First Logs w/ Evo16g

Here's my first log--Evo16g, 17 psi, 560cc evo8 injectors, stock fuel pump, stock sidemount.

It definitely looks good so far. Yes, I would log some more data points so it would be easier to see any small dips in timing you may have (though it doesn't look you have many, if any at all).

right on man. looks a little flat from 3900-4800. nice advance!

looks a little flat from 3900-4800


That's kinda what I was referring to. I'm willing to bet that if you added more data points you'd see a small dip in timing in that area. Obviously small dips like that are not TOO big of a deal since 1 degree of knock retard is only ~3 counts of knock (1 count of knock = .35 degrees of timing retard).

thats right. try leaning out a little around 3900-4800. see if that helps any

God damn I'm good :lol2:


Not bad for a "Jake Tune By Ear" lol!

Finally got the logger workin huh Joe? Its about damn time, AGAIN... Thanks to me! :icon16:

Thats ghey apple for ya :p!

You want a drop/flat spot in timing up to ~4000-4500. Thats where cylinder pressure are highest (torque peak) and rpms are too low for much timing. I think the curve looks balls on. A little too clean actually, I would start leaning out or even better raising boost and see what happens. Just be sure to write these settings down since they seem to be damn good.

Well, let's not forget I had to turn boost down...

...even better raising boost and see what happens. Just be sure to write these settings down since they seem to be damn good.


Yea that had to have been the best quicky job I have done in tuning. It definately helps by working on these cars everyday! And you were saying everything I was doing was breaking your car... I see how it is, no love!

:lol2::rofl::nutkick:

"Want to go for a ride now? I'll peg the boost gauge for you."


Actually that was me that said that. :lol2:


The day that you boost to 20psi is the day "me shat turns purple and smells like rainbow sherburt!" :rofl:

Yea that was some good times, running from the cops, talking shit about hondas... lol He didnt really know what to expect I guess.

Joe you need to drive your car back over spring break! I'm driving down but its in Panama City Beach and its the weekend that your spring break ends on... Should be a good time man. Too bad you gotta be a douche and miss out! :lol:

You wont get every 200 rpm on an OBD2 logger I don't think. At best you're gettng 20 samples per second, divided by the number of things you are logging. You're best bet would be to log only RPM and timing. If there is a problem in the timing curve, then go back and log with O2. Dropping from 3 logged items to 2 items will increase resolution by 50%. Man, thank God for dSMlink. Datapoint like every .03 seconds, logging 15 different things :D

Datapoint like every .03 seconds, logging 15 different things

Yeah, yeah, rub it in why don't you? :p

Joe, Are you using the Pocketlogger software. If so, the max you will get is 10 - 12 s/Sec. Did you do the byte timing test? If you are at that data rate (10 - 12) you should have about 2x more points than you do. Like Kevin said, log only timing and rpm and then post that log.

I just played with the byte timing, and lowered the delays quite a bit--I'm reading between 10-12 on my data capture rate. The inter-byte delay is 3000, and the inter-packet delay is 14. How exactly does the test work? If I'm getting "error counts," am I suppose to raise my delay settings until I don't get errors anymore?

Are you using v1.02 or v2.0?

With v1.02 you have to set the parameters manually. If you upgrade to v2.0 (Which I like much better), the byte timing test is automated.

By the way, the upgrade is free to people that have bought v1.02

EDIT: To answer your questions you start with the default settings and work your way down from there watching the error count. Start with the larger number (I don't remember which one it is I think default is 2400?). You want to decrease that number until the logger starts registering errors. Then increase to where the errors stop. Your objective is to get the timing as low a possible with NO errors. After dealing with the larger number....move to the smaller and do the same thing. You need to do this with the car stationary. Don't do anything else (like rev the engine, turn the wheel, A/C, etc.). Next, leave the logger on the "byte timing screen" and drive around for about 5 minutes watching for errors. Be careful not to wreck :iceslolan . If you see any errors just bump one or both of the timing settings up 1 notch. If you don't get any errors you are done. It will take a little time so make sure you write down your settings when you are done. You should also hotsync after doing the test.

If you have any other questions let me know!

I have a feeling that last point where timing dipped one degree wasnt even real. Even if it was, 1 degree of knock retard is nothing to even talk about. Most poeple set thier DSMlink CEL to come on at 5 degrees of knock retard. I've been using 3 lately though since I have shit luck.

At first I thought that the timing looked a bit low in the 4k rpm bracket, but I'm not so sure. With stock maps you don't get over 10 degrees until 5k rpm, which is what you seem to be seeing. With 560s though I would expect a couple more degrees. Odds are at 20 psi on a EVO16g even taking out that ~15% of the airflow signal you are still over 2.1 g/rev. Doing some quick math it seems that 2.47 g/rev will get you there (2.1/.85 (1 - .15).

A few other things to keep in mind when looking at 1 or 2 degrees of timing drop, Intake Air Temps over 84 degrees or below ~35 degrees will pull one degree of timing, coolant temps over ~206 degrees will cost you one degree, and over 216(?) will cost you 2. I can get the actual numbers if anyone is interested, but you get the point. If your coolant temp went up over 206 in the middle of that pull you'll apear to have lost one degree of timing, but it wsnt from knock. Etc.

