Close

Page 4 of 4 FirstFirst ... 234
Results 76 to 88 of 88

Thread: "Auto Tuning"

  1. #76
    Join Date
    Nov 2009
    Location
    Simi Valley, CA
    Posts
    8,060
    Rep Points
    9,036.5
    Mentioned
    637 Post(s)
    Rep Power
    91


    Yes Reputation No
    Click here to enlarge Originally Posted by shiv@vishnu Click here to enlarge
    By "unstable", I mean a system that doesn't achieve stability when supporting 2-3 times factory boost pressure with no CPS offsetting. It is almost completely reactive, relying on a fast retard rate (immediate 3 deg pull) and a slow decay rate (adding back the 3 degrees gradually over the next second or so). You can see how much it "learns" (ie, as opposed to reacts) by simply enabling methanol in the middle of a great. Without CPS offsetting, you will see situation were timing goes from 4-5 deg at WOT to 9-10 degrees almost immediately (or as fast as the decay rate will allow). If there was any "learning" going on, it wouldn't allow for that. Instead, things would gradually ramp up towards max advance levels. Almost like comparing short and long term fuel trims. One is a fasts moving, very reactive system designed for immediate corrections. The other is slower, more deliberate and far more STABLE system designed to keep the other from having to make big corrections.

    Think of the factory knock controls system as the short term-trim. And the Procede autotuning sytem as the long-term trim. Because that is an almost perfect analogy that can be understood by anyone who has ever seen a closed loop fuel control system work. Because it's only by monitoring long term trim that you can reliably gauge how aggressive or conservative the tune is. And then make routine, tiny steps in either directions to keep it at its target level.

    Shiv
    Your LTRIM/STRIM analysis above ignores the long term timing trim, octane adaption. Which takes the brunt of the timing reduction when tuned. On the BT reset menu anyone can clearly see knock adaption (short term) and octane adaption (long term) resets. The short term reacts and the long term learns (e.g. reacts at a slower rate).

  2. #77
    Join Date
    Nov 2009
    Location
    Simi Valley, CA
    Posts
    8,060
    Rep Points
    9,036.5
    Mentioned
    637 Post(s)
    Rep Power
    91


    Yes Reputation No
    Click here to enlarge Originally Posted by supracg Click here to enlarge
    I doubt the DME was given the authority to pull high degrees of timing and keep that setting persistent; as most systems would, the DME will always try to slowly move back towards the timing curve the car should normally have after a drastic (yet needed) timing pull. Small adjustments may be made persistent, but probably not big enough adjustments for 2-3 times increases in stress levels.
    The DME has a wide range over timing advance as one would expect given the huge variance in fuels people run around the world. It will happily chug away on 94 RM2 octane down to 84 RM2 octane. As you can imagine running such low octane takes a pretty hefty timing retard which the ECU learns as a long term octane adaption. In fact those huge timing retards do not trigger ignition/glow or super knock codes. Those are triggered by sudden short term trim changes (e.g. real knock) as one would expect.

  3. #78
    Join Date
    Nov 2009
    Location
    Simi Valley, CA
    Posts
    8,060
    Rep Points
    9,036.5
    Mentioned
    637 Post(s)
    Rep Power
    91


    Yes Reputation No
    Click here to enlarge Originally Posted by Laloosh Click here to enlarge
    Proof of what? People dumping the juicebox? REad e90 much, there is a jb3 to procede thread on there weekly. Try the for sale section of e90.
    Value in this econmy? Get feature for feature on both tunes and the price is nearly identical, except one comes in one box and the other comes in several add ons
    Certainly a natural progression to go from the JB3 to the V4 as both are currently implemented as we don't offer all the features and they have a 30% sale going on. And given the number of JB3s out there naturally it's a heavily traded commodity. But I do agree the value proposition on the JB3 can be much stronger than it is now with the BT, DPFIX, dash display, etc thrown in the equation. Our CAN tool will fix that in a couple weeks.

