Tuning with MAF on a 20V 4AGE

wrenchdad · Post by **wrenchdad** » Fri Aug 18, 2006 8:07 pm

Well, I plan on leaning out my cruise for better mileage but I want to first set it up for 14.7 and get a good base table and then play around with leaning it out, I have read of others that are running there 20Vs leaner as much as 16.5 at cruise, I may not go that lean but I am thinking somewhere in the range that you are running.

On the events, should I stay with the Msec type of adjustment or go to ignition events? I am thinking staying with what I have would be better, that low of a setting with ignition events might make it too fast a 3-3.4K rpm cruising. I think you are right after looking at the logs again, it appears that it does stay at the 95 reading for a lot of cycles before it starts to turn around.
SO I will not worry about the Air temps right now, and see what dropping the cycles will do.

Thanks again, wd

ps, My targetAFR table is set at 14.7 for cruise right now, covering all areas that I cruise in with the higher loads target set at 13.1-13.2.
later wd

superchargingmachine · Fri Aug 18, 2006 8:31 pm

I think that setting for 14.7 is a good idea to start then you can easily adjust to any AFR you want mathematically. Look in the megamanual and there is a simple formula. I have used it before and it worked perfect.

I use Msec but I have never tried ignition events. The nice thing about Msec is that it is independant of engine speed.

wrenchdad · Post by **wrenchdad** » Sat Aug 19, 2006 5:37 am

superchargingmachine

I will have to try that formula and see where my table is at. This morning I tried some different settings on the EGO stuff, drop the cycles down to 40 and then later down to 25, but was still seeing the swings. I bump the authority up to 15%, didn't see any improvement. So I finally just set the step size to ZERO and turn the EGO off.
Yes, this did stop the swings, well any swings that didn't come from me that is. I am of course still seeing rich swings on high map (96+) but it does appear that the problem is in the settings of the O2 stuff or I am just not as close as I think I am on my VE table.
This evening I am going to leave the step size at 0 and watch and log the O2 readings and see how the table looks without any correction. I have also push the AE up and out of the way, too.

I have a question, how much does your RPM move around when you are on cruise? Is it supposed to be rock steady or are 20-25 rpms swings OK?

thanks for your help, later wd

superchargingmachine · Sat Aug 19, 2006 7:24 am

I had simialr issues to what you are having and the solution for me was in my targetAFR table. When I looked close at the actual operating point versus what I thought I had in my target table I found that the target table was slightly off causing EGO to try and over correct. Try opening your target table while driving and have a look at what bin you are actually in, it may surprise you. Really your EGO settings shouldn't make a big difference if actual VE = Target VE(AFR).

I think the formula is a simple ratio of (AFRnew/AFRold)XVEold=VEnew

I'm not sure about the RPM thing. I watch my tach and it is rock steady and the Logs are 2 digit RPM values so I don't know if I would see a +/-25RPM swing unless I watched close in MT. I will give it a look today

wrenchdad · Post by **wrenchdad** » Sat Aug 19, 2006 8:41 am

superchargingmachine

Was your target AFR in volts or in AFR? Mine is in AFR for both the VE and Target AFR tables. Also what about the voltage switch point, I set mine to 2.5 but it changed it to something slightly different, which I figure is because it had to convert it over to a digital format. But when I look at the logs, checking for when the O2 switches over at 100%, I have not seen it even close to 2.5 vdc, it has been 2.8, 2.3. 3.0 or even 1.8 in the logs. Most of the time the 100% GEGO reading only last for about two lines in the logs or about .2 seconds and then it just passes on thru either headed leaner or richer. Not much control there.
I know the instructions tell you to make sure that the O2 is grounded at the same point as the ECU which they are both grounded at the very same point, so that should not be an issue but I think I will put a meter from the O2's ground to right on the ECU's ground on the DB37 connector and make sure there is no difference between them. Just to be sure.
I am also thinking about running a larger seperate wire to feed the ECU, right now the same wire that feeds the ECU fuse also feeds the O2 controller, so maybe, just maybe I am seeing some 12vdc swings when the O2 heater kicks on??? Is there a way to look at the 12vdc supply to the MS on the log viewer? It might help me track this down. Just a thought, you know the first rule in troubleshooting, check the power!!!

thanks and later wd

ps Is there a way to check my O2 sensor to know exacly what its switch voltage is? Maybe just using 2.5 isn't right. Is 2.5 just the standard switch point or does it even matter to a WB?
Mine is set up for 0-5vdc and 10-19.4 AFR just like the instructions are telling me to set it up. thanks wd

superchargingmachine · Sat Aug 19, 2006 9:07 am

I use a Zeitronics, which is non-linear so my target table is in volts. This makes it all the more difficult for me.

I would have to open my datalogs and look to see what the actual switchpoint voltage is when EGO corrects but I can't open them on this computer, sorry. But, it does seem strange that ego is switching at something other than approximately 2.5V. If you look back in the log is 2.5V shown a few lines before the switch? I am just wondering if EGO correction lags a bit.