Both runs look good to me. And pretty impressive that you seem to be getting away with 20 psi on a stock SMIC... Take that thing to a track, I want to know what the trap speed is with those settings.

Both runs look good to me. And pretty impressive that you seem to be getting away with 20 psi on a stock SMIC... Take that thing to a track, I want to know what the trap speed is with those settings.

Thanks for the vote of confidence, Kevin :D! I, too, am amazed that I can run this much boost and still have the car running with good timing--on a 9 year old stock smic that has never been cleaned, and the stock fuel pump! It was like 45 degrees outside when I was making these pulls, so I'm sure the cool weather is not hurting things ;) The track opens March 6th here, so I'm planning to go soon after that. With any luck, I'll have some cool weather to work with.

Another thing I should have mentioned in with my log is I still have an exhaust leak, either at the turbo to manifold gasket or at the manifold to head gasket. I'm unsure how much this would affect performance, but you can hear the leak under the hood. I'm going to get it fixed before I tune further...hoping for a little faster spool after this gets fixed ;)

Stock pump seems risky. I dont have my speadsheet here at work though. Ah screw it lets see. About 115 lph at 20 psi (rough estimate), is 30 gph, 190 pounds pe rhour assuming pump gas with a specific gravity of .76, 2090 pounds/hr at 11:1, divide by 60 for lbs/min, equals 34.81. What is your airflow at with 20 psi? 2g loggers can read airflow in lb/min IIRC. If its over 34 that may help explain that small dip at the end of the log. Check it at 17 psi to for the sake of comparison.

P.S. Spyderturbo007, thanks so much for the tips on getting the byte-timing and packet delay down. I figured the test out and set it accordingly--I'm pleased to say that I'm getting 10 samples/sec according to the logger, and you can tell the data is much more concentrated in these logs :D. Thanks!

No problem, you just have to take me for a ride!!! :naughty:

My car is in storage until spring, then it's off to slow boy racing for the EVO 16g, 190lph pump & 550 cc incjectors. They said they would even tune it with a wideband O2 if I wanted (I think it was like $135). Can't pass that up :smile:

You were saying about looking down to start the log. There is an option to change the way that the log starts (this is only for v2.0). It will start/stop logging by tapping anywhere on the screen. This makes it really nice and safe. You don't have to try and hit that really small button while dodging deer and foxes!!! :rofl: Are you using v1.02 or v2.0?

You don't have to try and hit that really small button while dodging deer and foxes!!! :rofl: Are you using v1.02 or v2.0?

Lol, it did give me that option, and I checked the box that said "Tap anywhere to start log," but it didn't work the first time. I'm using v1.02 :rolleyes:

Kevin, I only logged airflow on my first posted log (the original one), and my max airflow on that log was 24.xx lb/min. I believe this is a corrupt signal since my SAFCII is telling the ECU that less air is flowing than it really is, particularly since 24 lb./min is pretty damn near the max flow rate of the T25...

I am going to get my 190lph pump in as soon as I have a garage to work in and a jack.

reading this thread is like only knowing basic spanish and living in mexico, i understand some parts of tuning and im completely blown away by olthers..but i got alittle more moding to do before i start to worry about it...

reading this thread is like only knowing basic spanish and living in mexico, i understand some parts of tuning and im completely blown away by olthers..but i got alittle more moding to do before i start to worry about it...

Don't worry man...I was in the same boat, but I'm learning fast. There's just so much stuff to consider. Like today on Tuners a dude mentioned that my car could be knocking, but if the ECU bumped me up to a sharper timing map, it wouldn't appear to be knocking as much. I imagine this is true to some degree...but I think if I were getting serious knock, timing would be getting pulled a lot more than it is now.

Funny how your car makes you spike the learning curve :p There are a lot of guys here to lend a helping hand, though, thank goodness!

The only thing that matters as far as timing goes is the airflow seen at the ECU (post AFC). But for fuel requirements, you have to get the real number. Do 1 minus your correction precentage. Divide the logged airflow by that number. So if you are at -15%, and logging 30 lbs/min, its 30/.85, or 35 lbs/min actual.

Convert lbs/min to grams/min, then divide by rpm. That gives you airflow/rev, which is what the 2.1 number is. If you are still over that 2.1 grams/rev even after the AFC correction, you will still be on the "normal" timing map (the highest one). At 6k rpm, anything over 28 lbs/min puts you over 2.1 g/rev. Its over 6500 especially that you risk dropping back below that 2.1 since airflow drops sharply. Even if you are a little bit below 2.1, the change in timing is gradual. Not like all of a sudden timing doubles at 2.0 per rev or anything ;)

If you are knocking because timing is too high thanks to the AFC, you will see it in the midrage where peak cylinder pressuers are highest, if you are still under 2.1/rev. If you are meeting the 2.1 g/rev in the midrange, you are then most likely to see it on the top end. This gets tricky since timing will rise automatically with RPM, but also with the falling airflow/rev signal. ;) Thats why all you can do is to look for timing to drop back. If its not, then you probably arent knocking. Its also a bit random, so its not like you could have a perfect stock 2.1 g/rev timing curve, but be knokcing say 3 degrees the whole time because of the higher timing from the AFC. Etc.