  4. #79
    Join Date
    May 2010
    Posts
    30
    Rep Points
    29.0
    Mentioned
    0 Post(s)
    Rep Power
    0


    Yes Reputation No
    Click here to enlarge Originally Posted by Terry@BMS Click here to enlarge
    The DME has a wide range over timing advance as one would expect given the huge variance in fuels people run around the world. It will happily chug away on 94 RM2 octane down to 84 RM2 octane. As you can imagine running such low octane takes a pretty hefty timing retard which the ECU learns as a long term octane adaption. In fact those huge timing retards do not trigger ignition/glow or super knock codes. Those are triggered by sudden short term trim changes (e.g. real knock) as one would expect.
    'Huge timing retard' instances is exactly what is avoided with the PROcede's autotune; thats the point, that is what gives the PROcede the responsiveness it has

  5. #80
    Join Date
    Nov 2009
    Location
    Simi Valley, CA
    Posts
    8,060
    Rep Points
    9,036.5
    Mentioned
    637 Post(s)
    Rep Power
    91


    Yes Reputation No
    Click here to enlarge Originally Posted by supracg Click here to enlarge
    'Huge timing retard' instances is exactly what is avoided with the PROcede's autotune; thats the point, that is what gives the PROcede the responsiveness it has
    Before I bother answering, did you understand my post about the long term and short term timing adaptions?

  6. #81
    Join Date
    May 2010
    Posts
    30
    Rep Points
    29.0
    Mentioned
    0 Post(s)
    Rep Power
    0


    1 out of 1 members liked this post. Yes Reputation No
    Yes I did, but I am surprised by your response. So let me get this straight, lets say you are untuned, then you install a tune with ZERO cps offsetting etc.; you are saying you would be comfortable throwing 20 psi of boost at the ecu and allow it to take the time to make LONG TERM adaptions? This is assuming the ecu even has the authority to pull that many degrees of timing in the 'long term'.

    Your answer should be no. Through your logic, you are saying the ecu will encounter a strong knock event and retard timing, however this is short term, this will have to happen a few times over (risking the engine, whatup enrita) then eventually learn thats just how its going to be. But you are completely putting the engine at risk by doing that in those few instances. (Once again assuming the ecu even has the authority to pull that many degrees of timing in the 'long term'.)

    Now, PROcede logic: start off on map 1 or 2 at 100% ignition correction just for safety measure, and ramp up the boost until you start seeing SMALL knock events, then stop. And all the while monitoring and adjusting IC.

    It is plainly obvious which is safer.

  7. #82
    Join Date
    Nov 2009
    Location
    Simi Valley, CA
    Posts
    8,060
    Rep Points
    9,036.5
    Mentioned
    637 Post(s)
    Rep Power
    91


    Yes Reputation No
    Click here to enlarge Originally Posted by supracg Click here to enlarge
    Yes I did, but I am surprised by your response. So let me get this straight, lets say you are untuned, then you install a tune with ZERO cps offsetting etc.; you are saying you would be comfortable throwing 20 psi of boost at the ecu and allow it to take the time to make LONG TERM adaptions? This is assuming the ecu even has the authority to pull that many degrees of timing in the 'long term'.

    Your answer should be no. Through your logic, you are saying the ecu will encounter a strong knock event and retard timing, however this is short term, this will have to happen a few times over (risking the engine, whatup enrita) then eventually learn thats just how its going to be. But you are completely putting the engine at risk by doing that in those few instances. (Once again assuming the ecu even has the authority to pull that many degrees of timing in the 'long term'.)

    Now, PROcede logic: start off on map 1 or 2 at 100% ignition correction just for safety measure, and ramp up the boost until you start seeing SMALL knock events, then stop. And all the while monitoring and adjusting IC.

    It is plainly obvious which is safer.
    If you were to take a stock car and throw 20psi on it on pump gas, you'll get short term severe timing pull of more than 4 degrees and trigger ignition glow codes / limp mode. I would not suggest that. But clearly you're confused on how the OEM system works. Before you can understand how tuned systems work you need to understand the OEM logic the piggyback is working on top of. Short term is quick to pull and relatively quick to put back in. e.g. by quick I mean in the example above you'd be pulling timing long before you hit your 20psi target. Long term is slow to pull and slow to put back in. The combination of both systems can normally hone in on optimal timing within a few pulls. And this system is always in effect regardless of your CPS offset.

