Even though I have a lot of experience with FlashPro tuning, and have created and tested over 700 different calibrations, I decided to get an eTune a few weeks ago. Part of the reason for this was that I was getting a check engine light for misfires at idle, but the primary motivation was that I hoped there would be some “magic” fuel or timing curve that would give me some extra power. Basically, it seems that tuning is somewhat of an "Art", but I'm not artistic at all. Professional tuners are so tight-lipped about optimal ignition timing, so I figured there was some secret to it.

There were several eTuners that I thought about, but since it seems that Evan’s tuning gets so many raving reviews, I decided to go with him. He was more expensive than the others, but I figured it might be one of those “get what you pay for” kind of things.

Here is my modification list:
PWJDM Intake
630cc Siemens Deka Injectors
Hondata Insulating Manifold Gasket
Custom 3” Exhaust

The eTune process started off as I expected. He sent me a base map with a list of datalogs to send back. Unfortunately, the ignition timing in the base map was way too aggressive, and caused audible knock. I’ve never heard my car knock before, and have never had nearly as many knock counts in my datalogs. I decided NOT to do anymore datalogs, and reverted back to my previous calibration until Jeff sent an adjusted calibration. This really concerned me, since I knew that it would have been very bad to continue driving the car with that base map, so I really wonder how many people know that they should always revert back to a “known good” tune between eTune revisions. Honestly, I was surprised that there weren’t any recommendations about what calibration to use between revisions.

The knock counts were corrected right away, so I started doing the requested datalogs between each revision. I guess there were about 3 revisions per week for a couple weeks. Unfortunately, the misfires never did get resolved, but Evan’s said that Hondata plans to make the misfire sensitivity table adjustable to deal with this.

Now for worse news. After all the tuning was done, my full throttle air fuel ratios are WAY TOO RICH! Jeff believes that the AFR needs to be 13:1, and uses the FlashPro's "AF Corr" value to achieve it. Unfortunately, the AF Corr value is not accurate. I’ve posted proof of this on S2KI, as well as many others. The actual AFR that I wound up with was between 11.2 – 11.9! I emailed Jeff about this, and told him that my car had been on Dynojets before with external wideband sensors, and the non-corrected AF value was always MUCH closer to the wideband reading than the AF Corr value. Unfortunately, he just replied saying that he’s tested it before, and he believes the AF Corr value is more accurate. I’ll discuss this is more detail below with dyno graphs, charts, calculations, etc.

Now for the worst news. The datalogs that I sent to Jeff were exactly what he asked for (full throttle, stop and go, highway, idle, etc.). There was never a request for a datalog that included part throttle driving at high RPMs (ie. Part throttle in VTEC). However, every time he sent a new calibration back to me, the high cam (VTEC) part throttle fuel values always had negative adjustments (-3%, -7%, -7%, -9%). When I looked at the datalogs to see if there was anything in there that could possibly indicate a legitimate need to remove fuel from those values, there was nothing! After the eTuning was done, I created a datalog specifically to test these cells, and found that they were all WAY TOO LEAN. I drove in 4[sup]th[/sup] gear on the highway, and accelerated very slowly until VTEC engaged. As soon as it did, the AFR shot up to the high 15’s with STrim at +35%, then climbed above 16, then when it went to open loop, it shot up above 18:1 and the car started bucking. So I'm curious, how long would it take to burn up a motor at 16-18 AFR?

I emailed Jeff about this, and he sent a revised calibration with a note saying “You had bits and pieces that transient throttle swept across the high cam vacuum fuel map, and I based the changes on the small data I was seeing.” I already knew that was what happened, but was SHOCKED that Jeff would make that mistake! If you are going to do eTuning, you have to be able to figure out what data is valid, and what needs to be ignored! It seems that Jeff is relying on the automatic fuel change suggestions in FlashPro Manager way too much. I’ll discuss this in more detail below with graphs, charts, etc.

After everything was done, I did some very controlled back to back Soft Dyno runs to see how the eTune compared to my own tune. Unfortunately, the eTune didn’t make as much power or torque as my own tune. To make absolute sure, I did multiple pulls with each tune back to back within 10 minutes of each other, then repeated the test again the next day. The graphs below show the average of 3 pulls each from both tunes.

