[stealth-twntrbo]

Wow The procede has come so far, Shiv I am really impressed!

[folgrz]

lol this thread has so much win

[AltecBX]

Can't wait for this to finally take over.

[Dackz]

Originally Posted by jpsimon

I'm going back to Vegas on Friday, there will be enough teasing there. Get to the good stuff now!!

Jp, you going to Mfest??? Myself and some from CO are leaving in the AM

[GreenPlease]

Sweeeeeetttt!!!!!!

[JoeyFiasco]

hopefully the public release is sometime this week

[D.B.S]

Wish it was today... I have a test drive of the PROcede to give to this guy tomorrow local to me... He's on the fence between JB3 and PROcede... but I'll see if I can swing him our way...

[jpsimon]

Originally Posted by Dackz

Jp, you going to Mfest??? Myself and some from CO are leaving in the AM

na I'm gonna be there for a bachelor party, i didn't even know m-fest was going on haha

[OpenFlash]

First post updated

[Cloud9blue]

Shiv, when will this be available to the public? I have a JB right now, but I am interested to give this a try and see which one I like more.

[D.B.S]

Originally Posted by shiv@vishnu

First post updated

SWEET.... nuff said...

[dyong]

to team Vishnu.

Originally Posted by shiv@vishnu

Say goodbye to custom tuning and datalogging. Time to let a computer do it for you automatically.

Background:
This shouldn't come as a surprise to Procede users. But to others, it's important to point out that the Procede was developed to be the last tune you will ever need. And one that can satisfy the needs of any end user. For mild to wild. From those who just want to pick up an additional 70whp running the baseline maps and never use their laptop. To those who are always hunting for the last 10-30whp and aren't shy about conducting and analysing datalogs.

With v4, we've made custom tuning easier than ever since all our "wilder" users by adding datalog log channels that aren't available with other tunes (actual throttle, actual ignition advance, dme boost, dme target, etc,) These CANbus channels provide info that takes the guesswork out of custom tuning. Not to mention separating the factory's silly IAT compensations from our tune's boost control system. This meant no need to fiddle with IAT adjustments/decays when ambient conditions change in the quest for full boost. We've also stabilized throttle blade activity to give our boost control system a level of stability that no other tune, reflash or piggyback, comes close to offering. Add to that a total elimination of shift-bog, software-dependent lag, throttle delay and other n54 idiosyncrasies that we've all had to deal with for years. All in all, it's been a long and rewarding process to solve each and every issue that that has cropped up along the way.

Less Work. Better Results.
But now we are starting the next evolution of the Procede software. One that takes the user out of the equation. No longer will it be necessary for our power-maximizing users to conduct and interpret logged data in order to make adjustments to their own tune. This was always the end goal. We just took our time to make sure that we were feeding the Procede all the necessary input data. Without sufficient input data, auto-tuning simply isn't possible. But when configured correctly, auto-tuning has many many benefits over conventional custom tuning:

1) Auto-tuning is always active. And is basing it's tune customization on data that is recorded EVERY TIME the engine goes under load. Which means that it will constantly adapt to changes in conditions (ambient temp, octane, IAT, etc,) This simply isn't feasible with conventional manual custom tuning where the user-induced changes only really apply to the test conditions the few datalogs were conducted in.

2) Computers are far better at assessing data than you and me. This means that the Procede is not only keeping track of the number of ignition "drop outs" (indicating knock activity) but also the severity and persistence of those drop outs.

3) Laptop not required. Many of us do crazy things like run our cars at the dragstrip. This often means dumping in race gas, hooking up the computer and raising boost and dropping correction. The idea behind auto-tuning is that the user can simply add the race gas, do a few learning runs on the road (or strip) and the Procede will automatically drop ignition correction and raise boost to the limit of which the octane level will support. No need for fiddling with the computer at the track. This leaves more time for eating french fries and nachos at the concession stand.

4) Just in case you think that Auto-tuning will result in self-awareness and Judgement Day-like catastrophe, don't worry. The user will have the ability to put "limits" on max learned Boost and Ignition Correction values. So if you only want to run a max of say 12psi regardless of how knock-free and conservatively you are running, you can do that. So yes, this auto-tuning logic will benefit those who aren't wild power mongers as well.