    In the case of Enrita, preignition definitely kills. Which is why there are boost/timing limits for a given octane even if the car indicates it will take more. He read his logs with plenty of timing and assumed he was good to go. It's actually a flaw in the autotuning logic as well which is why upper limits must be put in place.

  8. #83
    Join Date
    Jan 2010
    Location
    SoCal
    Posts
    118,800
    Rep Points
    31,797.3
    Mentioned
    2084 Post(s)
    Rep Power
    318


    Yes Reputation No
    Click here to enlarge Originally Posted by Laloosh Click here to enlarge
    Proof of what? People dumping the juicebox? REad e90 much, there is a jb3 to procede thread on there weekly. Try the for sale section of e90.
    Value in this econmy? Get feature for feature on both tunes and the price is nearly identical, except one comes in one box and the other comes in several add ons
    Do not read e90much, thankfully. I'm not surprised threads of users switching to the Procede box are more prominent there. N54tech seems to have threads about people switching to BMS or about not wanting Vishnu for such and such reason. The lines are pretty clearly drawn. We really truly honestly don't know who has exactly what amount out there except that both have a large base and n54tech represents one and e90 represents the other. We welcome both Click here to enlarge

    You need to add the BT tool and so forth if you want to add some of the included features. However, you have a choice. You aren't forced, you can start at a lower price point and get those addons as you state if you wish. So, tune vs. tune is a fairly large price difference wouldn't you say?

    Stage 2 or 2.5 E9X M3 S65 V8 supercharger kit for sale
    : http://www.boostaddict.com/showthrea...r-kit-for-sale

  9. #84
    Join Date
    Apr 2010
    Posts
    254
    Rep Points
    328.3
    Mentioned
    11 Post(s)
    Rep Power
    4


    1 out of 1 members liked this post. Yes Reputation No
    I have tried various ignition correction settings with the Procede......before autotune I used to dial in around 75% correction using around 14 psi of boost. Timing curves were pretty good most of the time.

    But Terry is right about the octane limitation. Without meth or race octane, that amount of boost on a stock setup is going to generate some timing events.

    Even more so now with the procede on autotune I get timing dips of around 3 degrees which then recover quickly and timing continues to advance afterwards.

    So the timing dips can't be entirely removed from the equation unless you increase the octane at these boost levels.

    This seems to hold true for both tunes.

    But I did notice that when I manually set the ignition correction on the procede down to zero (aka JB3) the timing dips were more frequent and a little more severe. So I suspect that cps offsetting on its own may not eliminate these events entirely, but it certainly seems to be more proactive in reducing their occurrence.

    With that said.......I didn't really know about DME long term timing adaptation, so perhaps if I left ignition correction at zero for a longer period of time, the DME may have applied enough trim on its own to reduce the knock events.

    Kind of makes some sense to me based on my logs.

    Some logs show timing attempting to hit 6-7 degrees after a shift into 4th but it quickly drops down to 4-5.

    Other logs show it going straight to 5 degrees after the shift, so maybe that's the long term adaptation at work?

  10. #85
    Join Date
    Mar 2010
    Posts
    641
    Rep Points
    76.9
    Mentioned
    51 Post(s)
    Rep Power
    0


    1 out of 2 members liked this post. Yes Reputation No
    Click here to enlarge Originally Posted by supracg Click here to enlarge
    Yes I did, but I am surprised by your response. So let me get this straight, lets say you are untuned, then you install a tune with ZERO cps offsetting etc.; you are saying you would be comfortable throwing 20 psi of boost at the ecu and allow it to take the time to make LONG TERM adaptions? This is assuming the ecu even has the authority to pull that many degrees of timing in the 'long term'.

    Your answer should be no. Through your logic, you are saying the ecu will encounter a strong knock event and retard timing, however this is short term, this will have to happen a few times over (risking the engine, whatup enrita) then eventually learn thats just how its going to be. But you are completely putting the engine at risk by doing that in those few instances. (Once again assuming the ecu even has the authority to pull that many degrees of timing in the 'long term'.)

    Now, PROcede logic: start off on map 1 or 2 at 100% ignition correction just for safety measure, and ramp up the boost until you start seeing SMALL knock events, then stop. And all the while monitoring and adjusting IC.

