Toyota Celica Supra Forum banner

1 - 6 of 6 Posts

·
Registered
Joined
·
3,834 Posts
Discussion Starter · #1 · (Edited)
Here is information regarding the split second installation and requirements, in a single thread, rather then over 10 pages of information and non relevant gab. The unit is installed in a 1983 Toyota Supra. The car is completely stock 5 speed 5M-GE engine. The only modification is the SS unit and K&N cone air filter for the unit.

Note that all of this information is in this forum, and that I will try to give credit where credit is due.

Background:

Split Second Website : http://www.splitsec.com
Part # PSC1-005

This unit was originally made available through a group buy by Supra Bob. For under $300 you could have a unit that would allow you to use any 0-5v MAF from Ford, GM, Toyota, etc as a higher flow upgrade to the smaller stock AFM. In addition you get fuel tuning, SAFC style, through windows software on your laptop. The MAF advantage is that once your calibrations are made, the MAF will account for various air flow conditions and mod changes, just like the AFM would. Finally, the larger air flow MAF, along with the fuel tuner, would allow injector upgrades, to a point. Jason Moorman had the unit installed at Split Second, they gave him a base map and most use this as the basis for their installations. The unit only gave a slight performance increase feel as expected on a stock 5M. Here is the original lengthy thread:

http://www.celicasupra.com/vbforums/showthread.php?t=5

Jason tried the unit again with 5M turbo and had even more problems. He finally gave up and sold the unit to me. Others have had throttle transition issues with the unit, there is a discussion on why there are problems here:

http://www.celicasupra.com/vbforums/showthread.php?t=12478


Installation:


What you need:

PSC1-005 Split Second Unit ($389)
MAF of your choice (Here I am using a 3" Ford MAF, supplied by Split Second)
MAF harness (plug into MAF with wires coming out)
Wiring Diagram by SupraWes
Air temperature sensor (optional)
Starter Software MAP (link coming)

Explanation of wiring:

Red - 12V Tap into the OEM harness
Black - Ground Tap into the OEM harness
Violet - AFM signal from Converter Box to ECU - Tap into the OEM harness
Gray - Generic Air Tem signal - optional - Tap into the OEM harness if not using air temp sensor
Yellow/Black - RPM - Tap into the OEM harness
Green - MAF input signal to Converter Box - This will go to the output signal of your MAF
Blue - Fuel pump switch - This activates the ground to the fuel pump relay - Tap into the Green wire on your AFM line (E1)

Ford MAF A - +12v
Ford MAF B - Chassis Ground
Ford MAF C - Ecu ground
Ford MAF D - 0-5v output signal

Air temp - One wire goes to THA, the other to E2, easiest to tap into the original AFM harness

Notes:

These do not work on 82 Supras! Search cs.com for more information regarding them.

After installation, the original AFM harness will be left not connected and the new signal will pass from the new MAF, into the converter box, and to the ECU. Make sure you secure the old harness from moving around. If you need to return to stock you can simply remove the new MAF and re-install the AFM. It would probably be wise to cut the power to the converter box if you do this, just to avoid confusion. I would also suggest putting a fuse between the converter box and ECU to avoid frying everything if something happens.

Keep in mind that the MAF is much larger then stock (if using 3.5") and that you run the risk of running lean since the ECU is seeing much less air then is actually going into the engine. You should add in a air/fuel gauge and monitor the o2 sensor output at the very least. At best you should invest in a wideband o2 sensor and get REAL information regarding the air/fuel. I am not responsible for what you do to your car, I am only providing the information that I have found in my instance.

MAF options:

Look for ford/GM MAFs from the mid 90s on ebay. I've seen them go from 10-100 dollars. A 3" MAF would probably be best for a non turbo since larger MAFs could make you run lean. Most find that you can get the MAF and connector from a junkyard for about $50-75. Also, you can buy a NEW one from Autozone/Kragen for about $120+ core charge.

I am using Ford Part # F6SZ12B579AARM 3" MAF, 96-97 Couger/Thunderbird

Tuning:

Once your unit is working, you can adjust the voltage that is reported to the ECU. To do so, just change the number from your base map. There are many fields, so it will be best to change ranges of fields rather then box by box. To see what the different numbers mean, see the voltage offset chart listed here:

http://splitsec.com/products/psc1/PSC1005.htm

From what I can tell, it will modify the input voltage of the MAF. The MAF to AFM conversion factor always seems to remain the same. However, what the MAF reports to the converter box is what determines the final output. I found that the car will still run even with no MAF installed.

