I'm wondering if a VAFC (after tuning) will significantly affect gas mileage? not that it's a big deterrent but i was thinking about it today and being that energy is somewhat conserved in a car engine (though not totally) the increased hp ought to lead to a decrease in gas mileage.
also, how do people deal with VAFC's and cali smog checks?
Gas mileage will only be affected where you drive most of the time. In other words, if you spend 90% of your time between 3000 and 4000 RPMs, part-throttle, then your mileage will only be affected if you tune there. If you keep it WOT in VTEC most of the time and that's where you do most of your tuning, then you will definitely notice a change in mileage.
Most VAFCs are installed on cars with light mods, and most Hondas run VERY rich from the factory. These cars benefit (power wise) from removing some of this extra fuel to bring the air/fuel ratio back down to what is ideal for power. This means you will be leaning the map out some and improving efficiency (and hence mileage).
There is absolutely utterly NO REASON to modify the part-throttle or non-VTEC maps for a car with light mods that idles OK. The stock ECU (OBD1 or OBD2) is entirely capable of handling most bolt-ons without any trouble. You should only modify the part-throttle and non-VTEC maps if you can't get the car to idle right or have part-throttle lean/rich issues. I have yet to see an NA Honda do this, it is usually only turbos or vehicles with injectors much larger than stock.
This is where OBD2 becomes a large problem, is with the LTFT and the VAFC trying to "fight" the ECU. Even with OBD1 there is no sense in not letting the ECU do it's job.
You should have no problem with emissions by the VAFC itself (can't say the same for a smog visual inspection).
Originally posted by marcucci There is absolutely utterly NO REASON to modify the part-throttle or non-VTEC maps for a car with light mods that idles OK. The stock ECU (OBD1 or OBD2) is entirely capable of handling most bolt-ons without any trouble. You should only modify the part-throttle and non-VTEC maps if you can't get the car to idle right or have part-throttle lean/rich issues. I have yet to see an NA Honda do this, it is usually only turbos or vehicles with injectors much larger than stock.
This is where OBD2 becomes a large problem, is with the LTFT and the VAFC trying to "fight" the ECU. Even with OBD1 there is no sense in not letting the ECU do it's job.
You can modify part throttle with a VAFC, but you really have to know what you are doing. I am currently preparing a writeup on this (offline for now), but in particular the Prelude ECU goes open-loop after 30% throttle. There is a a reasonable amount of driving time spent in that domain, and proper tuning there might save you a little fuel. If you are looking for significant changes though, that is not where the juice is.
Generally, tuning part throttle with a VAFC is not possible unless you add a little switchover circuit. The circuit is simple really, and I am building one right now actually. I plan on including a description in my writeup. I'm not going into all the details right now, but I will when I do my writeup. Todd is essentially correct about the LTFT problem. Do a search, and you can find some more information about it.
__________________
I am not an active moderator on this site anymore. Please contact Phryxis or liquidxnitez for help with your issues.
I will argue that 30% (thought the actual number was 33%) is quite a bit of throttle. Even in a 1.6L Civic with **** for power I never cross 30% unless I am going WOT. Depends a lot on driving habits, though.
One important thing to note is that the wide and narrow throttle settings (what the VAFC considers wide and narrow) should be set to mimic the ECU (past 30% is wide). If you set these at 10% and 50% (default values) then the VAFC will interpolate where you ordinarily may not want any correction. WOT should be set to what WOT is on the ECU so that you don't have any gaps where the ECU goes from closed loop to open loop (ECU and VAFC go to WOT maps at the same time).
Artifex, what are you talking about with a "switchover circuit?"
great reading...
what is LTFT?
Todd, about the wide and narrow throttle settings, not sure if i understand this correctly... each car has different points of NT and WOT and u adjust them of vafc to match the particular car's ecu? am i even close?
Artifex, i'm really looking forward to your writeup.
I was told that screwing around with the vafc without wideband o2 or a dyno might cause some engine damage, so i'm wondering if i get A/F gauge, can i use it to adjust my AF ratio?
Long Term Fuel Trim. It's the ECU learning/applying changes based on data from driving over multiple cycles (what clearing the backup memory/power clears).
Each car (meaning make/model) can have different WOT points. They will run closed-loop (watching the O2 and making adjustments to hold the AFR at stoich) up until they reach a threshold value where they will go open loop (fixed fuel map). For Hondas, this value as far as I have seen (both OBD1 and 2) is around 33%.
The VAFC and the ECU need to be in synch otherwise you will apply WOT correction to the ECU in closed-loop, or let the ECU run a fixed (WOT) map before you apply WOT correction to it with the VAFC.
You CANNOT tune a VAFC with a narrowband O2. Period. All it will tell you is if you go critically lean or not. You need to get a wideband O2 or do it on a dyno with one. I would not recommend tuning on a dyno without one.
Anyways in response to your original question preludes run rich out of the factory...in 99% of tuning activities you are leaning out the fuel aka injecting less fuel. So your gas mileage will increase as you are not feeding as much to the engine, and less is wasted on incomplete combustion
One thing however, most of our driving (at least for me) isn't with my foot trying to push though the floor. AKA cruising. At this point Honda engines tend to run lean during these conditions. Also, due to our OBDII ECU it corrects any changes (if not made for open loop) that you put in so part throttle fuel changes will be corrected for quickly and you will probably not see any better gas mileage by installing a V-AFC.
