EZK ignition self tuning chip
09-14-2020, 07:29 PM
#16
Rennlist
Basic Site Sponsor
Basic Site Sponsor
Genius idea! This could be a real step forward in tuning of the Bosch system!
Ken: Do you try to eliminate all knocks or allow a small percentage to occur, so the ignition timing is right on the edge, with the knock sensors actively pulling out tiny amounts of timing?
IE....when we were tuning the air cooled 3.8 liter RSR engines, up to 75 knocks per 10,000 was considered maximum. The water cooled variants allowed more knocks.
I've got a plethora of stock GT and GTS chips, if you need them....and I don't need/want them back.
Just went and checked. I've got several of the following for EZK:
2 227 355 341, which I believe is 1993 GTS. (I've got both '94 and a '95 GTS here, that I can check to see if the chips are the same.)
2 227 355 584, which I believe is GT.
Ken: Do you try to eliminate all knocks or allow a small percentage to occur, so the ignition timing is right on the edge, with the knock sensors actively pulling out tiny amounts of timing?
IE....when we were tuning the air cooled 3.8 liter RSR engines, up to 75 knocks per 10,000 was considered maximum. The water cooled variants allowed more knocks.
I've got a plethora of stock GT and GTS chips, if you need them....and I don't need/want them back.
Just went and checked. I've got several of the following for EZK:
2 227 355 341, which I believe is 1993 GTS. (I've got both '94 and a '95 GTS here, that I can check to see if the chips are the same.)
2 227 355 584, which I believe is GT.
__________________
greg brown
714 879 9072
GregBBRD@aol.com
Semi-retired, as of Feb 1, 2023.
The days of free technical advice are over.
Free consultations will no longer be available.
Will still be in the shop, isolated and exclusively working on project cars, developmental work and products, engines and transmissions.
Have fun with your 928's people!
greg brown
714 879 9072
GregBBRD@aol.com
Semi-retired, as of Feb 1, 2023.
The days of free technical advice are over.
Free consultations will no longer be available.
Will still be in the shop, isolated and exclusively working on project cars, developmental work and products, engines and transmissions.
Have fun with your 928's people!
Last edited by GregBBRD; 09-14-2020 at 07:59 PM.
09-14-2020, 08:17 PM
#17
Inventor
Rennlist Member
Rennlist Member
I forgot to mention that I can incorporate any tune using the stock maps = SharkTuned. Note the WOT map is ignored but I can move/blend that into the high load area of the cruise map.
The nice thing about the chip is that it will compensate for fuel octane, cam timing changes either manually set or just belt stretch new to old. There are times where the battery or EZK plug should be disconnected to reset the learning. Moving to a higher octane for example. It should probably be reset seasonally too, IE. winter/summer.
The eight learned maps have a 100-200 rpm resolution. Some cylinders knock at the same place all the time with the stock maps. Those areas are retarded but the other cylinders are advanced in each rpm range until the limits are reached or they knock whereby the advance is rolled back a little and they will never be advanced again at that point.
Last edited by PorKen; 09-14-2020 at 08:47 PM.
The following 3 users liked this post by PorKen:
09-14-2020, 09:25 PM
#18
Rennlist
Basic Site Sponsor
Basic Site Sponsor
I don't know about how modern engines use the knock sensors....but this seems very far advanced, for this generation of computers.
Just out of curiosity, how fast can the maps "relearn" if, for instance, someone gets a tank of crap fuel? (Will the chip instantly pull out a whole bunch of timing?)
And can the maps re-learn when a new tank of good fuel is added?
I'm trying to imagine how many hundreds of hours you have invested, in this technology.....wow!
Very, very cool!
Just out of curiosity, how fast can the maps "relearn" if, for instance, someone gets a tank of crap fuel? (Will the chip instantly pull out a whole bunch of timing?)
And can the maps re-learn when a new tank of good fuel is added?
I'm trying to imagine how many hundreds of hours you have invested, in this technology.....wow!
Very, very cool!
09-15-2020, 02:19 AM
#19
Inventor
Rennlist Member
Rennlist Member
The chip should allow for running lower octane gas in an emergency better than the stock chip. There is a limit to how much it will retard but I found that each octane point requires only a corresponding number of degrees ± and the chip should correct for any really bad cylinders. (87 and early-88 with their lower compression pistons may actually run more efficiently on mid-grade octane, assuming correct cam timing, etc.)
Drastically lower octane will spoil all the maps and the factory knock retard will be working hard. But, if it has been through the first wave of retard cycles then bad gas would actually not affect the engine much as the worst cylinders like #6 will have already been retarded significantly.
