Ran Cobb stage 2 93 octane for a while. Never had any detonation (hit 15.7 psi). I had an extra ECU so I sent it to Vishnu for a mail-order reflash with a list of my mods. Told them I wanted a 91 octane map and that I would be moving to a high altitude soon.

Tried the Vishnu at sea level. Was fine for awhile. Started noticing detonation after 6000 rpm when hot outside. This was on 93 octane, same gas station used as the Cobb. Vishnu hit 17 psi peak.

Moved to high altitude. Now I'm getting detonation at full throttle 3000+ rpm. This in on 91 octane gas - two tanks worth. Hits about 15 psi peak.

Any suggestions? I'm about to reinstall the Cobb ECU before my engine grenades.

My mods: 02 WRX MT, gutted up, crucial shorty dp, factory 3rd cat, IC hoses, STI takeoff axleback

Hmmm, Vishnu's maps are generally considered to be on the conservative side. You should call them first thing (when CA opens for business) tomorrow. My guess, it may have been flashed for the wrong map (e.g., 93 octane or for a traditional Vishnu stage 1 TBE setup which is less restrictive that your exhaust setup, etc.). Any Delta Dash logs by any chance? If you have the accessport cable and a laptop, you can use the "simple logger" beging talked about in the engine management forum to pull some data like fuel trims and the ignition advance multiplier (which is probaby going to be zero).

The other possibility is that the gas was mislabled, though you indicate that this has happened with 2 tanks, so that's less likely.

Any Delta Dash logs by any chance? If you have the accessport cable and a laptop, you can use the "simple logger" beging talked about in the engine management forum to pull some data like fuel trims and the ignition advance multiplier (which is probaby going to be zero).

I do have the free logger and used it. What parameters should I monitor? I don't think the free logger monitors IAM.

If you log manifold relative pressure, ignition timing, knock correction, AFR, RPM, TPS and engine load that should be good to start with.

Mike

Logged a 2nd gear pull and here are the results:


(RPM)/(Manifold Relative Pressure)/(Knock correction)/(Ignition Timing)/(A/F sensor #1 current mA)/(A/F Sensor #1 lambda)

3144 4.80 10.5 31.0 -6.625 .843
3284 6.10 13.0 30.5 -6.750 .828
3457 8.30 13.0 26.0 -8.125 .798
3658 10.3 12.5 22.5 -8.375 .789
3818 12.9 8.00 14.0 -15.500 .75
4036 14.2 7.00 13.5 -14.500 .75
4295 14.2 8.00 15.0 -14.375 .75
4531 13.9 8.50 17.0 -14.500 .75
4795 14.4 7.00 16.5 -14.375 .75
4996 14.4 13.0 17.0 -14.375 .75
5219 13.5 7.00 19.0 -14.125 .75
5430 12.9 6.00 19.0 -14.250 .75
5681 12.2 12.5 27.5 -14.000 .75
5919 11.6 9.50 25.5 -13.875 .75
6132 10.8 9.00 26.5 -13.875 .75
6291 10.5 8.50 29.0 -13.875 .75
6432 10.3 12.5 34.0 -13.750 .75


Anyone make any sense of this?

Hmmm, Vishnu's maps are generally considered to be on the conservative side.
Quite the opposite when compared with Cobb's maps. Vishnu maps are considerably more aggressive next to Cobb... particularly considering the up-pipe controversy.

Quite the opposite when compared with Cobb's maps. Vishnu maps are considerably more aggressive next to Cobb... particularly considering the up-pipe controversy.
Well... maybe my perception is old. A year ago everyone was complaining about the 93 oc maps being about the same as the 91. Again, I don't have data, just summing up what the complaints (founded or not) used to be. My knocklink when I was running Vishnu maps remained very quiet.

Did you hear any det in this run?

The ignition timing jumps around a bit- goes from 19 to 27.5 to 25.5. Can you try a third gear pull by any chance? I'm wondering if there are problems at the higher load sites. You are only hitting 14.4 PSI in second gear, so a third gear pull might show this.

