Advanced Tuners Please Enter - Tuning Questions

[NJGSX96]

First, I apologize for the length. I was tempted to post this in Extreme Forum, but hopefully I'll get the guys I need to come in and offer some help and insight.

What is the best way to tune the car to get the most out of it without blowing it up? Do you tune for low knock? Do you tune for low AF (real AF, like wideband)? Do you tune to EGTs? Do you use a combo?

I know a lot of guys running 11s and lower and everytime it comes up, it seems they run with a lot of knock. Some tune to what feels best or yields the most power on the dyno. Others used a wideband O2 to tune to a particular AF ratio. Others just did it by seat of the pants, looking at EGTs. No matter what method they used, they all had a lot of knock, in some instances 43 knock count straight though the run.

I usually tune to knock. Just a touch of knock is where I like it. Gets me great timing advance, good power throughout the run and, for the most part, my car always ran decent. Recently I had some issues at the track, sputtering and pissing down the track from a combo of bad spark and too much fuel and nice cold weather. A couple alternators later, a battery, new plugs and wires and I am ready to hit the track again. Nate at DSMotorsports and I were talking and he lent me his wideband to assist my tuning. He mentioned that he gets lots of knock when tuning to AF ratio, as do many others I know, and to not let it worry me as much as it normally would. All people who tune with the wideband or WB/dyno say the same thing... they couldn't believe how rich they were and they couldn't believe that with a lot of knock they could still run fast. I have now begun to question my tuning methods.

Well today I went to the track, and if you go to the Track Talk section, you can read about it. Background of car... 1G, 95 pistons, 264/264, FMIC, internal gate TD06 20G, built trans, ACT2600, VPC/AFC, 650cc, Eibach/AGX, etc etc. Built pretty good. Compression test is 188 across the board, timing set to 5°, no leaks to 20psi (TB and 1G BOV leak at that point), new lifters, all new sensors. Car is sound. Anyway, I ended up running my best times, on pump gas to boot. I tuned more for seat of the pants and EGT instead of my usual tuning for knock method. I unfortunately didn't have a chance to install the wideband, but I did write down all my settings and logged all my runs to use as reference when I do. Long story longer, each run had gobs of knock and no timing. like 8 count steady dropping in 1st, 8 with a slow drop and then a jump to to 15 by end of 2nd, 15 to 20 to 28 (stepping up in increments) in 3rd and 20 to 30 to 38 (stepping up in increments) in 4th. Like this the car saw 820-840° EGTs and just seemed to go. Car has never felt better. Ever. Throughout the day I took out as much as 10% fuel and the whole time my AF stayed at .84-.86, EGTs 800-840°. Minus the knock and lack of timing, I would say the car is tuned excellent. That knock worries me though. Maybe becuase I always tuned to knock. I don't know.

So what to do? What are the recommendations from the advanced guys here on tuning? Do I stick in the wideband and tune for AF? Do I do the combo of AF and dyno tuning? What AF do I shoot for? What about knock? Do I give it as much notice? Do I got back to tuning for knock?

Any and all help will be greatly appreciated.
Thanks for reading. Again, sorry for the length.

 

[jw]

What is your AFC and VPC settings like? Can you post logs of the run?

We have very similar setup except I run an internally gated Green. The situation sound so familar that makes me think there is a trend in how VPC makes power. In the days where a logger is not easily accessible, people tune with EGT and O2. Although I can bet money that they have plenty of knocks throughout the whole run and the EGT will not even come close to 1600F, however, they make good power. I am curious to know also how others tune with VPC because I am seeing the same exact thing. Knock spikes and bleed off as rpm increase. It also seems the knocks come around 3800-4krpm (when it go open loop). I can hear my exhaust tone changes as the wastegate gate open with a loud nasty roar (still using extremely ported 1G O2 housing). Anyway, to cut story short, my guess is VPC run too lean when it go open loop.

 

[NJGSX96]

I have to pull the logs from my laptop which is still in the garage in the car. I'll try to get them up tonight if I can. From memory, if it still works, logs looked like this, if you can picture it. The knock "steps up" around 4000 and then again around 5500. IPW is good, around 18-19 I think, tailing off towards the end of each gear. AF was .86-.84. Timing was something like 16° dropping to 13° in 3rd and a pretty steady 11-12° in 4th.