So How Does it Work?
Quite elegantly as a matter of fact. Every time a few conditions are satisfied (above a certain load, in a gear, above a certain temp, etc,.) the Procede actively monitors CAN DME Ignition advance (Actual advance minus Procede-induced timing correction). It compares the real-time ignition advance value to the DME ignition advance target which we have internally mapped as a 16x16 RPM x Load table and subjected to IAT-induced trims. The difference between DME ignition advance target and CAN DME Ignition is DME-induced ignition retard. This is not to be mistaken for the ignition correction that the Procede induces based upon it's mapping. Instead, this DME-induced ignition retard is based upon, and reactive to, actual Knock activity.

The Procede then takes this instantaneous Knock activity value, which is represented in degrees (0=0 deg of knock retard, 10=10 deg of knock retard) and constantly updates an internally calculated running average function. This running average represents the current Aggressiveness of the tune. So as operating conditions get worse (ambient temp goes up, octane/fuel quality goes down, car is driven harder and harder, etc,), the average knock retard (aka Aggressiveness) starts to rise. And as conditions get better, it will start to fall. The rate at which this Aggressiveness value changes with time depends on the sample duration of the running average calculation. Set to one extreme, we can have the Procede respond to changes very quickly within 1 single gear WOT pull. But at the cost of stability. Set to the other extreme, it could take dozens of WOT pulls before the Procede can accurate assess the tune's current level of aggressiveness. Needless to say, we settled on a more moderate setting, requiring just 3-4 partial gear pulls to determine the tune's level of Aggressiveness.

DEMO #1
For this first demo, we are just going to show how well the Procede recognizes the tune's current level of aggressiveness. We are not going to show what it does with this data just yet. That is actually the easy part. The hard part (which is simply impossible with other tunes) is, and will always be, to internally monitor how well the current tune is working on your car. Reflashes can tell this since they can monitor the same knock data that we see. But they don't have the adjustment range to do much with it! On the other hands, conventional piggybacks have the some degree of adjustments (ie, boost level manipulation and, at times, timing control) but they have NO ability to monitor how well the tune is working! So that sucks.

In the following datalog, the tune is running approx 15psi of boost on 91oct. Ignition Correction is set to 0 (no Procede-induced timing retard). So it is completely up to the DME to react to knock. This is just an extreme case to show the effectiveness of this new logic. This is not a tune that I would recommend. I did 10 partial gear 3rd and 4th gear pulls. For the first 2 pulls, the was running on pump gas only. For pulls #3-7, methanol injection was enabled. For pulls #8-10, methanol was disabled again.

You can see the Aggressiveness parameter increase quickly from 0 to 6-7 in the first 2 pulls when meth isn't active. If I continued to run with no meth, it would stay in this range or thereabouts.

Then methanol was turned on. Methanol, as you know, cools IAT and increases effective octane. It has the ability to turn a "far too aggressive" tune that is knocking and running inconsistently into a great tune that offers solid consistent AND safe power. Once methanol was injected during the pulls, you can see the Procede IMMEDIATELY recognize the change in Aggressiveness. With in a few pulls, Aggressiveness falls to nearly 0.

Then, just to confirm, methanol was turned off for the last few runs. And, as you would expect, the Procede immediately recognizes the presence of knock retard and ramps ups the Aggressiveness level back up to 6-7.

So with the ability to recognize how over or under aggressive any given tune is, at any given time, the Procede will do things that you have always wanted a tune to do but were too afraid to ask

But you'll see that in Demo #2 later this week...

[SpanKyM3]

Hey Shiv,thanks for the heads up on the pm.I appreciate your comments & like what i see & read.So when is the availabilty for the PROcede to be back in stock?My luck i am from NJ & your on the west coast.Its too bad you dont have a shop here.Hint,hint.Or you can always open one up.I am sure you'd do well here.We have alot of speed racers here.

[prich2004]

Amazing job Shiv. Congrats to you and your team's milestone achievement! Expected beta/public release date?

[cu_north]

So on this Aggressiveness scale, 0 equals most adapted, correct?
Can this be made to adapt faster?
What if someone went to the track with 91 and filled up on MS109 or similar midway between runs? Would the car need 5 full pulls to adapt and maximize the benefits that higher octane provides?