How long has it been since you've done a reset?


Thanks,
Mike

where is the free loggign software?

funny....cause your still getting positive knock correction values....call Vishnu for sure.

Sorry if I open a can of worms with this but I just wanted to say I love seeing this thread after all the **** Vishnu talked about Cobb. :lol:

Sorry if I open a can of worms with this but I just wanted to say I love seeing this thread after all the **** Vishnu talked about Cobb. :lol:http://homepage.mac.com/mb38/woohoo.gif

Here's what I see:

When boost hits, timing drops down to normal looking values. KC drops a bit, too. A couple spots, the KC jumps up along with timing which indicates that the ECU was probably picking up a decent amount of knock in the areas where it's below 10. (Note to Drews_WRX, KC does not go negative or 0 the second knock is picked up by the ECU, it is only reduced. If further knock is heard again, it will reduce it further).

Around 5681 RPMs, timing makes a big jump along with the KC, but that's because the KC is down below and above that point.

Judging by the KC values, I'd guess that the IAM is still pretty high, maybe even 16. If it is pinging, for some reason the ECU is not learning quick enough for the operating conditions or remembering for long enough that it heard knock and reduce IAM to limit KC across the board.

Did you hear any det in this run?
The ignition timing jumps around a bit- goes from 19 to 27.5 to 25.5. Can you try a third gear pull by any chance? I'm wondering if there are problems at the higher load sites. You are only hitting 14.4 PSI in second gear, so a third gear pull might show this. How long has it been since you've done a reset?
Thanks,Mike

Yes, I heard detonation from about 3500+ rpm. Remember that I'm at an altitude of 5000 feet, so I'm not going to hit very high boost levels. At sea level, I was hitting around 17psi peak in 3rd, 4th, and 5th. 17psi is supposedly the target boost for Vishnu's reflashes.

I thought about the reset, but what is odd is that after 2 weeks of having the Vishnu installed, it started detonating above 6000 rpm on 93 octane at sea level on the same gas station I used with the Cobb. The Cobb never detonated and I ran the stage 2 93 oct map for 2 months.

I'm afraid to do anymore pulls. Hearing detonation is really bad news for my motor.

I'm not trying to bash Vishnu's products. I heard a lot of good things about Vishnu. It may very well be my car and just a coincidence that it happened at the same time. I would love to get a Vishnu custom tune next time they are in Colorado.

My impressions between the two ECUs at sea level on the butt dyno - Vishnu seemed to pull harder WOT, which may be due to the higher boost levels. Cobb seemed to have better response at part-throttle and below 3000 rpm. Off-idle it seemed you had to give the Vishnu slightly more gas as there was a slight flat spot there. The Vishnu seemed smoother in the transition when building boost from about 2800+ rpm.

where is the free loggign software?

Sorry to get off topic but I am curious as well.

Thanks

Sorry to get off topic but I am curious as well.
Thanks

http://forums.nasioc.com/forums/showthread.php?t=774757

nm....

Yes, I heard detonation from about 3500+ rpm. Remember that I'm at an altitude of 5000 feet, so I'm not going to hit very high boost levels. At sea level, I was hitting around 17psi peak in 3rd, 4th, and 5th. 17psi is supposedly the target boost for Vishnu's reflashes.If 17psi is the desired peak boost levels at sea level (where Vishnu is), 17psi is going to be pushing it at that altitude with the stock turbo. Not sure how much less boost you should be running at that altitude, but the ECU should compensate.
I thought about the reset, but what is odd is that after 2 weeks of having the Vishnu installed, it started detonating above 6000 rpm on 93 octane at sea level on the same gas station I used with the Cobb. The Cobb never detonated and I ran the stage 2 93 oct map for 2 months.The stock ECU stops listening for det above 6000 RPMs. Not sure if Vishnu extends this range when reflashing, but it doesn't sound like he does. Also, I have seen a few of Vishnu's dyno plots with A/F ratios, he seems to lean it out slightly above 6000 as well which doesn't help your problem. See that big jump in timing around 6000 RPMs I noted earlier? That doesn't help, either.I'm afraid to do anymore pulls. Hearing detonation is really bad news for my motor.Sounds like the Cobb tune works a lot better for you, I'd go back to it and call Visnu up and see if he has any ideas. With Cobb being up at the altitude you're at, it's no surprise that their tune runs better where you are. BTW, the latest 1.30 maps from Cobb are even better than their previous ones.