You can't update the MAP while the car is running, and any changes you make become the new MAP. Unfortunately, there is not an option to "save as" to have many different maps. If you want to tune on the fly, it is best to use an SAFC or external adjusting device. Participate in the fuel tuning discussion at the thread above for more information.

Diary:

8/2/05 - Installed unit according to diagram, did not add temp sensor at this time, started up on the first try. The box was reporting strange information and would not rev up properly After trying a few things, I ended up re-writing the information to the ECU through the laptop and it started to work properly. The car idles like stock and revs perfectly. My initial fear by using Moorman's maps was that I would get a throttle hesitation. SupraWes also noted a hesitation over 3000 RPM in 1st-3rd gear. My initial test drive shows a smooth operation from idle to redline. However, I can smell the fact that it is running rich. The idle rev is smooth, but under load it seems soggy and running rich. Tomorrow I will put on the wideband o2 and go for a test drive with the stock AFM on and with the MAF on to compare the results.

8/3/05 - Installing a Zeitronix Wideband o2 today to see what the stock A/F and the new MAF A/F is in comparison. I also have made a new map that is 10% less fuel to see if it will make any changes in WOT. I'm not so familiar with N/A tuning, but it seems that 12.5:1 - 13:1 is what you want to shoot for, but not more lean then 13:1. Considering the low compression nature of the 83 engine, I may try for 13.5, but I don't want to push it. Colder spark plugs will be put in before I start trying to go lean.

8/3/05 - Initial wideband results show LEAN. It seems the soggy WOT experienced was not richness as first expected. Not only that, but the low rpm range is rich. Can't get much worse then that. Here are the peak/lean results for WOT 1st - 3rd up to 5000.

1st - 15.20
2nd - 16.90
3rd - 18.20

Something isn't right. Before I do anymore runs I'm going to have to make sure that this unit is converting properly. It almost seems to be working backwards. As a test, I added 20% to all the values, and it made it so lean that the car wouldn't even run. I took 20% away from all the values and the same thing happened. So, I'll have to do some bench testing, or install the AFM in line with the MAF to make sure that their outputs are similar.

8/4/05 - It is possible that something happened to the original map to cause a problem. When I started fresh, I was still lean everywhere and full lean upon revving. I started richening the low air maps plus 20% and they became normal. I then decided to add 20% to the ENTIRE map and it all came back. Now the car is pig rich on the wideband (10.0:1) at all times, BUT I can drive around hit WOT without running lean like the MAP above that I started with. I believe that it might have been tuned for bigger injectors. The next step is to record a wideband run with the stock AFM and use that as the guide for what I want to shoot for. Then I will start leaning out the upper range map by 5% until I get closer to the stock map. I won't mess with the lower end so much, since the stock o2 sensor will take care of it, once re-installed. Here is the map if you want to take a look at it. Use at your own risk!!!!!!!!! This might be worth a try on a 6M which someone reported was running too lean with the SS. Also note that partial throttle will hesitate due to the richness, but WOT should pull smoothly.

http://www.935motorsports.com/mk2supra/fuel maps/


Later today I took 5% out from the whole map, too lean! So I tried 2% out, too lean! I took 1% out from the whole map, and now we are driveable, but lean in some spots. It seems that the percentages are not like SAFC. Just the slighest adjustment messes up the conversion quite a bit, so it looks like we have to deal in 10ths of a percent. I left the map at -1% and recorded a SS run and a wideband run. Here is some of my findings:

For the lower gears, the MAF records a max of 2.8V and about a 13 a/f. Once in third the flow really picked up to 3.5 and sent me into the lean. Luckily with the wideband recording it tells me the RPM, MAF voltage, and a/f ratio for that spot. The SS map goes by RPM an MAF voltage so i can go right to that cell range and add fuel as needed. This will take some time, but eventually through trial and error I should end up with my good MAP at 12.5 a/f all through the RPM range. Some of the MAP is already good around 12 and 13, so I'm tossing fuel in the higher flow portions to get that solved quickly.