Odd thing about my settings, last trip on race gas I ran 25psi on these same exact settings, seeing no knock, no nothing, but I had other problems. Now, on pump, they seem to make the car really go.

VPC - 550 chip with 650 injectors (I bought a 660 chip, or so I thought. It was a 550 chip - whatever)
Gain at 8 o'clock for first run. Set to 7 o'clock for the 2nd and 3rd run.
Response at ~5 o'clock

AFC
-HI- 1K 2.5K 3K 3.5K 4K 5K 7K
-1st- +9 +8 +7 +6 +5 +4 +2 0
-2nd +4 +3 +3 +2 +1 -0- -1 -3
-3rd +3 +2 +1 -0- -1 -2 -3 -6

1st run was a 12.9@107
2nd run was 12.5@110
3rd run was 12.31@112

Those are the only ones I wrote the settings down for. The knock in each run got progressively worse along the same guidelines as I mentioned above while the car got progressively faster.

So you think VPC may have something to do with it? Funny. I thought that too as my car became a different animal with the addition of it. I actually had an offer to go AEM. I may take it. Everything included, brand new, trade for my VPC, AFC and ECU with TMO Stage III chip. Open till whenever for me. Maybe I'll take it after Christmas.

 

[Zerocygnal]

Usually, I'll say in the past...I tuned with an EGT gauge. Not a perfect method..but we didn't have all of the nice software and computer/ECM interface capability. With the knock rising like that what does it peak at? What gear? What RPM? Forgive me if you listed it in the data above..I"m not completely familar with my pocketlogger yet. Was this on pump gas? Have you attempted the same runs with just race fuel? If so did it make a difference in your knock?

The only way I have ever known to reduce predetonation was to use a higher octane fuel, race fuel, lower combustion chamber temperature (intake of colder air whatever means nessessary) or retard the timing and experiement with the degree that works best by results. I wish I could speak for the VPC but I know nothing about it other than it is a replacement for the MAF. The advantage and disadvantage of it's functions is unknown to myself.

The obvious:
With the introduction of computerized tuning and the ability to see an actual knock value rather than blind tuning by A/F ratio or EGT's (only stated because without it you cannot see the real effect it has on your engines health) you can try different combonations of fuel, timing, and temperatures without damaging your engine.

Honestly I would attempt to use higher octane and adjustable cam gears to fine tune the combustion sequence and record the results. If your knock is still high then the only other means I can think of off the top of my head is to find a way to super cool your air, and or richen your fuel settings.

I hope this garbled mess of words atleast sparks an idea.

 

[sunflashx]

Rich,
Do you think it's safe to tune like that?

 

[NJGSX96]

Originally posted by sunflashx
Rich,
Do you think it's safe to tune like that?

I don't know. that is why I asked. I personally do not feel happy with that way of tuning, which is why I asked to get more opinions/experiences. A lot of the local fast guys I know tune with a wideband and say that knock is normal when tuning that way. I don't want to take there word for it and posted here, asking for advice from some of the more advanced tuners. I have since de-tuned my car some by lowering the boost and richening her up to get rid of most of the knock. I have my track settings written down so when I install the wideband I can see what kind of AF ratio I was running. Now I am just waiting for some responses. I am in Disaster Recovery training all day, but when I get home, I'll hopefully be able to post up some logs from that day.

 

[rdrkt]

Rich: On pump gas using the stock ecu you will get a lot of knock there is no real way around it. I think I have taken my car pretty far on pump gas 117 traps. No matter how finely tuned my car got with the wideband and pocket logger there was ALWAYS 7-10 counts of knock. This just seems to be the ecu seeing false knock.

Knock going up as you go through the gears make sense because at the top of 3rd and into 4th you will start to heat soak your IC and you are really starting to put a load on the car. When I tuned I did most of pulls through all of 3rd and through some of 4th and I noticed the same thing.

When Wes (HighPsi91) did my tuning he always tried to shoot for around 11:2:1 and even running that rich I would still see some knock counts and trust me the car was not knocking. I just don’t think there is anyway around that. One thing that did help out a little bit was brining the base timing down to around 5 or less and switching to a plug 2 ranges colder than stock. Eventually I got so tired of trying to fool the stock ecu into getting power that I bought a Haltech because you don’t have to worry stuff like knock and the ecu retarding timing. You just set the A/f ratio right and go from there.