[OpenFlash]

Originally Posted by SpanKyM3

Hey Shiv,thanks for the heads up on the pm.I appreciate your comments & like what i see & read.So when is the availabilty for the PROcede to be back in stock?My luck i am from NJ & your on the west coast.Its too bad you dont have a shop here.Hint,hint.Or you can always open one up.I am sure you'd do well here.We have alot of speed racers here.

Glad to be of help. Procede's will be back in stock within 2-3 weeks. But that batch is selling out fast. After its sold out, it will be another 5-6 week wait for the next production batch after that. We don't have installers over in NJ (which is my home state, btw!) but there are many many Procede users out there than can help you if you don't want to install it yourself. But the job is really easy so there is no reason for you not to do it at home

Cheers
shiv

[OpenFlash]

Originally Posted by AlpineCloud9

Shiv, when will this be available to the public? I have a JB right now, but I am interested to give this a try and see which one I like more.

I'll be releasing it to the private beta mailing list (of about 100 users) either tomorrow or the day after. After a week or two of expected results, we'll do a full public release. There have been a lot of updates in the last 2 weeks so I'm hesitant to release everything all at once. But even without the new stuff, I think you should try v4 and compare it to your current tune.

Shiv

[OpenFlash]

Originally Posted by cu_north

So on this Aggressiveness scale, 0 equals most adapted, correct?
Can this be made to adapt faster?
What if someone went to the track with 91 and filled up on MS109 or similar midway between runs? Would the car need 5 full pulls to adapt and maximize the benefits that higher octane provides?

0 implies that the tune is so conservative that there is, on average, no DME induced knock retard. 6 implies that there is, on average, 6 degrees of knock retard which is way too aggressive. The idea is to have the Procede target a certain level of aggression (say 2-3) by actively adjust ignition correction and boost level. This aggression target can be adjusted by the user depending on his/her needs.

It's easy to make it adapt faster but at the expense of stability. Currently, we've set it so that the Procede is able to recognize aggression level after only 6-7 seconds of WOT. For the time being, I think that is a good compromise between response and stability.

shiv

[dsswain]

Originally Posted by shiv@vishnu

0 implies that the tune is so conservative that there is, on average, no DME induced knock retard. 6 implies that there is, on average, 6 degrees of knock retard which is way too aggressive. The idea is to have the Procede target a certain level of aggression (say 2-3) by actively adjust ignition correction and boost level. This aggression target can be adjusted by the user depending on his/her needs.

It's easy to make it adapt faster but at the expense of stability. Currently, we've set it so that the Procede is able to recognize aggression level after only 6-7 seconds of WOT. For the time being, I think that is a good compromise between response and stability.

shiv

This answered one of the questions I was going to ask, as to when the boost portion is factored into the equation, the Procede will keep aggressiveness (an ignition correction factor) above 0.

The 2nd question is this. If someone decides that they only want to run 12psi, but are modded enough to support 15psi with an agressvieness of 2-3. Can the proceed add ignition timing above what the DME is requesting, and effectively run a low boost/high timing race map. Effectively increasing DME timing request to the point that the the tune is running an aggressiveness of 2-3 even at the low boost level.

[OpenFlash]

Originally Posted by dsswain

This answered one of the questions I was going to ask, as to when the boost portion is factored into the equation, the Procede will keep aggressiveness (an ignition correction factor) above 0.

The 2nd question is this. If someone decides that they only want to run 12psi, but are modded enough to support 15psi with an agressvieness of 2-3. Can the proceed add ignition timing above what the DME is requesting, and effectively run a low boost/high timing race map. Effectively increasing timing request to the point that the the tune is running an aggressiveness of 2-3 even at the low boost level.

Quick answer: Yes but we will put limits into maximum allowable positive ignition corrections. At first release, however, we will probably keep things from not advancing beyond stock DME ignition timing.

Shiv

[ridin135]

awesome this is really great
Now onto a question that most might like to know. Since we can switch on two maps on the fly. Will each map remember its auto tunning say if I am on map 1 and some guy in a ferrari wants to race I switch it to my high performance map 2, (saying that you have already ran this map before on days past to get it dialed in), and she is already tuned for max optimal performance right away?
Also how is your nitrous kit coming along?

[._.]

with autotune I think the idea of maps are no more, seeing as it can be considered a dynamic max performance map.