    It is plainly obvious which is safer.
    +1
    I am truly surprised why this concept is being so hard to grasp by some Click here to enlarge

    Also, its worth experimenting for those who want to see these DME dynamics in action. Here's what you do: Just run with 0% ignition correction for a day. Just so that you routinely see ignition drop outs so that TLOA (Terry's Long-Term-Octane-Adaptions) have enough time to settle. Then in the middle of a run, activate methanol. And see how immediately timing assumes it's normal setpoint values. You will see that this will happen immediately. And advance will stay stable under both steady state and dynamic conditions. Which goes go show just how "long term" long term octane adaptions are. And don't take my word for it. Try it.

    It's about as long-term as BMW expects it to be in a car that, at 5-7psi of boost, operates quite far from the knock threshold, 90% of the time. Double or triple that boost level without changing octane requirements, and it's only natural that the octane adaptions will be insufficient to protect the engine from knock. Which is why tunes with no CPS offsetting/timing control will see ignition glow codes when running the "wrong" map. When in fact, the only thing wrong about the map is that it doesn't offer any timing offsets to go along with the 100-200% increase in boost pressure. This is really basic stuff that any budding shadetree tuner would grasp. Unless, of course, they make profit on selling misinformation.
    Last edited by shiv@vishnu; 08-21-2010 at 08:38 PM.

  11. #86
    Join Date
    Feb 2010
    Location
    NJ
    Posts
    13,485
    Rep Points
    58.0
    Mentioned
    318 Post(s)
    Rep Power
    0


    2 out of 2 members liked this post. Yes Reputation No
    Click here to enlarge Originally Posted by shiv@vishnu Click here to enlarge
    +1
    Unless, of course, they make profit on selling misinformation.
    I was almost going to positive rep for at least continuing the discussion in a "semi" professional manner.. and then you had to say something like that . Its a bit of this dont you think?

    Click here to enlarge

  12. #87
    Join Date
    Nov 2009
    Location
    Simi Valley, CA
    Posts
    8,060
    Rep Points
    9,036.5
    Mentioned
    637 Post(s)
    Rep Power
    91


    Yes Reputation No
    Click here to enlarge Originally Posted by shiv@vishnu Click here to enlarge
    Which is why tunes with no CPS offsetting/timing control will see ignition glow codes when running the "wrong" map.
    We'll both agree ignition glow/knock codes are bad news but they are certainly not common on the JB3. Generally only seen if the customer attempts to run a race gas map on pump gas (which does happen) or if they accidentally put in a lower grade fuel. I don't believe those codes are common on the PROcede either but they certainly come up from time to time. In fact prior to the V4 I used to spend a lot of time clearing codes for local V3 owners and saw them fairly often. No doubt from trying to push too much boost on our poor quality 91 octane. The real difference isn't on what systems are throwing superknock codes but in what people post about them. When your customers post about them or other problems you intimidate them with emails and text messages to delete their posts ASAP and handle it "on the down low". We don't have that luxury on e90post as I'm not active there and rarely read those posts.

    PS. Here was a knock code report on your latest V4 autotuning firmware. May want to look in to that. Click here to enlarge

    Click here to enlarge Originally Posted by Alan l.
    Just scanned the car and got this under the shadow codes.

    30EF
    Glow ignition cylinder 3

    I remember reading that this isn't a good code (super knock) to have..

    I'm running firmware 25 8/6 maps. Map2 default settings.

    I've never gotten this code before even with 24 on the same map


    Alan
    http://www.e90post.com/forums/showpo...&postcount=242

  13. #88
    Join Date
    Nov 2009
    Location
    Simi Valley, CA
    Posts
    8,060
    Rep Points
    9,036.5
    Mentioned
    637 Post(s)
    Rep Power
    91


    Yes Reputation No
    Speaking of logging timing, was testing our CAN tool in dash timing gauge. Makes a great quick reference to see how the car is running. These runs are on map 8 with a 1000ml/min 50/50 mix of meth.

    Timing is on the fuel gauge scaled from 0-20 degrees, which works out very well. On the oil temp gauge I have IAT scaled from 60-160 degrees F. Just trying things there to see what's most intuitive and expect to change it when I think of something better.




Page 4 of 4 FirstFirst ... 234

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may post replies
  • You may not post attachments
  • You may not edit your posts
  •