AFR's jumping about and RPM hunting?

[robd]

Hi All, I've just been out in the car in my lunch break and using VEAL to dial in my VE table a bit more, this time with the settings Techsalvager(I think it was) posted up to try and stop MAT Correction doing much. I know this isn't really recommended, but my engine bay temps are very high at the moment so GAIR was going up and down quite a bit depending on speed, so for the mo I just want it so I can get my VE where I want it whatever the temps, then I'll adjust MAT Correction afterwards, or at least that's the plan!!

Anyway, sorry back on track. So I went out and VEAL was going great with GAIR not doing much, I stopped, got my grub and saved my map. IAT temps were just under 50C at this point so I thought I'd just log the drive back and see how it went. All was going well part way home, then I noticed my AFR's were jumping about and my RPM was rising even with my foot off the throttle and TS telling me my TP was at 1-2%! Now I've had this before and thought maybe it was just the throttle getting stuck, but my idle is about 900rpm (ish) and stopped at lights it was rising to almost 2000rpm! So I started logging from here to see if I could see what was going on. I also noticed that when off throttle in my overrun area, which is very lean, it was staying around 14.5AFR which is way off and completely different to what it was doing 20 mins ago on the drive out!

So yeah, AFR's are higher than they should be and RPM is raised, and it was always when either with the clutch down or off the throttle completely and it wasn't like it on the initial drive out or when the temps seemed lower(around the low 40's to 45C mark). If I blipped the throttle the RPM's would drop back to where it should be.

I've attached a copy of my log file for the drive back and my tune. The tune is still a bit all over the place at the mo but hopefully it may shed some light.
Any help greatly appreciated. Car is a '90 MX5 1600 running ITB's and 265cc injectors with no ICV.
Cheers
Rob

P.S. I zipped the log up as it said it was too big.

[jsmcortina]

Hi All, I've just been out in the car in my lunch break and using VEAL to dial in my VE table a bit more, this time with the settings Techsalvager(I think it was) posted up to try and stop MAT Correction doing much. I know this isn't really recommended, but my engine bay temps are very high at the moment so GAIR was going up and down quite a bit depending on speed, so for the mo I just want it so I can get my VE where I want it whatever the temps, then I'll adjust MAT Correction afterwards, or at least that's the plan!!

Just disabling MAT correction is similar to deciding that you don't want any gravity today.

James

[muythaibxr]

To clarify, disabling IAT correction doesn't disable the fact that changing temperature affects density, and that *should* affect fueling. Disabling it won't help any more than leaving it enabled... you'll just end up with unreproducible results every time you hit a part of the table that was tuned at a particular IAT and the IAT changes.

Ken

[robd]

No I completely appreciate that and I know I need it and I need to tune it. When tuning today my temps were fairly constant around 46C, but I just wanted to stop it making so many changes so I could get a solid AFR in that temp without worrying about it changing all the time. My plan was once I had the AFR's I wanted, I'll then put it back to 0's, let it do it's thing and then tune it from there to get my AFR's back to where they were tuned with MAT Correction effectively off.

It seemed to work ok today and was one of the first times the AFR's have dialled in quickly without lots of messing about. Whether that was becuase it was so hot and stayed around that temp or because of MAT Correction I'm not sure, but I'll leave it like that for the moment I think.

If you think that's what's causing the issue then by all means let me know and I'll change it.

[robd]

Been thinking about this some more and looked at my VE table, it looks like the AFR's are jumping about quite a bit due to the differences in the bins, e.g. at 45% fuel load and 800rpm my value is currently 39, whereas at 50% fuelload and 800rpm the value is 50. There are a few more like this where they're clearly not getting dialled in properly by VEAL for whatever reason or my map needs more work. Does this sound like the cause? Not sure if that would be causing the RPM's to rise and whether that's surging(?) or just something to do with my throttle?