There will be a limit to how far you can go with pump gas on your car. My suggestion would be find a safe power level that you are happy with and then concern yourself with tuning on c16.

 

[candela]

thanks rd... that helped me out as well

 

[Zerocygnal]

Will the TMO staged upgrades allow you to overide the knock safety and allow you to maintain your timing settings without dropping the dough for a Haltech?

 

[rdrkt]

Originally posted by Zerocygnal
Will the TMO staged upgrades allow you to overide the knock safety and allow you to maintain your timing settings without dropping the dough for a Haltech?

In a word no. I have the TMO ecu on my car and it doesnt do anything to change the knock signal.

When you think about it the Haltech or AEM isnt really all the expensive comparatively. By the time you have the SAFC, VPC, TMO, Pocket logger you have spent well over 1200 to have less functionality. It didnt make sense to me so I went standalone.

 

[NJGSX96]

I just recieved my plugs that are 2 ranges colder. I will try them out. I don't care if I have to change them more often as long as they make some type of difference.

i will also try to lower my base timing some. At the moment it is at 5°. I checked that last week and usually do during maintenance downtime.

I usually do tune with race gas when I go to the track as I want to run as fast as posible, but last visit to the track was a fiasco with many parts failing, so I wanted to get some baseline pump runs in to get the car street tuned. i was so pleased with the performance I may have tuned it too much on the edge. Knowing this, I did set the fuel settings back closer to the conservative side. When I get the wideband installed I can get a better grasp of the AF ratio I am running and also see what I was running at the track.

Regarding going standalone, you may have swayed me some more. Liek I mentioned, I was offered an even swap. AFC/TMO/ECU/VPC for a new AEM unit including MAP sensor, wideband, and whatever else is needed. I'll have to see what happens when I do some wideband tuning and finaly make it to the dyno.

Anyone else care to chime in? Noslaser? I know you have your tuning methods too. I'd love to hear them as well as anyone else's opinions, methods, etc.

Thanks again guys.

 

[larryd]

A couple quetions..

When see'ing knock, is it the same display on a 1g vs a 2g. Someone told me that knock sum on a 1g is different then knock count on a 2g.

With that in mind, does the ECU retard timing for knock on both 1g and 2g ECUs or is that just a 2g thing?

Sorry Rich, not trying to hijack the thread, just a couple sorta related questions

 

[NJGSX96]

Larry you thread hijacking biatch!!! HAHAHAHAH No problem dude.

I think someone expalined it well in another thread. Something about the 2G pulling timing from any knock, with a high count of 18 pulling 6.4°. You were in the thread so I am sure you saw it. On a 1G ECU, the count goes up to 43. From 0-6 count, the ECU gives timing. 7-43 the ECU will pull timing, moreso at the higher counts. I don't know the exact numbers. So to answer your other question, yes the 1G ECU will retard timing when it sees knock with the amount depending on how much knock count is happening. the 2G just seems to retard timing with any knock while the 1G allows a little before it pulls timing.

As it stands, I may be going standalone very soon. At the point I am at, I think the stock ECU is just becoming too much of a hinderance, making my next step a standalone like the AEM.

If anyone knows him or sees him, get Aslan in here. I am really curious as to his take on this whole issue. Any other wiseman too, like Tevenor, Leon, Batman, jw or Wes. I am looking for as many opinions as possible and those people's opinions, along with rdrkt, mean the most to me. No offense to anyone else.

 

[rdrkt]

Originally posted by NJGSX96
As it stands, I may be going standalone very soon. At the point I am at, I think the stock ECU is just becoming too much of a hinderance, making my next step a standalone like the AEM

Good call. You have a lot of great tuners in your area that will be able to help you out if you run into any problems.

 

[Zerocygnal]

Interesting...I read about the Haltech before and I really wanted it for my goal. After reading about the nightmare it is on initial setup (and the fact this is my daily driver) I didn't want to go that extreme. Mid 12's to low 13's can be achieved without it. Now I have this Haltech bug biting again, but the computer geek in my head says there is a way to get around it. It's starting to sound apealing again. I guess I will start a poll on how many people use the Haltech and start getting some more input.
Well I tried to help...but it seems I was helped.