The deal with entering 2.5 as a switchpoint and getting 2.509 or something back is in the 8 bit conversion. With 8-bit you get 255 points between 10-19.4 and it is linearly scaled. So whereever 2.5 sits on that line is what you get back.

Without being able to put the sensor in a known gas, I don't know of a way to check it.

wrenchdad · Post by **wrenchdad** » Sat Aug 19, 2006 9:45 am

superchargingmachine

I went back and look at the log with the correction turn off. Looking for what the voltage was compared to the reported AFR.
I was looking at the most lean and most rich places, somewhere I read that the WB acted backward from the NB in the way they work at the switch points, I can't remember where I read it at but I wanted to see which way the voltage to AFR was on this sensor, low volts= low AFR or rich.
Then I looked for what the voltage was at 14.7 and it is not 2.5 volts. Its is 2.373 at 14.746 AFR and at 2.51 vdc its 15.02 AFR.

SO I am thinking I have a .2 volt error that is making it not match the target like you were saying, So I think I will change the switch point voltage to as close as I can to the 2.373 and see what that does on the way home. I think changing the switch would have less effect on what I have tuned so far.

Thoughts???
thanks and later wd

superchargingmachine · Sat Aug 19, 2006 10:21 am

What sensor are you using and what does the company say the switchpoint voltage should be? This is what the output curve for mine is suposed to look like. http://www.zeitronix.com/Products/zt2/V_AFR.jpg

Are you sure you have Megatune configured for the right WB?

wrenchdad · Post by **wrenchdad** » Sat Aug 19, 2006 11:12 am

I am using one from DIYAutoTune made by Innovate Motorsports model # LC-1. I am not sure where to go to get the switch point vdc. I am just following the instructions that were on DIY's site on how to set it up, but I guess that could have changed and not been updated. But they really didn't say what the switch point would be. I just figured because they want the range set from 10 to 19.4 AFR and at 0-5 vdc with makes mid point 14.7 that 2.5 would be the switch point.
But looking at the logs with the viewer, that has let me at least see that 14.7 does not equal 2.5 vdc. Maybe I need to go to Innovate's site and see if there is anything there about the switch point for their sensors.

As far as MegaTune goes, I went into the ini file and changed it the "set" to the one for my sensor with 0-5vdc output, so that should be good. I also changed the one in the viewer software to match that too. Maybe its just that at 14.7 it is not at 2.5 vdc, and if you set it up that way then there will be offset that will make control impossible.

I will bump the Innovate site in a few and take a look, thanks wd

superchargingmachine · Sat Aug 19, 2006 12:14 pm

I am 90% sure my buddy Neil runs the Innovate LC-1 WB so I ust checked one of his MSQ's. His switchpoint is set to 2.352V.

So I think you are on to something. I have been to their site and it seems difficult to get any useful information.

wrenchdad · Post by **wrenchdad** » Sat Aug 19, 2006 12:48 pm

superchargingmachine

Thanks for checking for me, and that sounds real close to what I had seen on my logs. I did go to their site and found that maybe I need to do a free air cal, it said to do them often (how often, who knows?) for the highest accuracy BUT I think before I do that I will try changing the switch point to something closer to what I am seeing at 14.7 and then set the step size back to 1, authority to 10% and the cycles to about 40 msec. AND see how the drive home does. Which will be in about 30 minutes.

I will keep you posted, thanks again, wd

Post by **muythaibxr** » Sat Aug 19, 2006 2:44 pm

If you have the AFR tables filled in and turned on, the switch point does absolutely nothing for you. Instead, assuming you have the wideband settings configured properly in settings.ini, you enter AFRs in the table, which get converted to voltages before being sent to the MS. The MS will then use the voltage it looks up in the AFR table as the switch-point instead of using the switchpoint in the EGO settings. The switchpoint is not the problem, so you may want to look at the other settings, such as time, and percent per step.

wrenchdad · Post by **wrenchdad** » Sat Aug 19, 2006 3:13 pm

Well its time to back the bus up, I tried the new switch point, didn't help, tried different cycle numbers, not much help either, tried different authority settings, nope.

SO I turn it off, and..... that didn't help either!!!! This morning it seemed to stop it but I didn't have much driving time after the change, steady state, that is. This afternoon I had over half of the drive home, 40 minutes, to try it with lots of steady TPS and flat ground. I watched the O2 as much as I could and I could see that the O2 and map readings were not really changing when I felt the surges. So I also turn off the idle control from closed loop to warmup only, no change.

SO right now all I really know is something is changing and it seems to get worse when its real hot outside and I don't see it on the logs. I will be off on Monday and Tuesday, I plan on rewiring the 12vdc feed to the O2 and ECU and also check the grounds again make sure there isn't any difference between the MS and O2 stuff. With the MAP running so steady, it is hard to believe that I have a vacuum leak but I will look for that too.