later that day - I have edited my MAP and added .5% fuel in the spots that I found lean. Also, I find that the MAP as is does not richen for RPM. Meaning that if you have more RPM at the same MAF voltage, it starts to run lean. This would be say a small hill or something where you have the throttle part open but the RPMS are increasing. So I added .1 cell value higher in all the cells above 3000 RPM. Finally, I added .5% anyway to everything over 3.5V and 3000 RPM. My new map is called -1 plus 3000. I'm going to do another run later and if things are working how I think they are, I should now be less lean in the higher MAF voltages.

even later that day - My map is coming together. After adding fuel to the mentioned spots, I still found a lean spot around 3.2-3.4V MAF and after 4500 RPM, so I added yet more fuel to a large area there. If you look at the map I'm starting from, it doesn't make sense that it gets leaner as you go up to more voltage. I think this is the problem. Another run showed no more major lean spots, but still lean. My values are 15 before 3.2V and then 14.1 after 3.2v, which would mean taking out fuel. That makes no sense. Ive changed all the 14 values in the upper end to 15 from 14, adding in fuel. I expect that i will need even a little more fuel in the upper corner, but I'll see where 15 takes me before I add anymore. My midrange a/fs are around 12-11.5, rich but at least driveable. My low air a/fs are around 10 and it makes causes bucking from being so rich. To be able to drive around I need to get this more lean. My initial cell values were around 14, most likely due to my 20% increase in fuel across the board as mentioned earlier. That is too rich. I went down to 10, which would be no adjustment, but that was far too lean. It looks like originally the map was going from about 9 to 12 gradually (enrichment), so I'm going to scale it from 10-14 and see how that works.

8/5 - I devoted the first part of today's tuning to the idle/rev maps. Even though I realize that the o2 sensor will adjust this portion, I think it is best if i come close to where it should be so that the ECU has less work to do. Previously my idle a/f was about 10:1, I want to get that down to about 14, and about 14 on revs. I may also put the air temp sensor on, since that could play a part in my tuning. I found another really lean spot at about 3000RPM and 2.2 to 2.8V. This is when you rev up the car high, WOT, without any load. I've been playing with the figures for about 30 minutes and I am getting somewhere, but it is very time consuming to go map by map. I'm sure the results will be worth it. Once I can get the low load maps in, I want to try a 3rd gear WOT run and see how the a/f curve is after these modifications. I sure hope it is on the rich side!! I changed all the high air flow maps to be really rich.

later that day - my new maps are very good, i took the low values from the 16s down to the low15s and the car idles and drives nicely. I'm at about 10.5:1 in the low air mode, but it's good enough to get away with driving. Driving around town even with hard acceleration gives me smooth pulls. I was happy until I decided to do a WOT power run. I ran lean in the upper reaches, so it looks like I need yet more fuel. Here is a brief list of my WOT a/f in 2nd gear

3500rpm 3.0v 11.9a/f
4000rpm 3.25v 11.8a/f
4500rpm 3.45v 14.7a/f
5000rpm 3.57v 17.3a/f

So you can see where the lean starts to come in. I'm ok until about 4500, then it goes lean steadily from there. Logging a/f is like magic! Now I just go to the 4500 and 3.4v box in the SS software and add fuel on a upward scale up to 7000RPM! When I go to that area I find figures of 16 all across the board for RPM and higher air flow. I remember going from 14 to 16 as a general enrichment across the top end previously. So it seems that you MUST have a steady increase of cell values as the air flow goes up AND as the rpms go up. This makes sense, as that is how a typical fuel map would look.

What I did first was changed all values over 4v to 18. Then from 3.3v to 4v I used the auto fill button. The auto fill button will take 2 figures, say 16 and 18 and fill the values increasingly across the range you select. It's much easier then going 1 by 1 and helps you smooth out the MAP. One thing that I find is that you don't want big steps in cell values, you always want a smooth transition.

So at 4000 rpm, where previsly I had 16 16 16 16 16 16 16 16 16 , it now looks like 16 16.3 16.6 16.9 17.1 17.4 17.7 18. Then I changed all the cell values over 7000 RPM to 18 and used the auto fill option to increase by RPM from 16 to 18. So I have 2 points of enrichment, RPM and airflow. Hopefully this will be enough to keep me from running lean. Eventually the whole map should look like this, but I'm concentrating on the problem at hand right now. Keep in mind that I have yet to go over 4v. If I added mods or a turbo then I'd probably be having to tune farther up to the max 5v. For now everything over 5v is 18, but that might prove incorrect later on.

