nothing.
I see. The system I'm using doesn't rely on an external datalogger, but logs all engine parameters itself. This makes graphing the data pretty simple without the hassle you're going through now.
I'm jealous!
I've thought about putting the lambda sensor into the ECU so it can automatically compensate the fuelling to give an AFR of 12.8 (or Lambda 0.87 if you prefer), but I don't think the ECU 'learns' - i.e. it doesn't alter the map and save the changes, it just modifies the value in the map temporarily. If ever we stop using a lambda sensor, or it breaks, or it wears out, then I would want the 'good' map to remain, and if it doesn't learn (i.e. save an changes it makes from the feedback loop) then that's no good.
I wish I'd mapped the engine myself now. I'd have done a much better job than the person who programmed the ECU - he's a computer geek/programmer first and foremost, not an engine builder, an engine mapper, or a racer. He's ex Lotus, so knows how to make ECU software for road cars, which is quite different, I'm sure you'll agree, to building an injection map for an F3 car...
I've thought about putting the lambda sensor into the ECU so it can automatically compensate the fuelling to give an AFR of 12.8 (or Lambda 0.87 if you prefer), but I don't think the ECU 'learns' - i.e. it doesn't alter the map and save the changes, it just modifies the value in the map temporarily. If ever we stop using a lambda sensor, or it breaks, or it wears out, then I would want the 'good' map to remain, and if it doesn't learn (i.e. save an changes it makes from the feedback loop) then that's no good.
I wish I'd mapped the engine myself now. I'd have done a much better job than the person who programmed the ECU - he's a computer geek/programmer first and foremost, not an engine builder, an engine mapper, or a racer. He's ex Lotus, so knows how to make ECU software for road cars, which is quite different, I'm sure you'll agree, to building an injection map for an F3 car...
No lambda signal going to the ECU, so it runs of a preprogrammed map all the time? I've set mine up with a wideband O2 sensor to provide realtime feedback and the ECU adjusts accordingly if necessary. I do have a fuel table to run of but closed loop lambda feedback is nice to have. There's also an AFR target table with engine load vs. RPM so there's an AFR target for about 95% of the engine's load/RPM range to shoot for. It will adjust to that able, but only so much.
Here's an example of how my current target table looks like, in 3D available too.
MAP is manifold pressure (or vacuum really) but I'm sure you know that
Here's an example of how my current target table looks like, in 3D available too.
MAP is manifold pressure (or vacuum really) but I'm sure you know that
Yes, ours just runs the programmed map. We weren't originally going to run a lambda sensor, but after having misfire problems we needed one to know if it was too lean or too rich. So it's just logged, but doesn't influence the ECU directly.
I'd run it in closed loop mode if the ECU could 'remember' what injector timings it actually used. But as it only does the live adjustment (which I'm sure is great; but I don't want to keep the lambda sensor on the car all the time) I'm less interested...
I need to speak to the ECU people - maybe they can help. But they never answer emails and the person I need to talk to is never there when I phone... DAMN THEM ALL!
Edit: You run quite lean at 13.5... For best power I'd have thought closer to 13 would be better. And why so rich on the overrun (small throttle/high rpm)?
I'd run it in closed loop mode if the ECU could 'remember' what injector timings it actually used. But as it only does the live adjustment (which I'm sure is great; but I don't want to keep the lambda sensor on the car all the time) I'm less interested...
I need to speak to the ECU people - maybe they can help. But they never answer emails and the person I need to talk to is never there when I phone... DAMN THEM ALL!
Edit: You run quite lean at 13.5... For best power I'd have thought closer to 13 would be better. And why so rich on the overrun (small throttle/high rpm)?
My motor seemed to make the most power at 13.6 AFR. I added .1 for a little margin though. It runs at 13.3 now actually. Overrun target is set to about 16:1. Above 1500 rpm, high vacuum and throttle below 5% (or something like that, practically closed) it cuts fuel completely.
I'm interested in your ignition map, could you post it up just for giggles?
I'm interested in your ignition map, could you post it up just for giggles?
Attached (with a few other snippets tucked away).
As always, comment if you see something wrong or strange.
The 'hole' in the ignition map is to quieten the car during noise testing, which takes place in that RPM range with that throttle opening. Seems to work, without impacting on track performance.
As always, comment if you see something wrong or strange.
The 'hole' in the ignition map is to quieten the car during noise testing, which takes place in that RPM range with that throttle opening. Seems to work, without impacting on track performance.
Interesting. The TPS scale looks funky to me. Is it % throttle opening or converted ADC values?
You seem to have a lot of resolution on low throttle openings, and then only 3 rows between 50% throttle and WOT?
Could you also post the file itself? Or is the software you're using to view/change not free for use?
You seem to have a lot of resolution on low throttle openings, and then only 3 rows between 50% throttle and WOT?
Could you also post the file itself? Or is the software you're using to view/change not free for use?
The throttle opening is %
No idea why there is such resolution at low throttle openings - I've just asked the company that makes it if I can change that. I'll happily interpolate each value manually to build the same map but with better throttle resolutions. I'm guessing it's because the ECU was originally intended for road cars where smaller throttle openings are more important than large ones. But that's a guess.
I can indeed. But not yet - it's on my laptop. Maybe tonight I can zip up the map, the software and the manual and upload it to my site for you to download.
No idea why there is such resolution at low throttle openings - I've just asked the company that makes it if I can change that. I'll happily interpolate each value manually to build the same map but with better throttle resolutions. I'm guessing it's because the ECU was originally intended for road cars where smaller throttle openings are more important than large ones. But that's a guess.
I can indeed. But not yet - it's on my laptop. Maybe tonight I can zip up the map, the software and the manual and upload it to my site for you to download.
Sure thing, shoot me pm when it's up. I'll gladly have a look at it.
so you can either ignore it since it hardly matters or do test runs to produce some valid data and fill in the gaps
Excel Genius Required
(65 posts, started )
FGED GREDG RDFGDR GSFDG