http://homepage.mac.com/mb38/woohoo.gif


holy **** thats funny....... :lol: :lol: :lol: :eek:

Also, I have seen a few of Vishnu's dyno plots with A/F ratios, he seems to lean it out slightly above 6000 as well which doesn't help your problem. See that big jump in timing around 6000 RPMs I noted earlier? That doesn't help, either..

This is because almost all, if not ALL of the Vishnu dyno plots are of his custom tunes. His base maps are and/or should be more conservative. I believe his problem is too much boost.

Did you tell Vishnu that you were at that high of an elevation?

**Edit, did you call him up yet?

If 17psi is the desired peak boost levels at sea level (where Vishnu is), 17psi is going to be pushing it at that altitude with the stock turbo.
I ran the Vishnu at sea level first. That is where I saw the 17psi peak and starting getting detonation about 6000rpm. Now at altitude, I get about 15psi peak and detonation at 3000+ rpm WOT.

This is because almost all, if not ALL of the Vishnu dyno plots are of his custom tunes. His base maps are and/or should be more conservative. I believe his problem is too much boost.
Did you tell Vishnu that you were at that high of an elevation?
**Edit, did you call him up yet?

Yeah, I told him the exact elevation I would be at. Unfortunately, I'm over on my cell minutes and don't have a phone service installed yet so I can't call him. Hoping that maybe Vishnu or somebody might read this and give me some ideas. They don't seem to answer their email.

I'm probably going to go ahead and reinstall the Cobb for now. If I still get detonation, then I know something is wrong.

UPDATE: Reinstalled ECU with Cobb and flashed to 91 octane 1.30 map. I hear no detonation at all (same tank of gas). Boost peaks at 14 psi. Low-end torque is much better. Full throttle feels faster as well. :banana:

It'd be interesting if you can get a 2nd gear pull with the Cobb like the one you've got with the Vishnu above. Try to do it on the same stretch of road and start full throttle as similarly as possible, too.

Now a 2nd gear pull with COBB. Looks a lot different. Same road and same tank of gas. Looks like it builds boost way faster and the knock correction values are a lot different.


(RPM)/(Manifold Relative Pressure)/(Knock correction)/(Ignition Timing)/(A/F sensor #1 current mA)/(A/F Sensor #1 lambda)

3136 10.3 1.00 12.5 -11.875 .75
3357 13.3 1.00 11.0 -16.000 .75
3674 14.2 1.00 10.0 -14.375 .75
3855 14.3 0.50 10.0 -14.625 .75
4157 14.3 0.50 10.5 -14.250 .75
4350 14.3 0.50 10.5 -14.500 .75
4702 14.7 1.00 10.5 -14.375 .75
4804 14.4 1.00 11.5 -14.500 .75
5108 13.6 1.00 15.0 -14.250 .75
5311 12.8 0.50 17.5 -14.375 .75
5567 12.5 0.50 19.0 -14.250 .75
5832 11.9 0.50 21.0 -14.250 .75
6048 11.1 0.50 22.0 -14.125 .75
6229 10.6 0.50 25.5 -14.125 .75
6399 10.5 0.50 27.0 -14.000 .75
6619 10.5 0.50 27.0 -14.000 .75


Any opinions?