With the chip, the EZK will have to be manually reset when the engine can tolerate more advance for whatever reason. I tried different ways of restorative advance but decided against knock testing forever in normal running. After each cylinder reaches its advance limits the chip flags itself to only retard. (Some cylinders at some rpms will take as much advance as you throw at them without any benefit.)
Drastically lower octane will spoil all the maps and the factory knock retard will be working hard. But, if it has been through the first wave of retard cycles then bad gas would actually not affect the engine much as the worst cylinders like #6 will have already been retarded significantly.
With the chip, the EZK will have to be manually reset when the engine can tolerate more advance for whatever reason. I tried different ways of restorative advance but decided against knock testing forever in normal running. After each cylinder reaches its advance limits the chip flags itself to only retard. (Some cylinders at some rpms will take as much advance as you throw at them without any benefit.)
09-15-2020, 08:29 AM
#20
Drifting
With the chip, the EZK will have to be manually reset when the engine can tolerate more advance for whatever reason. I tried different ways of restorative advance but decided against knock testing forever in normal running. After each cylinder reaches its advance limits the chip flags itself to only retard. (Some cylinders at some rpms will take as much advance as you throw at them without any benefit.)
From your observations, can you rank the cylinders from 'worst knocker' to 'least knocker' for a certain RPM?
Also, which cylinders keep taking advance without benefit - would those include #5 and #8?
And just to clarify, the WOT map is not affected by your algorithm?
09-15-2020, 03:39 PM
#21
Inventor
Rennlist Member
Rennlist Member
If you have to ask... besides, this is of course a purely informational thread not to be sullied with such earthly concerns... 
What cylinders are good/bad or better said require more/less advance, depends on the engine speed, which mostly the cams and the flappy dictate. Generally, the inner cylinders, 2-3-6-7, flow better and require less advance. #6 and #2 need the least advance at their torque peaks but there are speeds (in the torque valleys) where they don't knock. The outer cylinders, 1-4-5-8, tend to run cooler and have difference flow rates so they can use more advance but have speeds where they get their act together.
The WOT map is bypassed. The cruise map is used for all loads. There is a load threshold for the addition of the per-cylinder maps. The custom maps are made at higher load levels but they are used at lower loads (with kind of a sliding scale). Once all the cylinders are 'trained' it feels to me like the engine seems to pull itself up into the higher rpms instead of having to be pushed there (with your right foot).
What cylinders are good/bad or better said require more/less advance, depends on the engine speed, which mostly the cams and the flappy dictate. Generally, the inner cylinders, 2-3-6-7, flow better and require less advance. #6 and #2 need the least advance at their torque peaks but there are speeds (in the torque valleys) where they don't knock. The outer cylinders, 1-4-5-8, tend to run cooler and have difference flow rates so they can use more advance but have speeds where they get their act together.
The WOT map is bypassed. The cruise map is used for all loads. There is a load threshold for the addition of the per-cylinder maps. The custom maps are made at higher load levels but they are used at lower loads (with kind of a sliding scale). Once all the cylinders are 'trained' it feels to me like the engine seems to pull itself up into the higher rpms instead of having to be pushed there (with your right foot).
09-15-2020, 04:16 PM
#22
Drifting
What cylinders are good/bad or better said require more/less advance, depends on the engine speed, which mostly the cams and the flappy dictate. Generally, the inner cylinders, 2-3-6-7, flow better and require less advance. #6 and #2 need the least advance at their torque peaks but there are speeds (in the torque valleys) where they don't knock. The outer cylinders, 1-4-5-8, tend to run cooler and have difference flow rates so they can use more advance but have speeds where they get their act together.
The WOT map is bypassed. The cruise map is used for all loads. There is a load threshold for the addition of the per-cylinder maps. The custom maps are made at higher load levels but they are used at lower loads (with kind of a sliding scale). Once all the cylinders are 'trained' it feels to me like the engine seems to pull itself up into the higher rpms instead of having to be pushed there (with your right foot).
The WOT map is bypassed. The cruise map is used for all loads. There is a load threshold for the addition of the per-cylinder maps. The custom maps are made at higher load levels but they are used at lower loads (with kind of a sliding scale). Once all the cylinders are 'trained' it feels to me like the engine seems to pull itself up into the higher rpms instead of having to be pushed there (with your right foot).