Does this mean VAFC wide throttle setting should set from default 50% to 33% before tuning or else the number come out is not optimize?
Quote:
Originally posted by marcucci I will argue that 30% (thought the actual number was 33%) is quite a bit of throttle. Even in a 1.6L Civic with **** for power I never cross 30% unless I am going WOT. Depends a lot on driving habits, though.
One important thing to note is that the wide and narrow throttle settings (what the VAFC considers wide and narrow) should be set to mimic the ECU (past 30% is wide). If you set these at 10% and 50% (default values) then the VAFC will interpolate where you ordinarily may not want any correction. WOT should be set to what WOT is on the ECU so that you don't have any gaps where the ECU goes from closed loop to open loop (ECU and VAFC go to WOT maps at the same time).
Artifex, what are you talking about with a "switchover circuit?"
Originally posted by Snoopy_cool Does this mean VAFC wide throttle setting should set from default 50% to 33% before tuning or else the number come out is not optimize?
Originally posted by pIOUs
33 instead of 10 is how I understand it.
There is still a problem with this though. The VAFC has the annoying property that any TPS value below the Lo threshold gets corrected at the Lo threshold. This means that if you set your Lo side to 30%, everything below 30% gets set to the same value. This is also the region that the LTFT will correct, so this is a no-win situation for OBD2. Currently, I have to set my Lo throttle at 30% with a correction of 0% across the board, and my Hi setting is 31%, with my WOT corrections accordingly. This is very non-optimal, and a real pain in the ass. I am working on a writeup which explains all of this, but I haven't had time to work on it lately. Combine this with the fact that you need all these extra tools and whatnot to make it work leads me to believe that it is just not worth it in the end. The other people who would benefit from this mod would probably all have the means just to go Uberdata.
Today I just decided to go P28+uberdata. The only thing I will miss is the immobilizer, so I am still trying to find an acceptable way to compensate for that.
__________________
I am not an active moderator on this site anymore. Please contact Phryxis or liquidxnitez for help with your issues.
Does this mean that you should set the low point to be 33% and the high to be 34%?
__________________
Black 2001 5-spd
I/H/Hi-flowCat/E/P/VAFC/DC Short Shifter
Eibach Prokit/KYB AGX/Neuspeed front strut brace/Tanabe rear strut brace/Suspension Techniques rear sway
Energy Suspension Motor Mount Inserts
Originally posted by axio Does this mean that you should set the low point to be 33% and the high to be 34%?
I can't say. It depends on what your part throttle corrections need to be. You would have to datalog your A/F ratios at part throttle (30%, 50%, 70% sort of thing) and determine how far off you are from where your target A/F is. If those are off by the same margin as your full throttle values, then you can set the Hi setting to 31% and get constant correction for everything above that.
I can't tell you a blunt yes or no, it is very specific to each car. With mine, all of the part throttle values were very rich (and still are), so 31% makes sense for me.
__________________
I am not an active moderator on this site anymore. Please contact Phryxis or liquidxnitez for help with your issues.
That's odd. I have yet to see a Honda (incl. H22) to run rich at all part throttle. OBD2 is the tightest, running right around 14.7 +/- .5 when closed-loop. Open loop it does go stupid rich.
Eh, I run pretty rich at part throttle too. Sometimes puffs of black smoke exit the exhaust even at part throttle. Marcucci, what do you suggest for the Hi Throttle point percentage then?
__________________
Black 2001 5-spd
I/H/Hi-flowCat/E/P/VAFC/DC Short Shifter
Eibach Prokit/KYB AGX/Neuspeed front strut brace/Tanabe rear strut brace/Suspension Techniques rear sway
Energy Suspension Motor Mount Inserts
Originally posted by marcucci That's odd. I have yet to see a Honda (incl. H22) to run rich at all part throttle. OBD2 is the tightest, running right around 14.7 +/- .5 when closed-loop. Open loop it does go stupid rich.
That is what I mean actually. In my brain, part throttle == 30% to 70%. I look at it on across the entire scale. Less than 10% to me is cruise territory, and 10-30% is street acceleration. I think it is just a terminology issue. Just ignore me.
__________________
I am not an active moderator on this site anymore. Please contact Phryxis or liquidxnitez for help with your issues.
Interesting experiment: I played around with the mixture on my 95 Civic (D16) that I just slapped a cam in (idling a little low). Instead of playing with the idle speed adjust I played with the fuel mixture. Low and behold the ECU would correct any idle mixture changes within about 5 seconds. I'd adjust it and the ECU would wrangle it back to 14.7 just like that.
For all asking about throttle settings, keep in mind that the VAFC won't do ANYTHING below the "LO" setting. It applies the part throttle settings to any throttle input between LO and HI. HI is where it applies high throttle settings.