Thanks
Ty

 

[larryd]

Originally posted by NJGSX96
Larry you thread hijacking biatch!!! HAHAHAHAH No problem dude.

I think someone expalined it well in another thread. Something about the 2G pulling timing from any knock, with a high count of 18 pulling 6.4°. You were in the thread so I am sure you saw it. On a 1G ECU, the count goes up to 43. From 0-6 count, the ECU gives timing. 7-43 the ECU will pull timing, moreso at the higher counts. I don't know the exact numbers. So to answer your other question, yes the 1G ECU will retard timing when it sees knock with the amount depending on how much knock count is happening. the 2G just seems to retard timing with any knock while the 1G allows a little before it pulls timing.

As it stands, I may be going standalone very soon. At the point I am at, I think the stock ECU is just becoming too much of a hinderance, making my next step a standalone like the AEM.

If anyone knows him or sees him, get Aslan in here. I am really curious as to his take on this whole issue. Any other wiseman too, like Tevenor, Leon, Batman, jw or Wes. I am looking for as many opinions as possible and those people's opinions, along with rdrkt, mean the most to me. No offense to anyone else.

yeah I saw that thread after I posted this question. Its quite an eye opening bit of information for me. I feel that I may be following your same steps very soon if I cant get the DSMlink to deal with the knock soon. Haltech or AEM EMS would be calling my name..

 

[Irks]

You could always tune for low knock, then disconnect your knock sensor when you actually go for an ET, to get better timing advance Not that thats a great idea... but it would work, and it would get around any sort of phantom knock the ecu is seeing.

 

[BatmanGSX]

Originally posted by Irks
You could always tune for low knock, then disconnect your knock sensor when you actually go for an ET, to get better timing advance Not that thats a great idea... but it would work, and it would get around any sort of phantom knock the ecu is seeing.

Don't forget to ground the knock sensor.

 

[jw]

Originally posted by NJGSX96

AFC
-HI- 1K 2.5K 3K 3.5K 4K 5K 7K
-1st- +9 +8 +7 +6 +5 +4 +2 0
-2nd +4 +3 +3 +2 +1 -0- -1 -3
-3rd +3 +2 +1 -0- -1 -2 -3 -6

Your AFC setting explain why your knock sum "step up" at higher rpm seeing you go leaner. I have it rich things up at higher rpm so my knock sum usually bleed down instead. The intial jump in knocks is something that I cannot tune away for some reason, I know it's tuning instead of mechanical because as I rich things up, the knock sum drops. However, the car feels like a dog so I just leave it lean instead.


Btw, I just switched to 550 chip and the knocks are still there, nothing change too much. The karmen reading does show a good increase tho even with the gain knob at 7 o'clock. I will try to play with my AFC settings to see if I see any difference. My lower rpm range look similar to yours. I will try to rich things up at 4krpm and lean it out later to see if it makes any different.

I am not a great fan of standalone because I dont feel the need to. Although the price of VPC/AFC/TMO combo justify going standalone but I still prefer simplicity. That's just me tho.

 

[NJGSX96]

I tried to go richer with the RPMs and all it did was make the car start to stutter. Maybe i was too rich, but I also had othr problems the night I tried that. Working from the last settings I ran at the track, I have richened up the top RPM range past 4K a couple % more than I had and that seemed to suppress the knock some. I need to do another logging run or two. Car still feels great though. I'd love to get some dyno time in. Maybe a Christmas present to myself. I also still have some wideband tuning to do and I want to get my coil-on-plug installed. I did a test on my stock coil and while the secondaries tested good, the primary was a little high at 1.1-1.2 (spec is .77-.95). Don't know how much of a factor that could be, but out of spec is out of spec.

jw, I hope you didn't switch the VPC chips to test for me. If you did, I really appreciate it. Regardless, I appreciate all your help. Let me know what you find or what settings you come out with.

Regarding the standalone... I really do like the setup I have with the VPC/AFC/TMO. Like you said, it's simple and does work. I may still go standalone, but if I do it won't be until next year. Like you mentioned, I am not sure if I NEED to. Regarding price factor, I got lucky and got everything used so my price was very reasonable.

