Hesitation When Punching Throttle
 

Hello, been on Megasquirt for the past 6 months and have been looking for a solution to this problem ever since I installed it. It's the same question everyone always asks, but I haven't been able to fix it with the common troubleshoots. Problem is when I step on the throttle (anywhere above 100 tps dot) the car has a severe hesitation (a full half/three quarters of a second) before the revs pick up. I have still been having trouble despite trying multiple "solutions" that I've heard of.

The things I've tried are:
Accel tuning (still has an initial lean spike EVEN THOUGH IT IS ACTIVATED)
TPS averaging lag factor
MAP averaging lag factor
Increasing VE cells above idle cells
Moving MAP nipple sources

Still always get that hesitation, and I know others have had it, but for some reason other "solutions" worked (of course it's not the case for me). Could it be the MAP line is too long? (goes from intake through drivers side firewall to ECU above pedal box) It is definitely an issue related to the ECU not picking up the change in MAP fast enough. And I didn't have this problem on the stock ECU eons ago. Any help would be appreciated.

Please don't say "You need to tune your car" - I have tuned it and everything is great and spot on (even accel values) except for this one thing.

My tune is also attached

Thanks

You need to tune your car. If everything was "spot on" it wouldn't have this lean hesitation when you tip in.
Look at your AE settings in particular.

And, since you've researched this already, you should know to post a tune file and a log as well that shows this lean/hesitation condition.
Otherwise we're all just making shots in the dark.

I'll post it later but I wondered if someone had the same problem and could just share their experience. Car does not really need to be tuned like I said because everything else is perfect. Whatever it is it's a little correction, not the whole tune

It sounds like it needs more fuel during AE. Your tune file and logs will help confirm that suspicion.

Make sure that you've got TPS, kPa, and AFR in the logs, since this will make it easy to see when you open the throttle (TPS) and how quickly the kPa and AFR respond.
Info helpful! Shots in the dark, not so much.

I thought the same because it does have a lean spike but there is no matter what a lean spike for a hair of a second right before.........every time. No matter if I put my AE to the moon or not, it still has the same lean spike and hesitation

I know how to give you guys a log lol. I'm not that newbish:naughty:

Log will come tomorrow :bigtu:

You need to tune your car. Your AE settings are bad.

https://cimg0.ibsrv.net/gimg/www.mia...9c05ab34ed.png

AE is very straight forward. Use a basic map from Rev or DIY and don't touch it ever again.
Miata engines like 14.7 for drivability, any leaner than that and you'll get slight hesitations. You can get away with 15+ at highway cruising.
But otherwise adjust your afr table for 14.7 where you have 15 currently.

Look I've already tried tuning AE. I'm coming here because after trying things like these I still get no improvement. I don't know how else to describe it, but there is always a lean spike that happens EVEN WHEN AE IS ON. I've tried setting AE literally to it's highest setting and everything in between. It's not AE. If it was AE I wouldn't be on here. I don't know what else to say, but if anyone has any other ideas other than AE that would be great. Still going to post a log later today so you guys can see what I've been talking about

What you are describing is AE, you're being stubborn. Or your TPS isn't hooked up.

Then I guess I'm describing it wrong. I've tried AE for 6 months and have had absolutely no improvement. The settings you guys would consider normal aren't even close to fixing it. I'm not trying to waste your guys' time with an AE question, I've read about it a ton already and have already tried it.

MAP hose length has been tested - it really doesn't matter. Unless you've got it pinched somewhere. Again, a log is needed.

Hey Melvin, I just downloaded your tune and have a few comments. Under Engine and Sequential Settings it looks like you are running on "untimed Injection" which seems to be batch injection instead of sequential injection.That might cause some of your tip in issues but not all. The second comment is that you are using TPS based AE with values added that are incredibly low. On my personal setup I am running a 1.6 engine with sequential injection and for value added I am running values added of (0,25,45,60,65) for tps dot values of (0,22,200,475, and 1000). I am not saying that I am doing it the correct way or that I am correct, just that you cantry upping your values of acceleration enrichment to follow some kind of curve. Maybe you can try fiddling with your slider scale to favor more tpsdot as opposed to map dot.
Your current acceleration enrichment values don't actually add any meaningful enrichment.