The short version is if you are not using some type of injector compensation chip or DSMlink, life is much much easier if you are over 2.1 g/rev all the time. In fact poeple with 14b and larger injectors for example may get some knock, but the same setup with a EVO16g or 20g that can push airflow over 2.1 g/rev and get timing back down may not knock, dispite having more airflow/boost. ;) The other thing is that the timing map at 2.1 g/rev is well defined (thanks to some DSMlink documentation), but when you drop below 2.1, its much more obscure and thats where you get stuck guessing at at timing should be. With DSMlink, you dont need to know any of this because it gives you knock in degrees of timing retarded. ;) WIthout it all we can do is compare the logged timing curve to a known good timing curve, and the only one we have is the maximum map at over 2.1 g/rev. Hopefully that makes some sense.

And this is where intercooling really helps. As we all know I always say that intercooling is the key to pump gas power, it also lets you run more boost so you can stay over the 2.1 g/rev more easily.

This is also the reason I always suggest running the smallest injector possible when using an airflow modifying tuning device like an AFC. The bigger the injector, the larger the drop in reported airflow signal, the higher timing will be, and the harder it will be to get over the 2.1 g/rev mark. ;) With injector compensation or DSMlink go straight to 950s, since it wont affect timing, or anything else for that matter.

Lets say you are dropping below 2.1 g/rev at 6500 rpm. Timing rises, perhaps 2 degrees, and you get some knock. Adding in fuel will help for two reasons. One is that the extra fuel may reduce tendency to knock. The other is that you are reducing the airflow signal less, raising the reported g/rev number, reducing the timing advance. So the key point is that when tuning by adjusting airflow you are changing BOTH timing AND afr AT THE SAME TIME. This is very important to understand if you intend to make any sense out this tuning thing.

I'll stop rambling for now...

I'll stop rambling for now...

Give me a few days to absorb that...:eek:

With regard to intercooling, I've been trying to decide whether to go with a full fmic (ebay core, custom piping, probably $300 all told), or the Dejon Super smic that my buddy Travis just upgraded from. Travis is the guy who sold me my turbo, and he said he'd let the smic go for fairly cheap--probably around $150. He made 320 fwhp on his Evo16g setup with that smic, 660s, and race gas, while still maintaining a pretty conservative a/f ratio. I don't remember exactly, but he said there was room for more power, even without cams, etc.

I know Kevin is not a smic fan, but I am contemplating saving myself $150 or so and going with a pretty efficient intercooler for this setup (the Dejon ssmic is rated at 400 hp) I'm a little worried about the pressure drop of the ebay cores with their 3" inlet and outlet...

You can't even look at what IC was on the car if race gas was used. It will bandaid even the worst of intercoolers. :) I know they make some nice SMICs now, but the problem is still like real estate. Location location location :D It sucks. There is a high pressure area in front of the tire, which limits flow through the core. So no matter how good the core is, you wont be getting any more air through it ;)

If you do a ebay type core, which is fine (its cheaper), jsut be sure to get decent dimensions. A long thin core is going to provide more pressure drop than cooling. The rules of thumb are these. Longer increases pressure drop with marginal increase in cooling (assuming a typical 20 plus inch core). Going thicker will reduce airflow to the radiator but improve pressure drop. A taller core also blocks the radiator more, but again improves pressure drop. So the idea is to compromise with height and thickness between cooling capacity and not blocking the radiator, and only go as long as you need. I personally think anything longer than 24 inches is unecessary, but admitedly I have never tested this theory. A 8" height will get plenty of air to the radiator, 10" may require some fan work and/or ducting if you keep AC and use it, a 12" core is really better for a race car with no AC condenser in the way at all. 3" thick should be the minimum, I've always used 3.5" though. Look for 2.5 inch inlet and outlet, it just works best for 99.99% of our cars. Welds have to be good! Welding is the most expensive part of core construction, so expect that to be where the skimping was done to lower the price.

Shopping carefully you can get a good deal on a homemade type FMIC kit. Joe G has done very well with the one you see on our site. I think it cost him ~400 total. Piping is just some mandrel bends from summit/JCwhitney/etc, and some couplers from various DSM vendors.

I would really rather see a FMIC on the car, but if you are not going to go bigger than the 16g the SMIC may be an acceptable compromise. If you plan to go bigger later I would just do the FMIC now, unless you dont mind dealing with selling that SMIC later when you eventually upgrade. Etc. Its all about compromise. As much as I try to avoid SMICs, there are times when they are perfectly appropriate :)

As you were talking about ebay intercoolers check this out.
http://cgi.ebay.com/ebaymotors/ws/eBayISAPI.dll?ViewItem&category=33742&item=7952706343&rd=1

I dont mean to change your thread. Sorry