Tune - Can't Fine Tune

[planetcaravan]

Hi All,
I'm running MS2 v3.57 using MSExtra 3.3.1 and TS v2.5.05. The setup is running a throttle body injection and EDIS setup on a straight 4 cylinder. After much tuning, on the dyno and on the road, the car runs well. The only thing I haven't been able to tune out, after much trial, error and experimenting is a stumble when changing up the gears. It is most noticeable between 3000-4000 rpm when getting back on the gas. From the data logs the engine seems to go real lean when getting back on gas between that rpm range but I have accel-enrichment on and it is kicking in at that point. I have tried giving it more accel-enrichment but that did not seem to help. It seems less noticeable at higher rpm. I have also tried adding more fuel to that area of the map but to no effect. Could anyone assist in pointing me in the right direction here?

In the attached data log i used the marker option in TunerStudio to mark when the engine leans on gears change (the marker is a little past the actual point due to my delay in hitting the spacer bar while driving). Also included is the current tune file.

Any help would be fantastic.

Thanks

[elutionsdesign]

Your accel enrichment does need some tuning. I'd turn OFF the wall wetting part for now, your accel time is quite short, lengthen it out to 0.8-1.0 seconds. Your accel threshold seems reasonable but your first accel point is essentially at zero TPS v/s so that's never even being used, move it out to your 150 threshold. Then lower it down to somewhere around 1.0 mS, build your curve up from there. Log that and see how it drives.

[planetcaravan]

Thanks for the reply. I will try out your suggestions for ACCEL ENRICH as soon as I sort out this new problem.

I've been running my MS now for about 8 months without any connection issues. Today when I hooked up my laptop the computer constantly goes offline. It pretty much goes online then offline every five seconds. The car starts, runs, drives like before, but it just won't stay online, so it makes it impossible to make any adjustments.

I'm running the old school serial port and cable. I assume it is not a cable issue as they either work or don't? Tried rebooting the laptop, re-starting TS, checked all the connections. I did a bunch of adjustments last week without issue, now this. I'm running TS 2.5.5.

Can anyone provide some assistance as I couldn't find anything the same on the forums.

Thanks

[Matt Cramer]

Things to check if communication drops out:

1. You aren't running non-resistor spark plugs, by any chance?
2. Are the grounds in good shape?
3. Any problems with the spark plug wires?

[planetcaravan]

Hi Matt, thanks for your reply

I am running NGK BP6ES plugs - these are non resistor type according to their website. What is the issue with that? I have been using them since getting my system up and running earlier this year without any problems until now.

I checked the resistance of the leads, they are all fine and are only a few months old.

The grounds are clean and solid onto the back of the engine block.

Any further advise would be great.

Thanks

[Matt Cramer]

Non resistor spark plugs put out a lot of noise. It could be the amount of noise was JUST below what was needed to cause a nuisance, and is now JUST above it - but I'd try going to resistor type plugs, as I'd give you a 9 in 10 odds that this gets rid of the problem.

[16vboost]

Once you get back into the accel tuning, try setting your MAP and RPM Lag factors at 95 or higher. With a low number, the signal lags too much and you're always enriching too late.

I'd leave EAE on for now, it's not that far off. But spend some time tuning your time based accell enrichment. I think you want to lower the accel amount at low TPS dots and higher accel PW for higher TPS dot. So a more slanted line. You'll figure that out easier once you set your Lags higher. (that's under "basic/load settings - general settings" btw)

-Alex

[planetcaravan]

Hi All,

Thanks for your assistance. Made some progress this weekend. The offline/online issue was solved with an update to TS 2.6.04 and a Java update. Thanks to Brian for his assistance there. I also changed to resistor type spark plugs as a precaution (thanks Matt). The issue was there even with the engine not running, so I imagine Brian's solution was probably the actual fix.