Also a question on the MAT Correction, how much would the air density affect my VE table from say 46C to 50C? I know GAIR is trying to alter the fuel and remove some. The main thing I was trying to accomplish with the MAT Correction not doing much was to try and see that I could get stable AFR's, as with this issue now I'm not sure what's affecting it and whether it's just my VE table itself or another issue.

Also does VEAL incorporate the GAIR adjustments when adjusting the VE table? I guess it does, but I have noticed that if my IAT temps are say 25C, it will adjust my VE values differently to if my IAT temps are say 36C. With the weather the way it is at the moment and my engine bay temps, I'm finding it very hard to get a stable temp to tune at and that's where I end up going around in circles, hence the MAT Correction adjustment.

Hope that makes sense and any advice anyone can offer is greatly appreciated, even if it's telling me I'm doing it completely wrong! If I have to start my map again I will do, I just want to get it running as smoothly as possible with my AFR's.
Also a note on my map, the AFR's are actually 0.5 higher than they need to be as my WB gauge and TS are currently 0.5 AFR out, so I adjusted the AFR table to suit as I haven't had a chance to re-wire the WB yet to try and solve the voltage difference.

Cheers.

[nismoautoxr]

I have personally had VEAL do strange things with my VE table in higer load and lower RPM bins like you are describing and I attribute that to the tip in lean spike that it sees . On my car in VEAL and my sons MR2 it will richen dramatically the 50-80 KPA cells from idle to 1200 RPMs . I might be mistaken but that is just my theory . I just started manually tuning those areas with logical numbers and locking those cells during VE analyze ...I do that in overun areas too because if i dont it leans them out so bad that it makes the car almost undriveable in a slight coast situation where you still have light throttle applied. That is just me but I got tired of chasing my tail on those areas.

[techsalvager]

injector settings effect your fueling a lot.
Which injectors? model #?
what dead times\lag times are you running?
I've had old rx7 460cc injectors with a flat ve table of 59 stay at stoich around cruise 14.3 - 15.1 with those ve cells just at 59.

[pit_celica]

Been thinking about this some more and looked at my VE table, it looks like the AFR's are jumping about quite a bit due to the differences in the bins, e.g. at 45% fuel load and 800rpm my value is currently 39, whereas at 50% fuelload and 800rpm the value is 50. There are a few more like this where they're clearly not getting dialled in properly by VEAL for whatever reason or my map needs more work. Does this sound like the cause? Not sure if that would be causing the RPM's to rise and whether that's surging(?) or just something to do with my throttle?

Also a question on the MAT Correction, how much would the air density affect my VE table from say 46C to 50C? I know GAIR is trying to alter the fuel and remove some. The main thing I was trying to accomplish with the MAT Correction not doing much was to try and see that I could get stable AFR's, as with this issue now I'm not sure what's affecting it and whether it's just my VE table itself or another issue.

Also does VEAL incorporate the GAIR adjustments when adjusting the VE table? I guess it does, but I have noticed that if my IAT temps are say 25C, it will adjust my VE values differently to if my IAT temps are say 36C. With the weather the way it is at the moment and my engine bay temps, I'm finding it very hard to get a stable temp to tune at and that's where I end up going around in circles, hence the MAT Correction adjustment.

Hope that makes sense and any advice anyone can offer is greatly appreciated, even if it's telling me I'm doing it completely wrong! If I have to start my map again I will do, I just want to get it running as smoothly as possible with my AFR's.
Also a note on my map, the AFR's are actually 0.5 higher than they need to be as my WB gauge and TS are currently 0.5 AFR out, so I adjusted the AFR table to suit as I haven't had a chance to re-wire the WB yet to try and solve the voltage difference.

Cheers.

If you're idling in the 45-50 kPa region and at 800 RPM, then for sure, the AFR will vary if the VE is varying from 39-50. I suggest using a single VE value for the 2x2 cell region you're idling in. This will make your AFR a lot more consistent.