For 99.9% of people (probably anyone asking these kinds of questions on a forum) I would suggest NOT making any "LO" (part throttle) corrections, setting your "LO" to 30% and your HI to 34% or thereabouts. Basically, I would only make WOT adjustments and set the WOT throttle point ("HI") to where WOT is on your ECU... which is 30-34% or higher on all Honda ECUs I've seen.
Originally posted by marcucci For all asking about throttle settings, keep in mind that the VAFC won't do ANYTHING below the "LO" setting. It applies the part throttle settings to any throttle input between LO and HI. HI is where it applies high throttle settings.
This is actually not true. This stupid behavior is the reason for all of my OBD2 issues. If you look at the VAFC manual page 35, you will notice that Apex'i shows a graph at the bottom of the page. That graph basically says that any corrections made at the Lo setting will be applied to all throttle positions below the Lo setting. I have confirmed this behavior using my Auterra datalogger. The LTFT will change according to the same percentage as whatever I put in the Lo category.
__________________
I am not an active moderator on this site anymore. Please contact Phryxis or liquidxnitez for help with your issues.
So if anything below the Lo setting has a -1 correction, should we just go and change the narrow throttle graphs to be +1 across the board? That way 33 for low and 34 for high sounds fine. This has been confusing since I first started setting up my VAFC.
Originally posted by pIOUs So if anything below the Lo setting has a -1 correction, should we just go and change the narrow throttle graphs to be +1 across the board? That way 33 for low and 34 for high sounds fine. This has been confusing since I first started setting up my VAFC.
It's very confusing. You really shouldn't touch anything on the Lo side. Just leave it at 0. If you change it to anything else, the LTFT will just dial it back out again.
Here's an example:
Let's say I want to tune my car for low throttle. Since anything less than 31% is always determined by closed loop (=14.7 and nothing else), the first throttle position I have to change anything would be around 35% (using some round numbers). So let's say, at 35% throttle, I do a datalog with my WBO2 and see that I am running 10% rich across the entire RPM range (which is pretty much true for me right now). Let's also say that I have had my WOT runs already tuned, and all the Hi settings have been set already for max power on the dyno (yay!).
So, common sense says that I should go into the VAFC and set the Lo value to 35%, and then go into the Lo tables and pull 10% across the board. However, this is not going to work. The reason is as follows:
The way the ECU calculates fuel in open loop is the following:
[VAFC setting alters MAP] -> [table lookup based on RPM and MAP] x [LTFT] = injector duration (fuel)
In closed loop it also checks the O2 sensor and keeps altering the LTFT until that value results in 14.7 AFR. Altering any one of the 3 parameters: MAP signal, base ECU fuel table, or LTFT, will change the fuel value.
Due to the behavior of the VAFC, it is now going to pull 10% fuel across the whole RPM range at 35% throttle and everything below (see post above for reason). This means that when I am now cruising at 10% on the highway, the ECU will be pulling 10% of my fuel because of the lower MAP value it sees from the VAFC (refer to the equation above). The ECU will notice that the base fuel table, when multiplied by the LTFT (see this post for details), is now running too lean. Over the next 20 mins, the LTFT will go up by about 10% in order to get the closed loop fuel value back to it's correct value of 14.7. So now the LTFT is 10% higher than it was before, and that value does affect open loop, which is what we just tried to tune.
So now, at 35%, I am back where I started at 10% rich again. When I first made the adjustment, the LTFT was at some value (usually at around -10% for stock Preludes). Now it is 10% higher, and so the 10% that I just pulled a minute ago has been dialled out by the LTFT.
That is pretty much the best way I can sum up the problem. The cliff notes are: don't touch the Lo setting.
Advanced readers:
I know of a simple circuit that would bypass this problem. Basically just use a voltage comparator and 2 MOSFETs to make a simple solid state switch. When the TPS voltage goes above whatever 30% throttle is, switch over to the TPS signal output by the VAFC. For any TPS voltage under 30% throttle, use the stock TPS sensor wire. The circuit is simple, but at this point I have decided just to go with Uberdata because there is less hackery involved, and it is a lot more tunable. If anyone is interested, I can post a picture of the schematic. I am not really an analog person, but the circuit is simple enough that I think it is correct.
__________________
I am not an active moderator on this site anymore. Please contact Phryxis or liquidxnitez for help with your issues.
The AutoGuide.com network consists of the largest network of enthusiast-owned enthusiast-operated automotive communities.
AutoGuide.com provides the latest car reviews, auto show coverage, new car prices, and automotive news. The AutoGuide network operates more than 100 automotive forums where our users consult peers for shopping information and advice, and share opinions as a community.
[/URL]
Currently, I have to set my Lo throttle at 30% with a correction of 0% across the board, and my Hi setting is 31%, with my WOT corrections accordingly. This is very non-optimal, and a real pain in the ass. I am working on a writeup which explains all of this, but I haven't had time to work on it lately. Combine this with the fact that you need all these extra tools and whatnot to make it work leads me to believe that it is just not worth it in the end. The other people who would benefit from this mod would probably all have the means just to go Uberdata.
The only thing I will miss is the immobilizer, so I am still trying to find an acceptable way to compensate for that.
Linear Mode