Test 1 using average of 3 pulls of my tune (blue) and 3 pulls of eTune (red). NOTE: The torque lines are the solid lines, and the thin dashed lines are the fuel injector pulse widths.





Test 2 next day using average of 3 pulls of my tune (blue) and 3 pulls of eTune (red)



Here is my AFR at full throttle



Here is the eTune AFR at full throttle



To show that my O2 sensor is accurate, Here is the AFR that I got from when my car was almost totally stock (stock intake, stock cat, and stock exhaust) on a DynoJet a couple months ago.



I datalogged those same pulls, and averaged them below. Pretty close, right?



I understand that there might be S2000’s with O2 sensors that aren’t accurate, but that doesn’t mean that all S2000’s need to be tuned using the AF Corr value! It is not difficult to test the sensor! Consider this simple test, which any eTuner can perform. If Jeff sent 2 calibrations to be tested back to back, and one of the calibrations had 10% more fuel at full throttle, then he could compare the 2 WOT datalogs to see how accurate the stock sensor is. If the AF value shows a 10% difference, and the AF Corr value shows a 6% difference, then the AF value is more accurate! If neither AF or AF Corr show the 10%, then spend a couple minutes calculating an adjustment manually. It’s just math!

Here is a test I did with my tune to illustrate it better. I made 3 calibrations to test back to back. The calibrations were all identical, except that one had a -5% fuel adjustment at full throttle, and the other had a -10% adjustment. The first calibration (blue) wound up with an average AF of 11.9. The other 2 calibrations gave average AF values of 12.43 and 13.13 respectively. The average AF Corr values for the same datalogs showed 13.29, 13.6, and 14.01 respectively.

AF Graph



AF Corr Graph



If you do the math, the AF graph shows a -4.5% and -10.3% difference between the calibrations, which is very close to what I expected. However, the AF Corr graph shows a -2.3% and -5.4% change. It’s pretty clear that AF Corr is not accurate at all!

Just for giggles, I decided to see how the WOT target lambda table would work. I used my regular tune, which has a pretty flat AF at 12.5, and also has a flat 12.5 WOT target lambda. I changed the WOT target lambda so that it would be 13.0 at the lower RPMs, then leave the middle RPMs at 12.5, but the higher RPMs got a value of 12.0. The AF graph below shows exactly that.



Now back to the part throttle tuning. I’m really surprised that an experienced tuner would not be able to tell that there’s a problem with a fuel map just by looking at it. Here is a screen shot of my own high cam fuel map, which looks very similar to every “canned” tune in FlashPro Manager. Notice the relatively even distribution of lines.



Here is the same map from my eTune, which shows an enormous gap between load lines in the middle! He should have known this would cause a very lean condition at part throttle! (Note: I added the yellow highlighting to illustrate the issue)




The fuel suggestions that FlashPro manager gives are pretty damn good, but FlashPro Manager doesn't filter out quite as many datapoints as is necessary. For example, below is a screenshot where FPM suggests large negative adjustments just from a brief trip into VTEC. It doesn't take much experience to figure out that only the suggestions circled in red are actually valid.



However, when I exported EVERY DATALOG that I sent to Jeff, and examined them in Excel, I found that out of 1,555,000 datapoints, there were only 667 datapoints that could have possibly been legitimate for part throttle VTEC. (Fuel Status = 2 or 4, VTS = On, MAP < 750 mBar, INJ > 0) That's about 3 seconds worth of data split up between ~15 datalogs. I don't see how a qualified eTuner could repeatedly make the same assumtpions 4 times in a row, and assume that he should repeatedly subtract 7+% based on obviously crap data. However, this issue did prompt me to suggest a few things to Hondata to improve the fuel suggestions in FlashPro Manager. You can see that thread here.

 

Interesting. That fuel map blows my mind.

 

Interesting...I've only heard the best about Jeff. Im going to take my car to him once I get my hands on EMS.

