[gcosens]

I'm just finishing the last of the wiring and plumbing to convert my Mini from carb to EFI. I only have to crimp the injector connectors and fit a throttle cable and then should be ready to fire up!

I've had speeduino controlling the ignition for some time so hopefully there won't be too many issues with crank trigger etc. However, if someone has time, I'd really appreciate a sanity check on the tune just to make sure I haven't made any schoolboy errors on the fuelling side before I turn the key for the first time.

I'm running a pair of short Deka 630cc injectors mounted just after the throttle butterfly. Fuel pressure is set to 3 bar. The injectors are wired to INJ1 & INJ2 outputs of my UA4C speeduino. The engine is a 4-cylinder, 1293cc A-series.

I've attached the tune - any feedback is much appreciated

[PSIG]

Yay! Always fun getting to the first-start. I'd take care with fueling as the flow with those two is enough for about 220hp. Are you near or hoping for that?

If tuning to 5800, remove/raise the soft rev-limit, so it can run cleanly to redline. Soft limit is a special function tuned for specific uses, when needed.

VE looks good, but be prepared to creep-up in boost, and alter AFR Table values to a cleaner AFR at higher boost. Seems safe, but there can be many issues throwing very rich AFRs at it before you get there. Save heavy retard and wet AFRs for when you need them to solve problems. Tune like it's NA until you can see what it needs, when it needs it.

With that much fuel available, be ready to use Flood Clear to establish if a no-start is excess fuel or not.

+1 on setting injector characteristics. Log everything. Save, then rename the tune for any significant changes, so you can go back to where you were easily, giving freedom to testing and experimentation. Use MLV to 'see' what it's telling you in logs. I didn't look at every setting, but it's looking good!

[gcosens]

Thanks for the review PSIG! I am excited for this - it has been a long time in the making!

The car was running 140bhp on the carb. The reason for choosing such high flow injectors is that I sized them for a target AFR of 11.5 at full boost and 90% duty.

Assuming my calcs are correct I'm a little worried about hitting my 90% duty limit - although I guess I've got a lot of tuning to do before I get back the same boost I was running with the carb.

Thanks for the note on the soft limiter - I had thought that soft limit before hard made sense but I guess running straight into the hard limit is probably safer for the exhaust valves / turbo! I'll switch off the soft limit and bring the hard cut down to 5600RPM (which is really my desired rev limit)

The reason I've got such high AFR targets on boost is that these engines tend to suffer from mixture imbalance between inner and outer cylinders due to the siamese inlet ports and I'm really worried about detonation! I've logged AFRs using the carb and my AFR target table is set up to match that log data. Point taken that such high AFRs may not be necessary though

Thanks for the hint on flood clear - I will stamp that throttle if needed!

I'm using GIThub to version control my project because I'm working between PC and laptop. So any tune updates will get a new commit with notes etc.

Regarding datalogs - point heard loud and clear.

Hopefully the injector connectors will arrive tomorrow and then I might be able to attempt a start tomorrow evening

[gcosens]

It runs!!!! Well... sort of.

First two attempts had the odd cough but after opening the throttle on the third attempt it fired up.
Running very rough after that but this feels like progress nonetheless

It's too late now to run it properly and try to dial things in - hopefully tomorrow I can try again.

[PSIG]

The reason I've got such high AFR targets on boost is that these engines tend to suffer from mixture imbalance between inner and outer cylinders due to the siamese inlet ports and I'm really worried about detonation!

I get it, and understand the approach. But let me offer a different approach that serves both safety, and peak performance, following why you are doing something based on real info. This is a much calmer and lower anxiety way to tune - better. It is the opposite of "well, I guess it's good – it didn't break".

When tuning from 'bottom-up' to build the foundation tune base, the trends and results are seen as you go, allowing prediction of nearby areas while heat, pressure, and loads are still low. By initially tuning to best performance (only), regardless of fuel and timing required to get that (hey - best is best), you can then also know immediately when you begin to find a find a 'bad' trend, and can apply only and exactly what it needs to avoid that bad trend. Tickle the dragon.

We can see that by applying anti-det fuel and timing ahead of time, that a developing issue is hidden, and worse you don't have much to counter it with. You used most of your weapons up already. The result is suddenly finding yourself deep into un-fun territory, with few tools. This situation is also coupled with reduced power and efficiency along the way, by running rich and retarded when you didn't need to.

So, I hope this highlights how we can develop a clean picture of the engine's response, allowing better tuning by working up to it. It is the basis of diagnostic tuning, and testing both sides of a result to find 'best'. If you don't know exactly where you and the engine are, don't walk deeper into the dark forest.

"Diagnostic" = "which setting is better?"

Find BEST Tuning-Diagnostic bracketing.png (16.1 KiB) Viewed 42195 times

BTW - have you considered running cheap per-cylinder EGT to know if one cylinder is hotter or colder at a certain point?

[runesm]

I can only strongly advice you get the injectors on top of the throttleblades. Look at how oem did it. The airspeed over the throttle aids vaporization and to homoginize the mixture.

I've seen melted outer pistons that way unfortunatly

[PSIG]

If positioned to inject across the high-velocity airflow off the blades, I've seen no issues with cylinder imbalance. Loads are too low to melt anything at idle and cruise. The issue disappears at high-loads and speeds, where the blades are wide-open and have little effect on shear and vaporization. Yes, there are fueling differences that you can't ignore, but I haven't seen any horror stories or even indicators of them - so far. YMMV.

[gcosens]

Thanks PSIG - would some thermocouples strapped to the exhaust headers be any use for EGT? Or do they need to be in the gas flow? From some reading it looks like others have tried correlating EGT to AFR on these engines but they struggled to get accurate results due to the middle header serving two cylinders and the outer headers only serving one each. Might be interesting to try though.

It's running and driveable now (strictly off boost so far) but I've made a couple of discoveries that lead to questions...

First discovery is that 4 squirts / cycle, 4 cyl, 2 injectors seems to actually give me 8 squirts / cycle total (4 squirts for each injector per cycle) when checked on a scope. I only looked into this because the idle pulsewidths were way shorter than I had calculated. I've since reduced to 2x squirts / cycle in order to get one injection event / intake event (also confirmed on oscilloscope). I just want to check this matches expectation for the squirts / cycle setting?

Another interesting discovery is that after some hand tuning and then running VEAL, I have a massive spike in the VE map at ~1500RPM / 70kpa. There's another smaller spike around 4kRPM at similar loading so I wonder if this is some sort of resonance or something related to the cam profile?

I'm pretty sure it's a real VE increase and not some dodgy result from VEAL because if I smooth out the value to match the surrounding ones I get a horrible hesitation / misfire when entering that cell on the VE table.

I've included a screenshot of the VE table. I should mention I also halved my R_F to get better VE resolution hence the enormous numbers...

[LPG2CV]

Perhaps turn the issue around. Is it feasible that you have a misfire at those times because of say vibration and a connector or some such issue?

Then, if you have a misfire it will appear as a lean mixture. Then VEAL will try to add fuel to compensate.

[gcosens]

Thanks LPG2CV, it was running very lean in this cell prior to running VEAL. Since the VE increase, the lean condition has improved and it generally drives very well in this area so I think it really did need more fuel there.

I'll have a look at the log anyway to see if anything dodgy is going on in this area