For the air density, assuming at sea level and 0% humidity, at 46°C, air density is at 1.103 kg/m^3. At 50°C, the air density is 1.089 kg/m^3. This means that the air at 50°C is 1.3% less dense. So, you need to remove 1.3% of fuel to stay at the same AFR. This is what Gair is doing. If the MAT is going up, Gair is going down. The problem is if the real air temperature entering the cylinder isn't exactly the one measured by the MAT. In this case, the air is denser than what MS think it is. This lead to a leaner AFR.

Gair is still used even when using VEAL, it's always there. It's in the basic fuel equation.

I suggest you to calibrate your AFR gauge more accurately. Even if there is a voltage difference because of the ground, you still can go around this by doing a custom wideband calibration. To do this, you need to know the ego voltage at 2 different AFR (11.0:1 and 15.5:1 by example). Set you EGO setting at "narrowband", but disable the ego correction by setting the controller authority at 0%. This will let you datalog the wideband voltage that MS is seeing in the datalogged field "EgoV" or "O2", I don't remember correctly the exact name. While datalogging, run the car at 11.0:1 AFR in your wideband gauge (not the one in TS). Do the same thing at 15.5:1 AFR. Play with the VE table to get these AFR at idle, try to have them the more stable you can. Then, check in the log what are the voltage at these AFR ratio. Finally, do a custom wideband calibration in TS by choosing "Custom Linear WB" and entering the two voltage/AFR points you have datalogged.

Sam

[robd]

I have personally had VEAL do strange things with my VE table in higer load and lower RPM bins like you are describing and I attribute that to the tip in lean spike that it sees . On my car in VEAL and my sons MR2 it will richen dramatically the 50-80 KPA cells from idle to 1200 RPMs . I might be mistaken but that is just my theory . I just started manually tuning those areas with logical numbers and locking those cells during VE analyze ...I do that in overun areas too because if i dont it leans them out so bad that it makes the car almost undriveable in a slight coast situation where you still have light throttle applied. That is just me but I got tired of chasing my tail on those areas.

Thanks for the reply, that sounds like I could be having a similar issue, as it does seem to constantly alter the same values even when it's just changed them and got them right a few mins ago!

injector settings effect your fueling a lot.
Which injectors? model #?
what dead times\lag times are you running?
I've had old rx7 460cc injectors with a flat ve table of 59 stay at stoich around cruise 14.3 - 15.1 with those ve cells just at 59.

I'm not sure of the exact details of the injectors, but they are 265cc injectors for a later model MX5, the MK2 1800. Here is a linkto what they are and they are the reds one, model 195500-3310 if that helps?
Injector dead time is 1.2ms

Been thinking about this some more and looked at my VE table, it looks like the AFR's are jumping about quite a bit due to
If you're idling in the 45-50 kPa region and at 800 RPM, then for sure, the AFR will vary if the VE is varying from 39-50. I suggest using a single VE value for the 2x2 cell region you're idling in. This will make your AFR a lot more consistent.

For the air density, assuming at sea level and 0% humidity, at 46°C, air density is at 1.103 kg/m^3. At 50°C, the air density is 1.089 kg/m^3. This means that the air at 50°C is 1.3% less dense. So, you need to remove 1.3% of fuel to stay at the same AFR. This is what Gair is doing. If the MAT is going up, Gair is going down. The problem is if the real air temperature entering the cylinder isn't exactly the one measured by the MAT. In this case, the air is denser than what MS think it is. This lead to a leaner AFR.

Gair is still used even when using VEAL, it's always there. It's in the basic fuel equation.

