Sort of, the timing i'm running right now is based off of the GP timing table and uses it at a guideline for the most part, but it isn't exactly the same. Due to the cars adaptations its difficult for me to know if the base timing table is what the car is running most of the time, or if its some kind of "upper limit" to the timing that cars typically never get to. I really need a dyno to figure this out for real

I have no qualms about sharing what I'm running though You can see the WOT timing curve in the high MAP areas starting from around 2500 RPM and 160 KPa to redline.

 

Arrrr! Pirate Tuning! The code be more guidelines than actual rules.... 🏴‍☠️

Good bet the other sensors have input into the final timing, so the ECU can pull timing to keep the engine safe. I am just starting to poke around in AdrianCL's safe starter tune while I wait for time to get a couple important parts in hand and installed before beginning the de-Bytetronik-ization of my car's ECU and proper WinOLS tuning from him to set up my inspection day ECU. I also picked up a used post-facelift Siemens-VDO ECU in case the Bytetronik can't be exorcised cleanly from my ECU and I need to get it coded to my car's info to start. I added 3/8 of a degree in the base spark table through the low and middle flow/RPM range, and 3/4 of a degree to the high flow/RPM range. It took the 9.1 AFRs I was getting down to around 9.9-10.0 with no knock at high RPM WOT. Middle range had three or four knock voltage spikes around 2V on WOT on my 50 mile test loop, and somewhat leaner AFRs, going from solid 11.2 it was seeing across the low and midrange RPMs on WOT to 13.2 ramping up to 11.2 at 4k RPM. I am adding fuel to the WOT table cover the knocks in that range in the next iteration. I also baked in the long term fuel trim that got set by the ECU adaptations so that it will start closer to where it wants to run....

Are the MAP labels scaled by ×8 or ×10 in ECUMaster? The table axis numbers look small for absolute pressure on boost....

 

They aren't scaled, I typically end a pull at like 196KPa which is about 28PSIa which at sea level (where I am) it ends up being around 14 PSI of boost. On my JCW 11% pulley I think that's not unreasonable. My dash gauge is more optimistic and reads 16 PSI.

I'm curious about what your knock voltage looks like during a pull, this is an area I've been pretty much blind in. What I did was set the timing very low, add MMT octane booster to the tank, and then set the "engine noise" threshold to be such that with low timing and MMT it didn't register any knock any more. My theory here was that I've seen the MMT octane booster work in my mustang and my old Focus ST and I know that it completely eradicates all knock. So if I'm running low timing and octane booster i can be pretty confident that there's no knock going on and I can use that condition to set the knock threshold. My car is fairly noisy, all the poly bushings i added to the front really come through on the knock sensor.

I'm not sure how much logging capability the Byronic system has, but does it allow you to take a log of the raw knock sensor voltage during a drive? I'm curious what kind of spikes you see and how often.


I'm running fairly rich at WOT, due to the closed loop fueling its pretty much always on target. I run 10.5:1 at WOT for the whole RPM range and then 14.1:1 at light throttle and cruising. I'm fairly rich at WOT, I hope to be able to dial that back a bit once I add an EGT sensor and get a better handle on the knock sensor response. I suspect i can make a little extra power if i'm not so rich.

 

Ah, there's the 10× scalar - it's kPa, not hPa. :D
Coffee helps with details in the early morning....

It's a squiggly line that usually bounces around at a low level unless the sensor thinks it hears a knock. There's also a knock noise trace. Most times it is not pulling timing, but if it gets high enough, it will pull a couple degrees for a short while. I can drop some log screenshots in, if you want to have a look. Adrian's timing map and fuel maps were set to a good starting point for optimization in subsequent tuning - far better than the base 550 tune that Bytetronik provided. I am trying to get a feel for how one "turns the distributor" digitally and how it affects other tables' required settings while waiting for stuff to arrive, get installed, and get Adrian"s expertise applied to it....

I can send some log data your way if you'd like a look at what Bytetronik logs. The dongle has logging built in, and you dump the logs into a computer afterwards. No real-time capabilities at all. Very old tech, and waiting since 2017 or thereabouts for a promised proggy update for bug fixes and the like. That, and relatively absent support, is why I am writing off a 4 year old $1000+ investment in Bytetronik in favor of WinOLS and ECUMaster....

I am working in that direction in my poke-and-test adjustments. Last timing adjustments to the upper WOT RPMs in the timing table bumped AFRs readings from 9.1-ish to 9.9-ish with just 3/4 of a degree more advance up there. Hoping for a similar jump with the latest increment in that area. Added fuel in the WOT table in the low/mid ranges, as well as in the closed-loop base fuel table, to get the long term adaptations within a nicer range. Shooting for +/- 3% max on those. Short term varies by the second and responds to user input, and is good for fine tuning the details into individual cells that make tables real, rather than looking like a tune stuck through instagram smoothing filters....

The pic on the right is what adopting the long term corrections into the tables does. The left picture is what subsequent adoption of short term adaptations to individual cells looks like.... :D

 

If you have a log with the knock sensor response in it i would be very curious to see what it looks like. I can share one of mine from the ECUmaster as well. The tricky thing is that on the ECUMaster we can control the following parameters with respect to the Knock sensor:
- Filter center frequency (no control on bandwidth or order...)
- sensor gain
- filter time constant (no idea how this exists in conjunction with the frequency parameter...)
There is also an "engine noise" table as well that lets us set the knock threshold. My issue is that the natural engine noise of the car at wide open throttle is enough to trigger the knock sensor so i had to increase the "engine noise" level pretty high. I'm worried that its not set up right. So seeing how it looks from the factory would be very useful.