Accel. enrichment got some looking at too. Thanks to '16vBoost' and 'elutionsdesign' for the suggestions. The MAP and RPM Lag Factors went from 50% to 95%. This improved the driveability a lot. Much smoother on the gear changes too. The Time Based Accel. Enrichment was another issue though. As suggested I moved the first point to the accel. threshold of 150, and played with the ramp a little more adding and removing. The was no one thing I did that showed a big improvement here. The fact that I don't quite understand the relationship of this graph to the accel. enrichment doesn't help. Can anyone provide some explaination on this?

Overall the accel. is smother but still stumbles on harder acceleration. I also noticed the the injector PW gauge goes red (briefly) when I plant the accelerator to the floor. Is this something that I need to address? Also noted under full acceleration the air/fuel ratio is floating around 14:1, is this a bit lean for full throttle through each gear? I have posted the latest tune and a data-log with me accelerating the car hard through each gear.

Some guidance with more fine tuning would be appreciated.

Thank you.

[elutionsdesign]

The graph is showing you how much pulsewidth is added to that being commanded by the VE table, for the most part now you're adding way too much fuel. The gauge is going red because the pulse width is so high (duty cycle is over 100%). For example, one spot in the log the table is calling for 1.5 mS of pulse width, you plant the throttle and the accel enrichment adds 14 mS for a total of 15.5 or 10 times as much fuel and your AFR sinks below 10. Maybe 2X or 3X is better, you'll have to try it, and I'd also extend the accel time from 0.2 to 0.8 mS.

You'll also want to add some fuel to the upper MAP areas, you're seeing 14 AFR because it needs more fuel to match your AFR table but the EGO correction is off above 70% TPS and 90 KPa MAP.

[planetcaravan]

Hi elutionsdesign

Thanks for the information. I will have another play with those setting know that I understand them better. Could you point me to where the accel time setting is? You suggested going from 0.2 to 0.8ms but I can't find that setting.

Thanks

[kjones6039]

Could you point me to where the accel time setting is? You suggested going from 0.2 to 0.8ms but I can't find that setting.

Is this what you're looking for?

Untitled.png

Ken

[16vboost]

Personally I like keeping accel time at zero or close to it and increasing accel taper time as much as possible without going too rich. With this throttle body injection I think you'll need less "added PW" and close to a 1 sec duration.

[planetcaravan]

Hi Guys,

I'm having a very frustrating time with the Accel Enrich settings. No matter what I try I cannot seem to get rid of the stumble on re-applying the accelerator. Originally I had too much Accel Enrich which was also maxing out the injector duty cycle too. Now the log shows the engine going lean on the re-application of the throttle on every gear change. I tried the following based on your suggestions:

1. Adjusted the MAP & RPM Lag factors from 50 to 95 then 100 (this made a improvement)
2. Adjusted Accel Time from 0.2 in stages to 0.8 then returned it to 0.3 (no noticeable improvement)
3. Adjusted Accel Taper Time from 0.0 in stages to 1.2 (a little improvement)
4. Adjusted TPS based AE adding more and less PW (little effect)

Can some further assistance please be provided. I have included the latest tune and log. The log shows a hard acceleration through all four gears and clearly shows the engine leaning out on every gear change when getting back on the throttle.

Thanks for your help.

[Matt Cramer]

Try more pulse width at around 400 %/s TPSdot and see if that helps. You may also want to increase the "Zero accel above" and "Full accel below" numbers by about 1000 RPM each.

[elutionsdesign]

Having an accel time of 0.3 and a taper time of 1.2 is a bit of a conflict? Hard to taper off extra fuel for four times as long as its on. I suggest again to turn off the enhanced accel and turn off the taper time (set it to 0) and just focus on total accel time and amount of fuel. Once you get a handle on how that works you can turn the other stuff on.

[16vboost]

I don't think there's anything wrong with a short accel time and long taper time. Here's a graph from the old manual that shows what these do.