I suggest you to calibrate your AFR gauge more accurately. Even if there is a voltage difference because of the ground, you still can go around this by doing a custom wideband calibration. To do this, you need to know the ego voltage at 2 different AFR (11.0:1 and 15.5:1 by example). Set you EGO setting at "narrowband", but disable the ego correction by setting the controller authority at 0%. This will let you datalog the wideband voltage that MS is seeing in the datalogged field "EgoV" or "O2", I don't remember correctly the exact name. While datalogging, run the car at 11.0:1 AFR in your wideband gauge (not the one in TS). Do the same thing at 15.5:1 AFR. Play with the VE table to get these AFR at idle, try to have them the more stable you can. Then, check in the log what are the voltage at these AFR ratio. Finally, do a custom wideband calibration in TS by choosing "Custom Linear WB" and entering the two voltage/AFR points you have datalogged.

Sam

Thanks very much for the info, especially the air density. For the moment I've adjusted the MAT Correction so at the higher temps it's not taking fuel away, as if I let it the car just bogs down and goes stupidly lean. I know it's not ideal, but it seems to drive the same from say 30C to 50C with GAIR only trying to take away and away 1% every now and again, when it does I can notice the difference and it doesn't feel as smooth as when GAIR is just at 100%. I've re-located my IAT sensor into one of the ITB runners now. Temps are possibly a little better and maybe quicker to drop down a few degrees, but still high. I think the high temp is more to do with just the temps in the way and all the heat from the radiator being blown at my ITB's and causing my high temps.

Thanks for that info on the WB calibration, I'm going to re-wire the WB first and see if that sorts it, if not I'll do what you've suggested as that sounds like it should work a treat.

I've stopped using VEAL now and just being tuning my VE table manually while driving and using the logs, this has led to a much smoother VE table so far and my idle now isn't jumping about and my AFR's are more stable. Still work to be done but it's def getting further than it was before.

[robd]

Been thinking about this some more and looked at my VE table, it looks like the AFR's are jumping about quite a bit due to
For the air density, assuming at sea level and 0% humidity, at 46°C, air density is at 1.103 kg/m^3. At 50°C, the air density is 1.089 kg/m^3. This means that the air at 50°C is 1.3% less dense. So, you need to remove 1.3% of fuel to stay at the same AFR. This is what Gair is doing. If the MAT is going up, Gair is going down. The problem is if the real air temperature entering the cylinder isn't exactly the one measured by the MAT. In this case, the air is denser than what MS think it is. This lead to a leaner AFR.
Sam

Just another question on this, the percentage that's being removed, in your example 1.3%, is that 1.3% removed from the value currently being used in the VE table, or is it worked out another way? I'm guessing it must be something other than just 1.3% of the actual VE bin value, as that would be such a miner change that it wouldn't affect the fueling in the way that I've seen so far?!

[pit_celica]

Here is how the final injector PW is calculated :

http://www.megamanual.com/v22manual/mfuel.htm#law

PW = Req_fuel*VE*MAP*GammaE + accelPW + injector_dead_time

where : GammaE = Gwarm*Gego*Gair*Gbaro

So, if you have EGO disabled, baro correction disabled and coolant temp at operating temp, the only factor in GammaE is Gair.

If you aren't accelerating (TPSdot/Mapdot is constant), then the final calculated PW relate only from VE, MAP and Gair. As VE depend on RPM and MAP, and Gair depend on MAT, all these inputs play a role in the calculated PW : RPM, MAP, MAT.

Have a good read!

Sam

[robd]

Wow, that was some reading there, still trying to digest it.

I have another question relating to what I asked earlier about air density affecting my VE and AFR's. So if I've mainly tuned my VE table at say 46C and then today the temp was about 36C but Gair was still at 100% and not making any corrections, would my AFR's still be different because of the difference in air density and if so how noticeable would this be?

The main reason this is still bugging me is becuase if I let Gair make adjustments, I can clearly see the car doesn't then run correctly. E.g. today, temps were cooler so a few times Gair just went to 99% or upto 101%, but even with that 1% adjustment it was then too lean/rich and I has smoke pouring out my exhaust when my AFR's were too rich. As soon as the temp in the bay got hotter, Gair went back to 100%(based on my current MAT Correction settings), my AFR's then went back to where I wanted them and it was no longer spitting out plumes of smoke.