09-15-2020, 04:18 PM
#23
Instructor
Hi Ken
any comment about performance gain of manual vs auto cars for the chip? Worthwhile for an auto to have the chip installed? Also how sensitive is the chip to old sensors - knock, hall and cps?
cheers
Richard
any comment about performance gain of manual vs auto cars for the chip? Worthwhile for an auto to have the chip installed? Also how sensitive is the chip to old sensors - knock, hall and cps?
cheers
Richard
09-15-2020, 04:23 PM
#24
Rennlist Member
Ken what would I need to do to use this chip in a stock 87?
The following users liked this post:
firemn131 (09-21-2020)
09-15-2020, 05:48 PM
#25
Rennlist Member
Thread Starter
Ken has pricing information available at Liftbars.com, as well as, information on which sensors should be new or in fully functional condition.
The following users liked this post:
PorKen (09-16-2020)
09-15-2020, 05:59 PM
#26
Rennlist
Basic Site Sponsor
Basic Site Sponsor
The chip should allow for running lower octane gas in an emergency better than the stock chip. There is a limit to how much it will retard but I found that each octane point requires only a corresponding number of degrees ± and the chip should correct for any really bad cylinders. (87 and early-88 with their lower compression pistons may actually run more efficiently on mid-grade octane, assuming correct cam timing, etc.)
Drastically lower octane will spoil all the maps and the factory knock retard will be working hard. But, if it has been through the first wave of retard cycles then bad gas would actually not affect the engine much as the worst cylinders like #6 will have already been retarded significantly.
With the chip, the EZK will have to be manually reset when the engine can tolerate more advance for whatever reason. I tried different ways of restorative advance but decided against knock testing forever in normal running. After each cylinder reaches its advance limits the chip flags itself to only retard. (Some cylinders at some rpms will take as much advance as you throw at them without any benefit.)
Drastically lower octane will spoil all the maps and the factory knock retard will be working hard. But, if it has been through the first wave of retard cycles then bad gas would actually not affect the engine much as the worst cylinders like #6 will have already been retarded significantly.
With the chip, the EZK will have to be manually reset when the engine can tolerate more advance for whatever reason. I tried different ways of restorative advance but decided against knock testing forever in normal running. After each cylinder reaches its advance limits the chip flags itself to only retard. (Some cylinders at some rpms will take as much advance as you throw at them without any benefit.)
For instance, here in So Cal we have times of random poor fuel, high ambient temperature, and incredibly poor air (like we have, right now.) I can envision the knocks being higher, during these events (or combinations of these events) and from what you are saying, your chip being "forced" to lower timing (if I understand you correctly.)
Then we will have times of good air, lower temperatures, and better fuel quality, where increased ignition timing is possible.
Either way...a re-flashed chip or a simple power interruption to clear the memory would work..just wondering what it takes, out loud.
09-15-2020, 07:38 PM
#27
Inventor
Rennlist Member
Rennlist Member
Manual reset means briefly disconnecting the battery or the EZK plug
Removing 12V battery power scrambles the static memory in a semi-random fashion
At startup the EZK looks at two static memory addresses to check if it is still sane
(Hex values 55 = 01010101 and AA = 10101010 which are written after a reset)
Removing 12V battery power scrambles the static memory in a semi-random fashion
At startup the EZK looks at two static memory addresses to check if it is still sane
(Hex values 55 = 01010101 and AA = 10101010 which are written after a reset)
09-15-2020, 09:41 PM
#28
Rennlist
Basic Site Sponsor
Basic Site Sponsor
Manual reset means briefly disconnecting the battery or the EZK plug
Removing 12V battery power scrambles the static memory in a semi-random fashion
At startup the EZK looks at two static memory addresses to check if it is still sane
(Hex values 55 = 01010101 and AA = 10101010 which are written after a reset)
Removing 12V battery power scrambles the static memory in a semi-random fashion
At startup the EZK looks at two static memory addresses to check if it is still sane
(Hex values 55 = 01010101 and AA = 10101010 which are written after a reset)
While I obviously have not tested this chip, yet, this sounds like something that every single S4 variant could benefit from.
09-16-2020, 03:14 AM
#29
Rennlist Member
After going to your site to see the price it is extremely reasonable given the amount of time and effort that went into this chip.
To put it into perspective most 944 chips cost twice this and are just static tunes without any self tuning capability like this chip.
To put it into perspective most 944 chips cost twice this and are just static tunes without any self tuning capability like this chip.
The following users liked this post:
F451 (09-27-2020)
09-16-2020, 02:14 PM
#30
Wow, timing is everything (bad pun)!!!! I am just finishing my LH tuning and getting ready to attack the EZK. Sounds like a $200 investment could save me a lot of tuning time. Ken, any issues with boosted engines and SuperMAF?