Edit: Also your trigger Accel Mapdot threshold kPa/s is set to 999 which is an incredibly high number. As in you won't get any accel enrichment until you go above that number. For reference, my number is 25.

Even if you've tried different settings, your current tune will give any miata hesitation when you punch the throttle. Your current AE settings add a maximum of 2% fuel. That is going to do jack shit.

I don't use much basic AE, mostly EAE, so I cannot comment on what are good AE-only settings. But I can assure you that my logs show at least as much as 50% extra fuel added on hard punches. This prevents leaning. So, if you are adding only 2%, I can assure you that is not enough.

Sooooo. Information provided. Let's hear how the recommended changes HERE IN THIS THREAD affect your situation.
Or, you know, you could keep telling us that our suggestions won't work.

Please don't be one of "those" guys. They rarely last long on the site.

Also, post a log file of the condition once you implement the recommended suggestions.

Remember when you punch the throttle, you're changing the map, which moves to a different section of the fuel table. Often cells above idle have a higher VE than the idle cells. So in some cases, I've found a car to respond well with AE as low as 2% where you have it. I could turn AE up and turn the fuel table down, but it's usually easier to turn AE down.

Also, start ignoring the lean spot, like, now. Tune according to how it responds to throttle input, not what the gauge is telling you. That's a delayed reading from a rush of air that entered the engine a few feet from where you're measuring. And with throttle response, every millisecond counts.

Fix your AE settings. Your current settings, you may as well have it turned completely off. It's not doing anything.

Thanks for more info. As far as my low settings, I left them there after trying higher AE values (all the way up to 40 added). The current tune would probably hesitate like you guys are saying, but since the higher values weren't working for me I just lowered it that's all. I will up the AE mainly just to please you guys lol and show that it still hesitates, or there is a possibility I missed something. I'll post a log with all the info in an hour or so and I'll change the AE and do another log. Till then

AE should start high and go higher.

Completely agree but I'll show you later that it still does it with higher settings

Ok. So logs taken and AE adjusted twice and two more logs taken. Using 100% TPS for reference. Pictures are below (log is under its AE counterpart). Car still hesitates despite going to both AE extremes (super lean and super rich) and even after hitting the sweet spot (second pic/setting combo). The second log/settings are definitely better as to be expected, but it still has a hesitation with them. Let me know what you guys think. I understand I may not have been clear in describing that I had already tried other AE settings and had then switched to my "anemic" super low table because I couldn't get it fixed anyway. Side note ignore the actual AFR values, they are off, but the actual magnitude of full lean is still a good reading. Thanks for the feedback
https://cimg9.ibsrv.net/gimg/www.mia...075a36a7b9.png
https://cimg2.ibsrv.net/gimg/www.mia...7bcc5e6a19.png
https://cimg3.ibsrv.net/gimg/www.mia...b456c1f654.png
https://cimg5.ibsrv.net/gimg/www.mia...5e5a6e8af0.png
https://cimg7.ibsrv.net/gimg/www.mia...bd6b5feae3.png
https://cimg9.ibsrv.net/gimg/www.mia...dbf1a6341a.png

Your AE settings still don't look anything like anything i've ever done on my own cars. Not enough, and too late.

Ok what have I done horribly wrong?

I'm still a noob to this...but what's up with your MAT correction and VE table?

https://cimg4.ibsrv.net/gimg/www.mia...eb36a39313.png

humor me and try this one, I think it's a combo of your mat correction and fuel table, not ae.

What do you mean when you say ignore actual AFR values? Are you saying your wideband is not calibrated and reading correctly?

Mat correction was something I changed because of heat soak. It's not that because I had it the way it came from DIY for a while until I fixed it. It still hesitated back when it was the default table. And my VE table is what works. Nothing abnormal

Yes afr isn't calibrated right in Megasquirt. Real afr values from the car are fine

You have an improperly calibrated wideband...and your "incorporate AFR" setting is on...