Now, I also found that at 3500rpm and 3v until 3600rpm and 3.1v that I'm getting into the 11s of richness. Now, this is not so bad, but we want to tune for power and this is our first chance to try and do that. So I'm going to take a teensy bit of fuel out of that area and see if we can get it to 12.5. We might find that it is hard to tune to that high of a resolution, but we'll find out. I took down values in this range by .1 cell value, so we will revisit those spots later and see what effect the change had.

Next, I find from 2600rpm/2v to 3800rpm/3.2v that I go into pig rich 9.6 a/f then back to 12:1. My values were all flat 16 here as well. So in the reverse of above, i need a steady lean curve here. I changed this range to 15.5, but used auto fill on the edges of the change to try and smooth it in. I expect I might need to take out more then .5, but I'm going in .5 increments to prevent wild changes. We'll see what happens in the next episode. I'm pretty close to having a decent map. Once I can drive through all gears with no major lean spots I'll post the update.

8/6 - Today I tested my lastest map. Still lean at very high RPMs and still rich and part throttle. I logged the part throttle problem and took out a value of 1 in that area. WOAH! Super lean. The part throttle area is extremely sensitive. I added back in .5 value rich in that area, and the problem seems to have gone away. It's still rich, but not so rich that it causes bucking. In first to third gear, I have a solid idle and solid part throttle. Very smooth. I richened up the top end some more, up to 18.5 max (out of 20). 1st gear pulled rich, 2nd gear went lean at 5000 and 3rd gear was rich at 4500. Lapop battery died, so no more tuning until later. I'm adding a little fuel to that area of the map, I will log 2nd gear only, from low rpm to high rpm and focus just on 2nd gear at the moment.

That night - After the 18.5 increase, I am still lean over 5000. I'm starting to wonder if this is going to work on stock fuel. Either the MAF is too large or my injectors are not flowing properly, or the fuel pump is not keeping up. I have changed the maps to go all the way up to 20, max rich. If I'm still lean, then I will not be able to add any more fuel. I will probably have to add a fuel pressure regulator or upgrade the injectors to get more fuel in the top end.

Later that night - The car drives great and feels strong. HOWEVER, even with full richness (cell value 20) over 3.3v and over 3500rpm, the car still is lean.

1st gear, 5000 RPM, 3.65v, WOT, 18:1 lean
2nd gear, 4600 RPM, 3.5v, WOT, 17.9:1 lean

So the conclusion is that I can't add enough fuel with the SS unit to compensate for the larger MAF. To verify that the MAF is at fault, I will install this setup into another 5M MK2 and see if my results are similar. If the problem exists again, then it's the unit. If the MAF works better in the new car, then I know that my current test vehicle has a fueling problem that I must address.

Ford 3" MAF map is saved until I work out fueling. I will now move on to the Supra TT MAF installation.



PLEASE DO NOT REPLY TO THIS THREAD. USE THE LINKS ABOVE FOR DISCUSSION!!! THIS POST IS JUST A REPORT OF MY FINDINGS.



Credit list of those who have provided wonderful information:

Supra Bob
Moorman
SupraWes
 

·
Registered
Joined
·
3,834 Posts
Discussion Starter · #2 · (Edited)
Stage 2: Supra TT MAF installation

A few folks were interested in using the TT MAF for the Split Second conversion. Considering most of us know MKIV Supra guys, it is easy to pick one up for a few dollars, or in my case, free.

My buddy recently went AEM, so his MAF was just laying there. I picked it up today. The bonus of using this MAF is that it is smaller then the Ford, so hopefully I can get it to work with stock injectors. Even if I can't, and it still requires a fuel upgrade, I think I might rather have the Supra MAF in for these reasons:

- It's a Toyota part, makes me feel better
- I could possibly pass smog with this installed, since it looks stock
- It has a temperature sensor built in, so no extra sensor is required
- Great for NA or turbo, and easily replaced

Now on to the wiring. Here is a picture of the Supra MAF. Looking at the MAF head on, we will call the rightmost pin, pin #1. The Supra TT factory harness colors are mentioned as well, for easy reference.

