mastacox

I've decided to make a separate thread detailing my tuning experiences on my supercharged 1998 4Runner SR5 with the URD Supercharger Fuel Upgrade Kit. Enjoy. (This first post is a sort of a continuation from THIS thread.)

I thought I'd post a little update on my tuning. Sorry for the book, but I'd like to try documenting some of my tuning efforts. Just so everyone is clear, all of this tuning is being done in Los Alamos, and the highway between LA and ABQ. Elevation is anywhere from 5500ft ASL to 7800ft ASL. There are lots of big hills between the two which are excellent (and horrible) for inducing ping, so usually if I hear ping it's in top gear under high boost going up a 6-10% grade.

So with my fiance's help I have semi-successfully made myself the beginnings of a "Summer Tune" map. She has gotten very good at driving for me while I sit and stare like a little nerd at the computer screen, trying to spot problem areas. I have decided I definitely need to make a little momentary switch connection for ECU power for easier resets though (I'm copying you Mark), as I have been pulling the ECU fuse after each modification to the map. For this new summer map I pretty much started from scratch, and have embraced a few functions in the tuning software, specifically the "fill cells with X" and "adjust by X%" buttons.

Strangely, as things have been starting to warm up, I actually began experiencing ping again in the mid-to-high boost region (3-5psi on the boost gauge, but more like 1-3psi absolute) at 2000-2500rpms. This was a bit of a surprise to me since I hadn't had any ping all winter after I upped my timing retard to 6.5deg in some spots. I had to up the retard to about 7.2deg IIRC in order to fully eliminate it again. Perhaps from the winter/summer blend gas switch? Winter blend gas has more ethanol in it, it may have a cooling effect on the intake charge helping with ping in the same way a WMI system does. Anyway, I really need to take a long hard look at my timing map this summer because I get the feeling it just covers WAY too much area, and the only way to tell if I need retard in a specific spot any more is to lower the retard and see if it pings, which is a time-consuming process to say the least. Still, better safe than sorry I suppose. I may be interested in that "knock sensor light" you mentioned for something a little more accurate in detecting problem areas.

As for the fuel map, I have pretty much revamped it and have had pretty good luck at keeping the most of the long-term fuel trims near zero for closed loop operation (they usually hover around -0.8 to +0.8). I discovered a very useful feature in the tuning software is the "change cells by X%" button, because the fuel trims are also a percentage value. So I can take a look at the long-term fuel trims in a selected area, select the cells that I want to change by using [shift] highlighting, and chage them by the fuel trim's value (say -6% for example). This really has made tuning closed-loop much easier.

My current map is actually much more uniform than my original one, it seems to me that fuel map values are largely independent of the amount of boost applied (thanks to the FPR mod perhaps?) So looking at a specific RPM row, the values accross it are all the same, until I pass into boost and open loop. As a matter of fact, all of the values are the same from about 2000 rpms to 6500 rpms (I know that's way past redline) and -25inhg to 0inhg/psi, that value being 9.2 IIRC. When I pass into boost, the values decrease to 9.1 for under about 3000 rpms, and above 3000 they stay at 9.2. I have yet to tune open loop, but I suspect they will need to be decreased to 9.1 or even 9.0 because I am running at about 10.0:1 in open loop hard throttle right now.

On another note, the fuel map values at idle ended up having to be much different. I have them set at 9.8-9.9 from 500rpms to 1500rpms in vacuum after some warm-start and open-loop idle testing. I don't have it perfect yet (and I'm not sure I ever will) but I have helped alleviate many of my problems for warm-start lean conditions and cold-start lean conditions (as well as open-loop idle after heavy throttle applications, which is kind of unexpected). Getting the idle just right is the hardest part by far, but worth the effort. It is also difficult because I need to think about tuning open-loop engine decleartion (engine braking) where I'm running at about 12.0:1 (and have been ever since I installed the URD kit and wideband stuff) and I would prefer it to be closer to stoich. Problem is, some areas between idle and engine decel overlap and idle will have to take priority.

So that's what I've done. It's a long read, but maybe some of my experiences will help you in yours.

EDIT: On another note, my Walbro fuel pump has been getting louder for some reason. After driving for several hours it will begin making a pretty loud whistling/whining noise (by loud i mean easily heard over the engine while parked and idling). At first I thought it might be some kind of vacuum leak but sure enough it's coming from the gas tank. It's only been installed for 6 months, it had better not be going bad on me. If it is, it's easy enough to replace but still... Maybe I'll bump up to a 255lph if I need to replace it?

mastacox

Well, it's really more of a "coasting" thing. If I do "real" engine braking (engine pulling over 1750rpms with 0% throttle) the fuel cuts to the cylinders, so yes the AFR shows off-the-scale lean. But if I'm driving along and I take my foot off the gas to coast (usually under 50 mph) the reading goes rich to 12:1. Since the ECU is actually in open loop for this area, I suspect this has to do with the supercharger adding more observed resistance to the intake and the engine actually pulls less air than it would otherwise expect. I just need to turn down my fuel map in that area, but like I said it is very close to the idle region and I prefer nice idle to stoich coasting.