From the look of the Cobb log boost and timing are a lot more conservative in the Cobb map. From the looks of it timing is very conservative down low and only starts hitting the 20's as boost is starting to taper off the closer you get to redline. The more conservative timing tables on the Cobb tune is probably what's keeping your car in check. Knock correction if almost nonexistant here. .5-1 degree of correction is amazing. I'd say Vishnu rides the boost maps and timing maps a little harder then Cobb does based on these 2 logs with these 2 maps on your perticular car.

Also, having just been reflashed, I understand that your IAM starts out at 8 and the ECU adds timing as it relearns. I had already been driving the Vishnu for a couple of weeks. I'm going to run the log again in a couple of weeks and see what happens.

I whipped up a little plot to make the data visualizating a bit easier between the 2 logs. I left off data below 3900 RPM since the Vishnu pull hadn't hit full boost yet.
http://www.we-todd-did-racing.com/wetoddimage.wtdr/wOTc2Mzk1NnM0MTNkZmQzMXk1NDE%3D.jpg
This plot fully visualizes the difference in timing between the Cobb and Vishnu map. Often the timing difference approached 7* while under load! Boost remained nearly identical, while any fueling differences can't be told without a wide band so I left that off (no need to log that in the future, either).

I've never seen KC values as low as the ones in the Cobb log (0.5-1.0). Even with the IAM at 8, you would see KC values up to 4-5 with a stock map or their S1 map. How long after the reflash did you make this run? Did you hit boost for any amount of time before logging this run after the reflash?

Vishnu in California, COBB in Utah, Rick in Colorado... Anyone think that might have anything to do with it??

How long after the reflash did you make this run? Did you hit boost for any amount of time before logging this run after the reflash?
Nice graph! This was the second drive after the reflash. The first drive it definitely saw boost as I ran it to redline in 2nd and 3rd and did some highway runs. This drive, I just drove the car until warmed up (not much boost) and then ran it on the same road.

Vishnu in California, COBB in Utah, Rick in Colorado... Anyone think that might have anything to do with it??

Except that I ran the Vishnu ECU in Texas at near sea level and got detonation about 6000 rpm. This was 93 octane on what was supposed to be a 91 octane map.

I'm thinking now that maybe the Vishnu ECU was flashed with the wrong map. I can't believe there could be that much difference between these two experienced tuners.

I went out and did some logging on my Cobb stage 2.5 v1.3 map and my suspicions were confirmed that my car just didn't feel quite right ever since my 30k tuneup. My timing dipped down as low as 6 degrees before TDC. Granted, I've only driven the car about 100 miles since the ECU was reset, but that just doesn't seem quite right. I'm going to be doing the opposite and switching to ECUtek in the next couple of days from my Cobb.

You also have to remember that more then likely the Vishnu tune has the IAM locked or starting at 16. Shiv has always done this with all his maps. He argues that if you start the IAM at 16 and tune your timing just right then the customer doesn't have as many issues with timing advance causing det. My only problem with this is that once the IAM is at 16, it takes a lot of det to bring it back down from what people have logged in the past when logging problamatic Vishnu maps.

Also, I think we can all agree that the timing on the Cobb map is pretty conservative... I mean real conservative. This is going to be your #1 contributor to no knock and the low KC values you're seeing. Also, the Cobb maps builds boost a lot quicker then the Vishnu one. The Vishnu map in this case is clearly not a safe one.

Actually, conservative timing would tend to give you large KC and IAM values as the ECU tries to max out timing in search of finding the detonation limit. In my experience with the WRX ECU, it is generally very eager to advance the IAM along with KC right after a reset. It's very strange to see Rick's logs showing next to no KC advance, but I haven't logged a Cobb Stage 2 map, either. Too bad the free logger doesn't log IAM, it would be interesting to see what it is.

I'm pretty sure that Vishnu doesn't lock the IAM at 16, but only sets it to 16 after a reset.