SUPRA MAF PIC

1 +B (power; 9 - 14V) (Tap to switched ignition source) (Red)
2 VG (O2, MAF Output) (Connect to Green Split Second Input Wire) (Yellow/Red)
3 THA (temp) (Tap to THA on ECU or THA Wire on AFM Harness) (Pink/Blue)
4 E2 (Tap to E2 on ECU or E2 on AFM Harness) (Brown/Black)
5 E2G (ground) (Connect to any Ground Point) (Brown)

See the Wes diagram above for where you can tap the existing AFM harness. You could cut your harness and solder in a Supra TT harness, but it's best to leave your stock harness available, in case you need to go back to stock.

8/8/05 - Supra MAF installed, using the Ford MAF MAP as a base. This MAF idles around 1.5V instead of the .9V of the Ford MAF. So if you are looking at a Ford MAF map, you must instantly lean out the lower end sections. It was so rich black smoke was shooting out. I got it tuned in to be driveable, leaning out some sections. However, the top end was still lean and in fact showed less voltage at WOT then the Ford MAF. This means that the Ford MAF flows more air then the Supra MAF and the engine accepted this, meaning that the Ford MAF would be a slight upgrade from the Supra MAF.

10/20/05 - Turns out the fuel pump was bad and wouldn't deliver fuel in the RPMS. I have installed a new oem pump and now am ready to re-test.

o2 voltage reading: If you do not have a wideband handy, but want an idea of where your a/fs are, you can use a a/f guage. Most have 10 LEDs, some have 20 or 30. The more LEDs the better. A recent test of a single wire o2 on a NA GM motor showed the most power at a voltage of .76 or higher on the o2. To add a little bit of safety, let's say .80. That would be the 8th LED on a 10 LED scale. So we don't want the o2 reading to drop BELOW that during WOT throughout the RPM range.

My current reading is over .98 at WOT, insanely rich for NA. I will continue to pull back fuel in the SS software until I can get to around .8v. After that, I will install the wideband again to see what .8 would correspond to as far as a real A/F.

If you are running a boosted setup, you don't want to drop below .91v. A .92 reading is nice and safe and .95 is generally what most stock turbo setups are at. While some do run a .90, it is not suggested. In our case, think of .90 as a danger zone.
 

·
Registered
Joined
·
3,834 Posts
Discussion Starter · #3 · (Edited)
Stage 3: 295cc


11/5/05 - I have sinced replaced the fuel pump with a working stocker. I also have 7mge 295cc injectors installed. Everything else is stock. I started with the Ford MAF and a stock MAP. A quick drive shows rich on all accounts, WOT, idle, cruising, etc. I took out 5% from the entire MAP and the car runs lean at part throttle, low air flow. The MAP is extremely sensitive in this area. The air flow equation is very narrow and even the slightest adjustment can upset it.

11/7/05 - I now have moved back to the Supra TT MAF. I finally decided that I wanted a Toyota MAF under the hood. It has the air sensor built in, looks stock, and could possibly pass smog. Big if on that one! So I started at zero again. The smaller TT MAF will show more air flow then the FORD at any given moment. So with the stock SS map it was incredibly rich. So much so that it would barely run. After a few adjustments, I got it to play nice. I pulled out -20 at idle and -10 at low air flow higher RPMs. However, there is a certain range where I can only pull out about 5% or less or it is too lean. This is the same problem as shown above. This area is real touchy for some reason due to the hard coded conversion. So with -20 idle, -10 higher RPMs and -5 in this mystery range, I went for a drive. I left all the upper air flow maps as supplied. The car ran very smoothly, though did not seem as strong as the ford MAF. Probably due to being more rich. There was a certain range that was so rich it hesitated. I need to log this range and tune it down. The top end pulls were very smooth to 6k and pig rich. I'm going to start pulling fuel from the top end -5% increments until i can get it at least out of the pig rich area. So far the results are encouraging. Once I can get a good driveable map I will post it for others going the TT route.