So I suppose my question is, even on a cool day, once warm my IAT temps are at least in the 30's(C), so with this in mind, realistically how much is my AFR going to change if Gair was staying at 100%? Also as it appears my IAT is suffering from heat soak, which means so is the air entering the ITB's, should I let Gair alter the fueling or should I stick with what I'm doing now to alter the MAT Correction table to minimize the changes made to my VE table where the car seems smoother and doesn't cover the car behind me in my wallet full of fuel!!

Cheers

[pit_celica]

A colder air temp means a denser air induction. To have the same AFR with the same VE table at a colder MAT, you need to add fuel. This is where Gair goes into action (if you let it do its job). If Gair was at 100% at 46°C, then, at 36°C, Gair should be at 103.17% which is rounded at 103%.

If you let your engine run at 36°C with Gair at 100% with a VE table that was tuned at 46°C with Gair kept at 100%, the AFR should look leaner than before. In your case, you said it's richer, so there is something wrong. A 1% difference in Gair shouldn't yield to a large difference in AFR. I'm pretty sure something else is playing a role here.

You need to do a datalog of a similar run at each MAT temp (36°C and 46°C) with the VE table staying the same for both logs. During these logs, avoid hard throttle change and try to keep the RPM in a zone (let say between 2000 and 3500 RPM). The same must apply for the load (keep MAP in a zone : 20kPa to 70kPa). Disable Ego correction and baro correction (if applicable). Remember what you are trying to achieve : you want to see the effect of a change in MAT (and only MAT) on your tune. So, only the MAT should change between both tests. I suggest doing a log in the morning or in the evening to attain your coldest MAT target. Then, do the exact same run in the exact same manner (RPM zone, MAP zone, slow TP change) when the temp are higher like at noon during the lunch break.

Post both these datalogs and your msq used in these datalogs. Then we will be able to talk about it deeper.

Sam

[robd]

Thanks very much for the reply Sam, you're a great help.

I'll get some logs of both runs and put them up. Looking at my AFR's this morning where my IAT temps were about 20C, it was indeed leaner as you say. I think where it had gone richer at different temps is where the MAT Correction settings I have were adding fuel for a moment depending on what temp it was at.

I can see where I've gone wrong and need to adjust my VE or MAT Correction. I'd done most my VE tuning lately at hotter temps in the 40'sC, but I also had my MAT Correction table altered so Gair would stay at 100% most of the time as I wanted to work out what was going on with my AFR's all the time and GAIR.

This is my current MAT Correction table values and the MAT Correction Value is set at 50%...

20.0C - 0.0%
30.0C - 1.0%
40.0C - 3.0%
50.0C - 5.0%
60.0C - 8.0%
70.0C - 12.0%

So when I was tuning my VE at 46C, GAIR was still at 100%, AFR's were spot on where I wanted. So does this mean MAT Correction was adding between 3-5% fuel to keep GAIR at 100% when it should of say been at 95 - 97%?

So I suppose the question is, what best's to do from here, can I adjust my VE table based on the above correction so that at 20C my AFR's are correct as it's not adding any fuel as VE values are good for that temp? Obvioiusly I'd then have to set all the other MAT Correction values to 0.0% and see if the correction was then ok at higher temps. If so how could I work this out, from what I've understood(or tried too) so far, the GAIR/MAT Correction percentage is a percentage of the PW not the VE is that right? Also if I am still getting bad readings from my IAT sensor, could I adjust the MAT Correction so that it added fuel at the lower temps to match the correction at the higher temps, I know that's probably not the best way and I'm botching it, but just wondering if that will work?!

I know I've probably done all this completely arse backwards but it's a good learning curve at least, at least I can see for definite now that is is the IAT temps that are affecting the fuel and not anything else, as before I wasn't sure if it was this or another factor which was driving me mad!