Ok thank you for pointing that out, I'll change it. It doesn't affect hesitation if that's what you were thinking

Thanks gooflophaze for the video. I have already watched it and have used his technique. I watched it again and will have to check pulsewidth like he said

You should really humor curly.

And if my quick changes don't work, start adjusting AE on it. Go up or down by 1%. I've had throttle response be crappy at 7%, great at 8%, and crappy again at 9%.

Edit: and by 1%, I mean 1% steps. You're throttle blips in in your logs are around 500 tpsdot, which is currently at 6%. So try 3, 4, 5, 6, 7, 8, and 9% before saying it doesn't work, which it might not.

And to double check for my sake too...when you guys are saying "add 1%" - you're not using 1% and 1ms interchangeably, right? Since my WOT can be ~11ms of pulsewidth, 1% would be +0.11ms, and 1ms would be 9%.

Why don't you start by fixing your AFR problem because garbage in is always going to create garbage out and your logs are going to be worthless

Thanks curly. Yeah I understand, maybe some fine tuning could be worthwhile. I don't fix my afr calibration because I can remember the AFRs. I can still tell when the AFRs change in an unusual way too, it's really just off by 1-2 points. And the full lean and full rich are generally good indicators. It hasn't been something that has affected me really

He should ignore AFR error, and the time delay, but not the value. If you god from an area where AFR target and actual under strady state is 14, to an area where target and actual Uber steady state is 12, then there should be a slight delay until AFR goes from 14 to 12; but it should never go above 14, if AE is doing its job.

Irrelavent. Incorporate changes the fueling equation, but not the active adjustment. That is done with EGO. I’m not saying that correct WB reading by MS is not important, only that it is not important because of “Incorporate AFR”.

Listen to Six. Tune steady state, then tune AE.

And regardless of if you like the shape of your MAT Correction, 100% should be at the temp where you usually tune. Not saying it won’t work like you have it, but it doesn’t transfer well. Kind of like setting a REQ_ FUEL to a random value, then trying to make sense of your VE table, or Cranking Pulses, both of which have relationship to RF.

Agreed on the AE afr thing you said. And my afr although not reading right is tuned correctly, so that's why I moved on to AE. It's just not reading the right translated value in Megasquirt

..... 1-2 POINTS?

Fix that. Immediately if not sooner.

It's just what Megasquirt reads but my AFRs are fine. It's not something necessary to do

I'd probably go back to square one on this. I couldn't imagine a world in which i'd skip calibrating sensors i'm using to tune.

And i've also followed Scott's AE video a few times and not once have i ended up with an AE curve that looks like yours. It's entirely possible that i rely more on AE than someone like Curly would, but your logs look exactly like what i'd expect with something that doesn't have enough AE going on. And that you're only having problems when applying more throttle also points to AE issues.

At least you've got EGO correction off i guess. But that sounds like it might be more because your megasquirt is being fed garbage by your wideband.

Must've shadow edited. The "current tune" I downloaded earlier had both incorporate AFR and EGO correction on. I could only imagine what would be happening, if the gauge said 14.7AFR, but MS says it's 12.7AFR, and is trying to lean it out to 14.7.......

How did you tune your VE anyways? If your wideband reading is "off" by 1-2 points...but you're memorizing it... I take it then you're driving the car, noting the spot on the VE table, holding that condition steady, looking at your physical gauge, take a mental note, adjust VE, revisit the same spot on the VE table, and confirm if your changes were correct?

Or do you drive for a while.....datalog...and then change the VE table one by one by analyzing your log manually (aka not using the VE analysis)? Because if you knew exactly how "off" your wideband reading (in megasquirt) was, I'd simply just adjust the entire AFR target table, and turn EGO back on - that's if you don't want to fix your wideband connection.

This could all be easier...and more efficient...if you fixed your wideband connection.

OP is well on his way to a motor replacement IMO

OP, what WBO2 set-up? If Innovate LC-2, there are known values to use for correction / calibration. Don't use mental offsets, or fudged numbers. Use correct calibration. I'm not sure how far off yours is.

My set-up with LC-2 and MS3 is 0 = 7.2 and 5 = 22.2 (vs Innovate's published 0 = 7.35 and 5 = 22.39). Brain publishes other numbers but they calculate to almost exactly the same corrections. These corrections are for the grounding error. Your errors may be different.