11/8/05. The 1.6 to 2v area of the split second conversion map is very touchy. I found that a certain area of the map is very sensitive. It is the area where you have low rpm high air flow, such as a throttle tip in. I have noticed that other MAF conversions like the MAFT popular on 7mgte engines, have this same problem. For example, 1500 rpm at 1.6V is is measured at cold start up (high idle). This runs rich. However, in 5th gear 1500 rpm 1.6V, this runs lean. The SS unit does not know the difference, so basicall you can't tune for idle very well, you have to let the ECU take over there.

All the fuel that i had taken out at idle, I had to add fuel in those same idle spots so that throttle tip in (say less then 10% throttle), was too lean. Finally, the car is driveable without noticeable lean spots.

The stock map supplied from SS, goes up to zero correction at one point, then starts leaning out in the top end. This doesn't make sense to me. You would think you would want it even at the top. Anyhow, I found that even with 295 injectors, the stock map cause it to run leat at times at the very top end. This is a good thing since we are trying to make it lean at the top for big injectors. However, we don't quite want it that lean. So I took the highest # in the map and instead of it going down in the top end, I just made it flat from the highest fuel number, across to the end. This means that at a certain point it just stays at slightly less then zero correction. After this change, I was able to pull to the end of each gear (1st through 3rd), without a lean condition in the top end. I'll have to make more pulls to decide if this is working right. If I can pull from 1st through 5th without going lean, then we have had a success in the conversion. At that point I can start taking out fuel and tune for power.

I'm about 75% smooth in my map. I have a slight lean condition, though not a big deal at a certain rpm. I also have an extreme rich condition at another RPM. Only driving and making changes will solve these. I'm also undecided about the top end fueling. I want to make a nice long 4th gear pull with logging, so that I can find out what the real story is.

In a few more days I'll have my map to post up.
 

·
Registered
Joined
·
3,834 Posts
Discussion Starter · #4 ·
11/15 -

More dial in. I had a lean spot at throttle tip in. My correction was going from 9 to 13 in the next cell. I have learned that you can't jump numbers like that. It makes drivability terrible. You have to use auto fill function. If you select a range that starts with 9 and has 13 at the end, it will average the numbers from 9 to 13 all the way through the selected area. This makes for smooth transitions in correction.


Top end - I'm trying to work with the top end. I took out 5% correction from the top end and ran too lean. I changed to 4% and still too lean. I went back to normal and it's pig rich. The adjustments are very sensitive in this software. You have to deal with 10ths of correction.

Idle issues - It happens that idle air flow is measured the same as cruising air flow. This means that if I dial in the idle from being too rich, the cruise is too lean. If I richen up the cruise, then the idle is so rich it barely idles. The software does not know the difference between moving and not moving. Hopefully I can find some compromise. I think I will be moving to a heated o2 sensor to make the closed loop operation work at maximum.
 

·
Registered
Joined
·
3,834 Posts
Discussion Starter · #5 ·



The supra MAF is now hardwired. Meaning that the supra TT AFM connector has replaced the stock afm connector. I made the wires a little longer to allow movement of the location if needed.

To wire it in full time, it was pretty simple. The original VC wire (voltage reference) was 5 volts from the ECU. This was cut at the ECU, and connected to the 12 volt ignition power. This wire from the AFM harness goes to the 12v power on the MAF. VS, the voltage output from the AFM, was disconnected at the ECU. The ECU end went to the SS unit, and the other end went to the MAF input on the SS. On the AFM plug end, the VS wire went to the the MAF output on the TT harness. The 2 grounds and THA, the air temp signal, were connected like stock. So now I have full air temp correction, like stock.

The only wire yet to be figured out is FC, the fuel pump jumper that activates the pump on the original AFM harness. There is no corresponding wire at the ECU, so I have to find it.
 

·
Registered
Joined
·
3,834 Posts
Discussion Starter · #6 ·
Latest update:

W/ 295cc injectors, the tune is somewhat dialed in. 2 problems remain. The car takes a few extra cranks to start when cold. There is also an off throttle slight hesitation. I believe that is is due to the fact that I no longer have a wideband to dial in the final tune.

Otherwise it works as stock with the Supra TT MAF. Unfortunately I will no longer be tuning the setup, since I don't find that it is adding any more top end power. It does offer great throttle response, but that is all.

I may next try to convert to map sensor using the SS.
 
1 - 6 of 6 Posts
Top