Maybe you do need more accel time. You don't have any room to go up in accel pulse width since you're close to 100% duty cycle during the accel event. So add more fuel by increasing the accel time. I think you're getting there. And I like the way EAE is working for you. I'd leave it on, but turning it off wouldn't hurt if you want to focus on only one thing. But you'll have to tune AE with it on once again.

-Alex

[planetcaravan]

Hi Gents,

I wish i had your faith that I was getting somewhere with this ACCEL ENRICH. Spent another 3-4 hours adjusting settings and test driving. Small improvements have been gained from your suggestions, but there has been no magic bullet fixit here. I have adjusted as follows:

ACCEL TIME from 0.2 to 0.6 (didn't feel any improvement going higher than this)
ACCEL TAPER TIME from 0.0 to 0.4
FULL ACCEL BELOW (RPM) 2500 to 3500
FULL ACCEL ABOVE (RPM) 5000 to 6000

All the above helped a little in reducing the flat spot on rapid acceleration - it is okay on steady quick application of the accelerator, but if you slam it to the floor (it is a club racer after all!) the flat spot appears. I have tried less fuel and more fuel and neither has made the flat spot disappear completely. The best result so far is attached in the latest log and tune. I have also adjusted the TPS BASED AE as follows:

TPSdot% 150 ADDED MS 1.5 (was 1.5)
TPSdot% 400 ADDED MS 4.0 (was 6.0)
TPSdot% 800 ADDED MS 6.0 (was 8.0)
TPSdot% 150 ADDED MS 8.0 (was 12.0)

I was starting to think if the design of the engine is an issue here. The MGB engines are an inline 4 cylinder with siamesed intake ports (i.e. only two intake ports) that is why I am running throttle body injection and not port injection. There is a lot of charge robbing that happens when the cylinders that share the same intake fire next to each other in the firing order. Could this be robbing the injected fuel/air mixture also from one cylinder to the other? The engine drives great in normal driving and stead acceleration, but when asked to go rapid full throttle it flat spots - maybe an intake velocity issue to do with charge robbing only at high intake speeds? Don't know, just running out of options to try!

Thank for your assistance.

[racingmini_mtl]

With the B-series engine, you need to run 4 squirts per cycle. You're running only 2 which means you will have a very bad fuel distribution with the charge robbing. With 4 squirts, that should be minimized as much as it can be with your configuration.

The MS2/Extra code does have a mode designed specifically for siamese port engines that allows you to run port injection. It uses the correct injection pattern that allows you to fuel each cylinder properly and completely compensate for any charge robbing. However, that requires you to tune the injection timing and you need at least 2 WBO2 sensor to monitor the inner and outer cylinders separately; that's the only way to tune the needed injection timing and adjust the fueling.

There a few Minis running with this code and they have an A-series engine with the same siamese port configuration (and there are also a couple of MGBs). The people running them have been able to get very good results not achievable with any other setups.

Jean

[16vboost]

Oh I forgot to mention this last time. You need to set your MAP lag factor and RPM lag factor to 95 or higher. What happens is you get on the throttle, map instantly goes to 100 kPa in the manifold, but with a lag factor of 50, megasquirt doesn't see 100 kPa, it sees 50% of the change in map, then another 50 until it eventually gets to 100. It only takes a moment but in the mean time MS is fueling based on the wrong VE cells. I think this might be your magic bullet. I for one can't tune transients with anything less than 95 MAP lag. Give it a shot.

-Alex

[planetcaravan]

Hi Guys,

alex - actually you did mention the RPM and MAP LAG factors in your first response. I did up these from 50 to 95 then to 100. It helped the most out of all the suggestions but did not eliminate the problem completely. Can this setting go higher than 100?

jean - i will it a try changing from 2 squirts to 4 squirts per cycle. Does this mean all my fuel mapping and settings will be out? I have read some of the mini megasquirt forums, i thought a lot of those exotic setups were so they could run port injection. It my case, i am running throttle body injection (a 45mm Jenvey TB with two Bosch 39lb injectors.

What do you think?

Peter