 

I had Jeff E-tune my turbo setup and I really didn't have any problems per say. While i'm no tuner I know enough to check the logs for knock, lean conditions, to much timing, etc. Only problem we had was with the vtec adjustment. He intially had it set too low for my setup, but that was more of a miscommunication about my manifold type. I check the logs often and I've had no issues thus far. I'm sure he does the best he can. I wonder though, with the huge popularity of the e-tuning now, if he's taking on too many cars to tune at once. You know, maybe not checking the little things. I can't say anything bad about my experience though, and my car still runs great.

 

When I first started the etune I still had my AEM wideband installed, I noticed that it was reading closer to the uncorrected than the corrected up to the upperr RPM band (+7000) where the uncorrected seemed to be skewed, may be the corrected AFR accounts for that skewing, I’m not a 100% sure about this.
I also sent him additional datalogging of various driving conditions in order to make sure that all bases were covered.
Car feels great.
Cheers

 

Gernby, I have learned a ton over the years reading your posts and your findings. I am not surprised that another tuner could not find any more hidden power in your set-up (no offense to anyone). You have the time and passion to pour hours into your car - like most car enthusiasts. Would strapping onto a dyno have any different effect?

 

Hey Gernby....

I've had an e-tune done by E-Tunez....and recently had a wideband installed in my car. I have no interest in sides....just offering my own observations.

My wideband reads pretty close to the AF Corr value in most cases.

part throttle the wideband is about .2:1 richer than the AF Corr

WOT under 6k rpm is .5 to .65:1 richer than the AF Corr

WOT over 6k rpm it's about .2:1 richer than the AF Corr

I'm going through the process of re-tuning now with the info from the wideband. The good news is my tuner has been able to find a little more response in the lower rpm's, and a bit more power overall by getting to that 12.9 - 13.1 range. Part throttle I'm pretty steady at 14.55:1.

I don't know about all tuners, but I do know he's been able to get my car tuned very well in a fraction of the time it would take me to tune it half as well. I know you have spent considerable time on yours, and know more about doing it yourself than I. Appreciate that you've been willing to offer up lots of detailed info.

Hope this helps....& this is just what I'm seeing in my logs. My wideband is an Innovate LC-1 and it's wired up to the ELD pin to log into FlashPro.

 

Jeff,

I understand that you think that my motives are bad. My mother even says that I'm negative. However, I'm also a degree'd engineer who is an hourly contractor with crappy private health insurance, and have a stay-at-home wife with cancer and a 4 year old daughter with a budget that doesn't allow for me to blow $200 just for the fun of ripping someone apart. I promise that my intention was to get benefit from your eTune!

I also submit that 290 of the calibrations I've created and tested were in the last 10 months, and took way more hours than you estimated. I've also had my car on 3 different Dynojets this year, and all of those widebands read WAY closer to my AF value than the AF corr value. BTW, I forgot to provide the link to the test I did with my own wideband years ago here. Whatever facts you've collected about the AF Corr values are no more valid than my facts. They are all just datapoints. That doesn't change the "fact" that you should test the accuracy of the sensor in some way. I believe that my suggestion above is a huge gift to the whole eTuning industry. All it takes is another datalog!

Honestly, I believe that I've create at least half a dozen threads that I believe benfit eTuning (like this). I also developed the Soft Dyno, and provide it publicly for free. I was totally taken off gaurd by your statement that I've always been against you. I don't recall ANY occasion that I've posted anything negative about you! Of course, I'm not saying that I haven't posted anything negative about you ... I'm just saying that I don't recall anything negative. I truly regarded you as the "safe" choice!

Another thing that I should mention is that I truly did NOT want to post a thread ripping you or any eTuner apart. That's why it took me so long to post this thread! I've been debating what to do about it! I know that I could have told you up front who I was, and what I've done, and that would have changed the outcome. However, that would have eliminated the "intellectual value" that I believed I would get from doing the eTune.

 

I thought that a stock S2K had more power then that?

 

He put down 220whp on an edyno with basically an intake, exhaust, testpipe and tune. I think somewhere between 210 and 215 is where he would be on a dynojet research (the real dyno would probably hurt because the jdm scoop wouldn't function on the dyno vs the street).

A bone stock ap2 would put down maybe 200 to 205whp on the dynojet. Mustang dyno probably 5% lower.

Churchs dynapack around 235 to 240ish since it is calibrated to show flywheel hp

Sent from my T-Mobile G2 using Tapatalk