I will put a timing light on it in the mornings before everything in the engine bay gets heat soaked and again in the afternoon to see if the timing or dizzy signal may be falling off when hot. I am running wasted spark using two dual coil packs with built in ignitors from a 99 Corolla, firing them off the LEDs which is what several people are doing. I am using the stock dizzy with two teeth cut out of a 24 tooth wheel. Maybe my coils are getting too hot? Its hard to tell when I get home, everything under the hood is too hot to touch. I meant to check, how hot the coils were getting on the drive in this morning but forgot, I was running a little late.

I read somewhere that there are some resistors that are supposed to be checked in the MS to make sure they are not getting too hot, I have not done that, I will have to pull the MS and take the cover off and drive the car in the afternoon and see if they are heating up too much.

SO thats all the ideas I can come up with before I do some testing, can you think of anything else?
thanks wd

Post by **muythaibxr** » Sat Aug 19, 2006 4:35 pm

The resistors you're supposed to check to see if they're getting hot are not being used in your setup most likely... at least not the ones for ignition.

I'd start looking under the hood before I'd look at the MS... Maybe a fuel-pressure issue or something like that?

Ken

wrenchdad · Post by **wrenchdad** » Sat Aug 19, 2006 5:31 pm

Ken,

Thats something I have wondered about in the logs, where does it get the fuel pressure numbers from? I mean I see them in the logs but I know there is no pressure sensor. Because I am seeing swings in the reported fuel pressure?

I will have to try to rig something up because as you know these 20V don't have a test connection. I did a dead head pressure test back when I was having first start trouble, but that was without the reg, the pump was putting out plenty of pressure so the pump shoudl be OK but haven't tested the reg, could be swing on me, that is what it kind of feels like, like I am giving it gas when I am not and the pulse width nor the duty cycle are changing so the MS is not telling it to send more fuel, so the O2 may just be the effect of the fuel pressure. I have a inline test fitting and I know I have one old hose that I used for the deadhead test, I also have an old fuel hose from my 99 Corolla. But it uses a quick connector but maybe the other end will work.

got to go but I will keep ya'll posted, and thanks again wd

Post by **muythaibxr** » Sat Aug 19, 2006 6:49 pm

The fuel pressure logging does nothing unless you've got a sensor hooked up for it.

Ken

wrenchdad · Post by **wrenchdad** » Sat Aug 19, 2006 8:04 pm

I figured as much, just wondered what was making it change. I right now I am wishing I had one so I could track it with the O2 reading.

I will look into the fuel pressure as soon as I can. I am thinking it should be running in the mid to high 30s at idle and mids 40s at high map readings, that sound about right?

thanks wd

wrenchdad · Post by **wrenchdad** » Sun Aug 20, 2006 9:09 am

Guys

I appears that I may have a miss, althou I haven't been able to check the fuel pressure, yet. Just after I get out of warmup, I tried to cruise in a lower gear (3rd) at about 40mph, and could feel the car missing.

I remember that when I first got the car running and was checking the base timing with a timing light, ever once in a while I would not see a flash on the light, it was like it missed a spark? I wrote it off as my light being 25+ years old, But now I wondering if that single missed spark was a minor problem that now after driving about a 1000 miles has become a major problem?
I also looked carefully thru my logs last night, and I am seeing a timing spike that lasts right at .1 sec that has been as much as 18 degrees !!
Seems to happen during shifting in the gears but only happens once or twice per log? Mostly at the beginning of the logs as the engine temp is coming up to full operating temp. I had missed it because I had been looking at the parts of the logs for steady cruise.

any thoughts on what to look for? I have gone back from my first install and changed out both of the coil signal wires to wire that is shielded at least thru the firewall, they don't go all the way to the MS like the dizzy signal does. And I pulled the dizzy and removed all the G pickups before I had them wired to ground thru 220 ohms resistors, now that are just gone.
I did all that trying to get rid of the missing spark flash but it didn't change anything, and then the car was running pretty good, real rich but none of this surging. I also wrote in another thread that I am not seeing the 5vdc square wave signal to my coil/ignitors off the LEDs, I changed out the pullups from 4.7K to 1.2K to get at least 3.5vdc but that was the best I could get, I put in a couple of 10k pots and adjusted them down while watching the spark output on a scope, anything higher in voltage would just kill the signal. Ken told me not to go lower than 1K that is why I settled on the 1.2Ks
SO maybe I haven't put my spark issue to bed, yet??? And maybe I am about to kill my coils if I haven't already??? For dwell, I am running 8ms cranking and 3ms running with .5ms for min. if I remember right, the running maybe at 4ms, I will have to double check that.

Well, what do ya'll think? Sorry if I didn't think about this sooner
thanks wd

superchargingmachine · Sun Aug 20, 2006 5:27 pm

WD,

It sounds like your running the VB921's and a distributor. I can't be much help on this setup. I only have experience with my EDIS setup. I hope someone else can chime in and give you a hand.

I

Dr.Hess · Post by **Dr.Hess** » Sun Aug 20, 2006 7:06 pm

Hey guys.

SCM, wd is running two 1ZZFE 'rolla coil/igniters in waste-spark, with timing off the dizzy wheel modified as per Ken's destructions of grinding off two teeth. I have the same setup, but with Camry coil/igniters which are functionally identical but have different part numbers.