I have an AEM UEGO which isn't supported with calibration so it's kind of hard to adjust values. It's honestly not that hard tuning without an accurate wideband reading in Megasquirt because I do more myself. It's what where'stheboost said (his first described method). Megasquirt doesn't have a problem with my values and doesn't correct them because it's not an extremely different value

I would like us to talk about the hesitation problem not about my AFR being off in Megasquirt. I will check out some pulsewidth settings like Scott said but if anyone has any non AE related info I would appreciate it

I still will try to get good values for wideband calibration, but it seems like it's non linear and I've had places where it lines up and places where it doesn't

Do you not understand that so many of your problems are rooted in the fact that you wideband isn't calibrated properly? Fix this! The systems of the ECU aren't independent, they're inter-related and AFR is the basis for many of them. We aren't getting sidetracked to make your life harder and avoid the question.

Everyone should take a turn to say fix your wideband...

Fix your wideband.

Then again, I think everyone's said it already.

If it's off by entire points, that's not just a calibration issue. That points to your wiring being garbage and should be addressed.

It's also not hard to find the published calibration numbers/voltage points for that wideband on the internet (or in your user manual). I bet it would take entire minutes.

OP keeps telling everyone that they are wrong every time advice is given. It must be true.

After all the fuss about wiring...I just went and used the power and ground from the cigarette lighter. Gauge matches MS. Also AEM UEGO.

I've been trying to input wideband bounds for 6 months, ever since I got the wideband. I put in the bounds that came with the instructions (voltage vs afr) and it still doesn't work. I'm not a dumbass I know how to plug in numbers, and I've already tried. So I got it close and made sure nothing was affected by them. Maybe I'll try a new ground as that sounds like a cause. Ecu doesn't mind these off values as I watch what it tries to fix (nothing). Back to AE, some of you said my chart looks off. I didn't care that it wasn't finished or finely tuned because even when I max out the added fuel in AE, there's STILL a lean spike. The whole point of AE is to nullify that. So if it still lean spikes then there's something else wrong. I keep trying to tell you guys this. There are other things that can cause this hesitation. If you guys only know of the AE solution then just don't respond. If you want to come and try to figure it out like I've been doing, be my guest and you'll see why I'm asking this question and why I'm telling you AE won't fix it.

Thanks for a forum record probably -28 cats in one day

Have you read any of the responses on here? First we said it was AE - since that's usually what causes this. Then we saw the issues in your MSQ, then we saw the other issues. Then we list more issues...all surrounding your wideband. AE can only fix so much, but you need to fix your VE table. Can't fix your VE table if your wideband is botched. Fix your wideband. Get the gauge and the MS reading to match. Then we can reconsider looking at other things.

For fun's sake...when you do "fix your wideband", post up a video showing the same reading on the gauge and laptop ;)

No one's saying AE anymore. Fix your wideband. How is it currently wired up? What's the source voltage and what's it grounded to? I didn't need to change any numbers on my AEM UEGO. I literally hooked it up, turned on the MS, and picked the AEM UEGO XX-XXXX series, and everything matched. So...in short...

Fix your wideband.

Fix your wideband.

Fix your wideband.

There will be a voltage offset if the power and ground are sourced in different places from the wideband to the megasquirt. There will always be differences unless you source power and ground in the same place for both units. But since that advice is in pretty well every discussion about hooking up your wideband and hooking up a megasquirt and I'm sure you did it that way, that can't be your problem.

OP, you (apparently) have maps that use AFR Incorporate and EGO correction, both of which is entirely reliant on the AFR value that your Megasquirt sees. So before we even discuss your trainwreck of a VE table, which has to be completely tuned before you attempt to perfect AE settings, your fucked-up wideband input is wreaking havoc on all of the fueling calculations.

Do what we tell you to do. Fix the basics, and then we can discuss the settings that build upon those basics. Or, ignore us, and keep racking up the negcats.

You telling us that you want to talk about AE settings is like a 3rd grader who insists on talking about calculus without knowing what 7x7 is.