Hmm, not really, I have to really put my foot down to induce open loop. Could it be you AFR Sensor calibrator is kicking quicker than expected? It takes less than 50% throttle for me to engage the O2 Sensor faker-outer in my system, all it needs is 1psi of boost and on it goes.

Yeah, I need a $50 laptop to keep under my seat... but I'm just so broke right now. I need to sell my old 4Runner stat! Good to hear about the engine temps. Summer will be the real test, eh?

mastacox

Well, it's not super loud but the noise is definitely distinctive and noticable especially at idle. I might try taking a video if I can later on.

Last time it was whining with a little more than a half tank though so I dunno what's up. But, it only gets loud when I've been driving around for over an hour or so (gas warms up?). I think I'll just chalk up the noise to its higher flow rating and wait for it to die on me

I have heard people say the pump is loud, so maybe mine just had a little break-in period.

mt_goat

Hummm...still sounds a little weird, I'll take a note of what my mild coasting A/F is next time, but I seem to remember the 16-20 area.

How do you know that? Is this a URD O2 sim?

Yeah the real test will be on 100 deg day this summer.

mastacox

It's there, I promise. Like a little hidden "open loop" area. It might be easy to miss if you're not looking for it though. It's basically a small area in the ECU's memory between closed-loop idle and fuel-cut engine braking. I usually see it when I'm coasting or using very small engine throttle inputs for driving around town.

Yes, I have the FTC1-E with built-in Oxygen Sensor Signal Conditioner, which works in the same vein as your AFR Sensor Calibrator, but mine is for narrow-band oxygen sensors and yours is for wide-band (AFR) oxygen sensors. You AFR Sensor Calibrator should be set to begin faking out the ECU with a "false AFR reading" at about 1PSI. You'll need to tune the areas of the map where the AFR Calibrator is active as if it were open loop.

mt_goat

Oh wow....a missing piece of the puzzle has been provided. But is the AFR cal actually sending the ecu into open loop sooner?

mastacox

Well, the description is a little stickier than that... I'm sure Mark is dying to jump in right about now.

Basically the idea is that in closed loop, the ECU uses feedback from the AFR Sensor to decide if it needs to richen or lean out the mixture. The difference is that in open loop mode, the ECU goes to a pre-set table of values set by Toyota to decide how much fuel to dump into the engine. Because you do not have access to reprogram this table, you need to instead lay your own set of values "on top" of it in order to get the desired AFR. But, because in open loop the ECU doesn't actually see what the AFR is, you can tune at will.

For closed loop however, the problem this presents for tuners is that while the ECU is actively trying to maintain a stoich mixture, it will agressively fight any adjustments you try to make to the mixture while in closed loop. If you try to add fuel while the ECU is fighting you might see a brief moment where you mixture richens, but in the end the ECU is relentless and will win; that is unless you can fool it. Running in closed loop can be a problem because you can be in boost BEFORE you hit open loop mode, which can be dangerous for the engine in the form of high EGT's. While in boost, it is best to try and maintain a 12.0:1 rather than the stoich fuel mixture the ECU is aiming for. So, there is an area of engine operation that you need to control, but the ECU want's to fight it. What to do...

Enter the "oxygen sensor calibrator." For a narrow-band sensor you may know it as the Split Second ESC-1 (or URD Oxygen Sensor Signal Conditioner), and for a wideband sensor URD has the AFR Sensor Calibrator. The idea is that if the ECU thinks that the fuel mixture is stoich, it won't make any fuel trim adjustments to the cylinders. So what the oxygen sensor calibrator will do is intercept the oxygen sensor's signal to the ECU at a specific MAP pressure value (1 PSI is a nice start) and will instead send a "fake" signal to the ECU that makes it look like the air fuel mixture is exactly 14.7 (stoich). It's important to note that the ESC1 is not programmable; however, the FTC1-E with built-in signal conditioner (what I have), and the AFR Sensor calibrator (what you have) are. Because the ECU sees 14.7, it is happy and makes no fuel trim adjustments, leaving you free to adjust the fuel mixture at will without fear of reprisal from the ECU!!! Very cool idea. The ECU is still technically in "closed loop mode" but from the tuner's perspective it's the same as open loop, in that you have direct control of the fuel mixture.

Little clearer now?

mastacox

Here's a link to a thread where Mark, I and you had a conversation dealing with this very subject actually, starting on post #17...

https://www.yotatech.com/forums/f123/need-info-splitsecond-esc1-experiences-92688/

mt_goat

Oh yeah, that thread is where you said this:

That's where I'm at now, did you get that figured out?

mastacox

Well, I mostly got there. I do have it so that when the oxygen sensor signal conditioner kicks in at 1psi now instead of 4psi, and as such the AFR goes down to 12:1 and I can stay in boost/closed loop and not worry about the engine roasting itself. However, I havn't yet tuned open loop delivery fully. Once I fully tune open loop to be at 12:1 also, everything will be happy.