 

[jw]

Nah, I just want to try the 550s myself to see if there is any different. Well apparently not too much. The lower aircount on 660 chip does give me more timing tho.

Yeah, I guess trading for AEM sounds like a deal but I am just curious why the other guy want the trade? I guess he must be as tired as me just wanting something that has proven to work instead of guessing what's wrong with the setup.
AEM software looks decent but I dont know how well they work at the track. Haltech works but DOS sucks.... I would pay for Haltech if they support both for notebook and palm but they dont... Oh well..

 

[NJGSX96]

I figured the 660 chip would give more timing due to the lower aircount. I will probably stick with my 550 chip. No big deal there. For reference, what settings are you at with the gain and response and what does your AFC settings look like so far?

The AEM looks very good and has some nice features. Each day that passes someone figures or finds something new. Figure by early next year I will know if it is the way to go or not. Haltech is nice too and from what I hear they have or are coming out with a Windows based version.

As for my trade, it is with a friend of mine who owns a parts business. Since I don't want my car down while I try to sell my stuff to get the money up for the AEM, he offered the trade. Not that it would be hard to sell what I have, I just don't want any downtime and don't want to dip into the house fund or new car fund to get it outright and then hope to sell what I have.

 

[stevetford]

When you think about it the Haltech or AEM isnt really all the expensive comparatively. By the time you have the SAFC, VPC, TMO, Pocket logger you have spent well over 1200 to have less functionality. It didnt make sense to me so I went standalone. [/B][/QUOTE]


Amen to that. When I first started posting to this board everyone was asking me why I am using the Haltech instead of the stock ecu with this piggyback and that piggyback, a 2G MAS, a TMO stage 3 chip and so on.
rdrkt is right, once you start adding the price of all these add on's you have just paid for a standalone ECU.
I had my Haltech running in a day and 3 days later the car was running better than it ever did with the stock ECU. Starts up like stock, idles like stock drives around like stock and so far has killed every similar equiped stock ecu car I have raced. I will never look back. Good luck with your setup rdrkt.

Steve

 

[jw]

Originally posted by NJGSX96
I figured the 660 chip would give more timing due to the lower aircount. I will probably stick with my 550 chip. No big deal there. For reference, what settings are you at with the gain and response and what does your AFC settings look like so far?

i dont know if my setting is worth mentioning since I am still tuning with this 550 chip. Currently, here is it:
Gain - 7 o'clock
Response - 5 o'clock
Hi setting:
1k - +3
2k - +2
3k - +1
4k - -2
5k - -4
5.5k -5
6k - -5
7k - -5
Lo: 25% Hi: 40%

 

[95GST]

Hi, on the DSMLINK the knock retard, each deg is about same as 3 knock count in 1G terms. So if you see 3 degree knock retard, is about equal to a 1G having 9 knock counts.

I found I have to run 9:1 - 10:1 (really rich) to have almost no knock. As I lean it out, I get more knock. You can't deny that is not real knock!

It seems that, to run almost no knock on pump gas, you either have to turn down the boost, retard timing, or run really rich, all of which can slow you down.

I'm not sure how a standalone could help, I'm not familiar with them. With the stock ECU and knock sensor, it's telling you that you have real knock. This is a physical thing that is happening in your combustion chamber.

I thought about ignoring the knock to a certain degree and tune on the wideband targetting to a 11-12:1 ratio. It's is definitely true you make more power as you reach towards stoich, but the power gain is mostly from the leaner mixture. On the dyno I started at -26%, I leaned out -6%, I have picked up 20HP, I leaned out another -6%, I picked up another 10HP. However each time the knock increased. I am making more power, but how long is my engine going to last with the knock's presense with this method.

Each time it knocks, it's applying excessive pressure on the piston, on a bad day/bad gas you can detonate bad enough to take a chunk off your piston top, or kill the ring land.

It is possible that the ECU listens inteprets the knock sensor signal and over react and translate it into excessive knock count/knock retard. In that case a standalone could help overcome this "hinderance". But how do we know for sure how much timing should be retarded at x amount of knock? The mitsu engineers probably have programmed the ECU to respond to knock very aggressively, as they cannot afford to have cars getting piston damage from knock during a wide range of weather and different octane all through out north america.

I'm in the same boat trying to decide how to best tune it on pump gas.. This is a great thread.