Thanks again
Rob

[pit_celica]

Again, if your VE table is able to make your AFR spot on at 46°C, do not touch your VE table anymore unless you add/replace a part on your engine that modify the VE of the engine.

If, at a different temp, the AFR aren't spot-on, play with your MAT correction curve.

Here is how Gair is working mathematically (it's not written anywhere, but this is how I understand it) :

Gair_final = 100 + (airden_factor - 100) * MAT_correction_value/100 + MAT_correction_curve_point

Example : let say that airden_factor is equal at 100% when MAT is at 70°F. Here are some parameters for the equation : Actual MAT : 50°F, MAT correction value : 50%, MAT correction_curve_point at 50°F : 3%.

Result : airden at 50°F = 103.89% : Gair_final = 100 + (103.89 - 100) * (50/100) + 3 = 104.95% = 105%


This is how I understand it.

Sam

[robd]

Thanks again for the reply Sam, you're clearly better at maths than I am as right now that's going to take reading a few times!

One question though, if I do leave my VE table as it is with my current MAT Correction settings, would that not mean that if I then zeroed out my MAT Correction that I would then be too lean at higher temps as well?

So I guess I either need to tune MAT Correction further like you've suggested, so it's also adding fuel at lower temps to make up for the difference in my VE. Or adjust my VE to be correct at the lower temps so it will be richer VE values compared to what I have now, then let Gair do it's thing and see how much it leans out my VE table at the higher temps.

[pit_celica]

You're right. Currently, if you put all zero in your MAT correction table, you will run lean at 46°C. If I were you, I would put all zero in the MAT correction table and use set the MAT correction value at 100% and retune the VE table while trying to keep MAT constant when VEAL is doing his job (the use of the MAT filter in VEAL could be useful for this). Before doing this, add maybe 10% of fuel to the whole VE table to make the change easier at first hand. Then, tune the MAT correction table to be able to keep the same AFR when you are running at a colder or at a hotter MAT (vs 46°C). This is how it should be done, this is how the pros suggest to tune.

You could also keep everything as it is right now and tune the MAT correction table only. Keep in mind that the goal of this is to be able to reach the same AFR with the same VE table even at different MAT temp.

Sam

[pit_celica]

Here's a visual way to see how Gair is affected by MAT and MAT corection value (100% or 50%) :

viewtopic.php?f=91&t=45364#p324374

Sam

[robd]

Thanks very much Sam, that link was very useful as was your reply again.

I'll try a long run down the motorway and see what temps I get at slow and high speeds so I can then get a decent MAT filter going like you suggested, then I'll adjust my VE table to suit that temp and try tuning it again. I'll try it at 0's for MAT Correction but with 100% and 50% to see what works best, as before at 100% it just seemed too excessive the changes Gair was making but I'll give it a whirl.

Thanks again and I'll update with progress.

[robd]

Ok, got a little update on this, I adjusted my VE table as suggested and set the MAT Correction to 0's and the value at 100%. I then let VEAL do it's thing on a colder day and it started to dial my AFR's back in. I noticed that still at 100% MAT value it was affecting my AFR's too much and generally making it too lean at hotter temps, so I set the value to 50% and things were a little better.

I tuned my idle to be at 14.7AFR when colder, then when the temps got to 25C with Gair at 99% my AFR is still 14.7, then at 32C with Gair at 98% my idle was also still at 14.7AFR. The only problem seems to be when it gets even hotter, so when the car had been sat idling for a while, the temps were at 42C and Gair was at 97%, this was then leaning it out to about 15.2AFR! The only way I could then bring it back to 14.7 was to adjust the MAT Correction table to add fuel which effectively put Gair back at 100%!

I'll do some more testing at lunch time today as the higher temp of 42C maybe wasn't as accurate as the car had ben sat for 5 mins. But so far things are going better. Only main issue I have is catching my idle at lower temps when Gair is normally 100% as the engine bay temps rise so quickly once it's warmed up and when the car is stationary or at slow speeds below 40mph!