Overall I'm happy with my tune thus far but I've been taking it easy on the 4Runner to help conserve gas, and more importantly take it easy on the transmission until I get the VB upgrade. Once I get the VB upgrade and a tranny temp gauge, the green flag is dropping I'm confident my tune will be excellent though.

Here's the steps I would try to tune your engine. Make sure each time you change the map to reset your ECU, so that long-term fuel trims don't mess up your tuning.

1) I would say your best bet is to first fill your entire vacuum region with the same number (perhaps 9.0 or 9.1), pull the fuse on the ECU to reset it, and then drive around a bit to watch your long-term fuel trims. Watch what the fuel trims look like as you drive in various areas of the map. If they all look the same, adjust as needed using the "adjust cells by %" button, selecting the entire group of cells by selecting the top-left cell, and then holding [shift] and selecting the bottom-right cell. If you can, it's good practice to try and identify regions that have similar properties (whether it's fuel trims or AFR's) so that you can adjust entire regions of the map at one time instead of single cells. Repeat this process a few iterations until you have nice low fuel trims in vacuum. Keep in mind you'll probably have to up the numbers in the first two rows (idle region, 500 rpm to 1000 rpm) up to closer to 10, maybe 9.7 or 9.8. It'll be a little boring, but a nice excersise in getting "regular" driving under control.

You're also going to have to watch your hot-start idle and cold-start idle, since these are open-loop scenarios where the ECU isn't looking at the O2 sensor, so you have to tell it how much fuel to put in the engine. You'll get the hang of it.

2) "Fake" open loop, and real open loop. Here, you have to set your AFR Sensor Calibrator's critical pressure to where you want to start tuning (probably 1psi for engine safety reasons). Now, you'll have to get your driver to practice pressing the throttle enough to induce small to moderate boost and keeping it there, so you can watch the AFR and decide how much change must be made. It's probably best to try doing it with RPMs constant (or in a tight range) and pressure increasing, since from what I've seen your fuel map values will not be very dependent on pressure. It's usually easiest to do this on large hilly terrain, but since you live in OK you might have to improvise Pay attention to the area where the AFRSC first engages, and try to "smooth" the transition so that when you go back into vacuum from boost you don't have a spike in fuel trims. You'll also probably see a richening of the mixture when the ECU goes into open loop, so just be ready for it and adjust your map accordingly.


How's that for a start?

TrackRunner

Thank you for sharing your knowledge with us

mt_goat

That's great for a start Brian, thanks very much. Couple of questions, how do I set the AFRSC activation pressure and is it obvious when it engages? As you can tell I haven't ever messed with it.

mastacox

Well, I'm guessing (although I don't know for a fact since I don't have one) it came with some form of software that should be able to do that for you, by plugging in to the serial port on/in the AFRSC box with your laptop. It could be that the default setting is 1psi, and it might tell you in its manual (assuming it has one). Otherwise, URD are the people you should call. I should mention it is within the realm of possibility that although the box has a serial port, it is meant to be set up like the ESC1 in that the trigger pressure is set to 1 psi and is not meant to be changed by the end user.

As for telling when it is working, you need to watch the MAP reading from your FTC (with the SplitSecond software on your laptop), and the short-term fuel trims on the OBD-II scan tool. Assuming the AFRSC is set up correctly, as soon as you hit 1psi on your MAP, your AFRSC should engage causing your short-term fuel trim to go to zero on the scan tool, and also cause your AFR reading to go somewhere other than 14.7 (unless by some fluke your map is set such that the air/fuel mixture is exactly 14.7:1, which is unlikely). When the AFRSC is engaged, your AFR reading should proportionally follow the values on your fuel map, rather than the oxygen sensor's signal. Changing your fuel map's values should precisely change the AFR of the engine, so tuning should be the same as tuning in open loop (as shown in the U-Tune guide).

As with the vacuum area of the fuel map, it's my opinion that the best way to tune is to try changing entire areas of the map at a time with the "change by X%" button, rather than individually changing cells. Also as before, in my experience the cell values will not need to change much as pressure (boost) increases, so rows of cells should be the same or at least very close to each other. This will hold true except for areas close to the AFRSC's transition pressure, in which case you will want to tune it such that there is a smooth transition between stoich (14.7:1) and 12:1 over the span of 2 or three cells. Make sure this "transition area" is on the right-hand side of the AFRSC's transition pressure, not in the closed loop region.

Good luck! Let us know how tuning goes!

mastacox

Reading a little bit more about the AFRSC, it sounds like it actually has a map in the unit for defining when it should be active. So based on that, make sure it is "active" only in the boost region above 1psi.