Raslian, want to start a new thread with some datalogs of your car? I'm curious to know why you suspect the Cobb mapping only after your 30k service? Why not go Street Tuner or get Pro Tuned?

get a ap

get a apHow about posting something useful instead of reading a sentence from every active thread on NASIOC and then spewing a few words of useless drivel? :rolleyes:

If you had actually read this thread, you'd know that Rick Schu and Raslian already have an AccessPort.

I suspect the mapping after my 30k tuneup because of the fact that they probably disconnected my battery and reset my ECU. So I'm guessing, well more like hoping that it learns itself back to what it used to be. I can take some more in depth logs fairly soon, and maybe shed some light on what my ECU is trying to do. And as for Streetuner, or Protuner? It's honestly a money saving move. ECUtek is only $650 dollars, whereas a Streetuner liscence is $400 on top of my AccessPort. If I go ECUtek, I can sell my AP for a good chunk of change, and only spend $200 tops for an ECUtek reflash. Oh, and to clarify, this change is not because I'm dissatisfied with my AccessPort, but I'm purchasing a FMIC and would like to get the most that I can out of it as far as management.

It's very strange to see Rick's logs showing next to no KC advance, but I haven't logged a Cobb Stage 2 map, either.

With the Cobb ECU, the free logger gave a message that it didn't recognize this ECU, but everything seemed to be logging correctly. Maybe its not recognizing the KC values correctly on this ECU. The Vishnu ECU was recognized and I did not get this message.

Also, I'm pretty sure the previous version of the Tari logger had IAM - it was called "advance multiplier". I used this old version on the Vishnu ECU, but most of the values were way off, including boost and rpm. It did show an advance multiplier of 13 - this was at sea level before I noticed the 6000+ rpm detonation. The new version doesn't show this value for either the Vishnu or Cobb ECU.

With the Cobb ECU, the free logger gave a message that it didn't recognize this ECU, but everything seemed to be logging correctly. Maybe its not recognizing the KC values correctly on this ECU. The Vishnu ECU was recognized and I did not get this message.Ah, that explains it. It probably wasn't reading the KC values correctly then like you suggest.
Also, I'm pretty sure the previous version of the Tari logger had IAM - it was called "advance multiplier". I used this old version on the Vishnu ECU, but most of the values were way off, including boost and rpm. It did show an advance multiplier of 13 - this was at sea level before I noticed the 6000+ rpm detonation. The new version doesn't show this value for either the Vishnu or Cobb ECU.A value of 13 for the IAM sounds about normal for 91 octane gas depending on the tune. And if you're getting det above 6000 the ECU most likely won't pick it up. The ECU relies on the IAM to reduce KC above that point which is why it's a good idea when tuning maps to be conservative with timing, fueling and boost above that point.

Ah, that explains it. It probably wasn't reading the KC values correctly then like you suggest.

Maybe not - I just read this on the Tari website:

"Please note that even if your ecu type is unknown it will not affect the functionality of the application."

Well, either way, KC values of .5-1 are very abnormal. Do they go to 0 when out of boost? Maybe try the latest version of the logger just released...

Here's a log with the latest logger version. This run is with the wastegate arm tightened 2 turns from stock (had to run stock setting with Vishnu because of their custom restrictor - caused too much spiking). Car was a little heat soaked.


(RPM)/(Boost)/(Knock correction)/(Ignition Timing)

3173 7.54 0.00 14.0
3348 9.72 1.50 15.0
3635 13.8 -0.5 9.50
3814 14.1 -1.0 9.50
4118 13.9 -1.0 10.5
4278 14.1 0.00 11.0
4578 14.4 0.50 11.0
4770 14.4 1.00 11.5
5060 14.1 1.00 16.5
5285 13.5 1.00 18.5
5551 12.6 0.50 20.0
5690 12.2 0.50 21.0
6019 11.6 1.00 23.0
6122 11.2 1.00 24.0
6399 10.6 1.00 27.5
6607 10.4 1.00 28.0