Unfortunately my car has upgrades suspension and a few poly bushings, I suspect that this is adding noise that the sensor is picking up. Previously I used octane booster to kill all knock and I was still picking up spikes on the sensor, I suspect I need to change my filter time constant....hmmm....

 

Will post some after work. Bytetronik logs knock voltage and knock noise, and what it calls "knock retard" but is probably just an overall short term spark adaptations value. You're never 100% sure what you're actually adjusting with Bytetronik, due to the occasionally mislabeled tables and table axis and body units....

Mine too - all solid poly bushings, as they last a lot better here in the desert and keep the engine and suspension bits from bouncing around like a Mexican jumping bean. I am not so far into pushing timing advance that I am getting more than a handful of knocks in a 50 mile test loop, so I haven't done the octane booster yet. We only have 91 here west of Kansas City.

Finally got some good news on parts delivery - ECS had said mid-November for the rest of my new stock airbox parts, but they're now due next Monday. Will be able to fix the thermostat, rewire the water pump relay stuff, and add the thermatic switch that will feed the cooling fan's low speed wire power through a resistor to create an "ultra-low" fan speed that is independent of ECU control, and not have to put the aluminum "cold" air intake back afterwards only to redo it all when the factory plastic bits finally arrive. This means I may get to the next step with AdrianCL before November - extracting my ECU rom around Bytetronik's read block so it can be tweaked with WinOLS cleanly using his map pack of over 80 properly labeled tables. Might even swap in the facelift Siemens VDO MS5150 I picked up, since it seems most folks feel it is more advanced and manages the facelift engines better than the EMS2000. I want to have a good backup inspection day setup in reserve before swapping in the standalone....

 

This is what the log output from Bytetronik's dongle looks like. Sample 451 in the attached log file shows a knock event. Knock voltage was 3.x, Knock Noise Level was 165, and 2.614 degrees of timing was pulled. Knock Retard is, IMHO, mislabeled, as it occurs without knock voltage and/or knock noise level spikes also, so I think it's just a collective timing retard value for timing retard from any cause.

Sample 6738 has another, lesser knock event.

There are also a number of events that look like possible knock that the ECU did not pull timing for. The lines are very "squiggly", but following a center line through the squiggle isn't hard. I have started deleting the first 5-6 and last 5-6 samples, as starting and stopping the engine produces a false knock voltage and knock noise that skews the scale on the trace in my log viewer, and makes it more difficult to find the worst *real* or * real-ish* knock events using its search tool....

Let's see 'em! The more info we trade, the more info is out of the dark recesses of secrecy, and we all benefit - not just you or me, but anyone tuning....

 

I got a borescope and checked my cylinders today, happy to report they all look really good! the car has 150k on it now, and theres a little bit of marks on one cylinder but nothing crazy. No pitting on the pistons, lots of carbon but that is normal for this cars mileage.

Cyl 1

Cyl 2

Cyl 3

Cyl 4


I also rebuilt the front drivers side brake caliper because it appears to be dragging. However its still behaving the same even after the rebuild....hmmm. I can't turn the wheel by hand when the car is in the air, if i compress the caliper it will spin easily, but as soon as i pump the brakes even a little it locks tight. not sure its an issue, the caliper benefitted from a rebuild no matter what. Eventually I will do the other side as well. Maybe I just need new brake hoses? Not sure.

 

To clean the caliper, I did a hot ultasonic bath in alcohol, then blasted it with a pressure washer. Worked ok, could really use a strip and repaint but i'm not up for that.

 

Time for the R56S caliper/rotor upgrade?

 

I think I looked into this once before and the R56 front brakes are actually the same size as the R53 JCW brakes. . I think eventually I'd like to do a multi-piston Brembo upgrade or a Z4 master cylinder. I will say that I bled the front driver's caliper that I rebuilt and the passenger rear caliper that gave me issues in the past and I noticed that the brake feel is much better now. I think there was still a tiny bit of air left in the system from the giant fiasco I went through last winter haha. I will rebuild the front passenger caliper too just because i have the new parts and having new seals and a new piston would make me feel better

As for the log! Here you go, i've also included the tune file that goes with it. Loading the tune and the log will allow you to see the live data cursors on the tables as the log plays.
https://drive.google.com/drive/folde...5j?usp=sharing

This has two map revisions, my version control is Map_XX_Y aligns with log XX_Y. So 15_1 aligns with the log file with the same name. These logs dont have any full pulls in them, i'm actually re-dialing in the fuel table after adding some 550CC injectors.

 

Bosch 550cc/min injectors. PN: 0280158117

Note: These injectors have a preset in the EMU black software, it is very wrong. The dead time is way off. I did some searching around and found much better dead time data online. This is what I used:


The built in data caused the car to be extremely rich at idle. When switching to the data shown above it was basically right on target. There is a little error in my fuel map that i need to tune back out, but I knew it wasn't that much error. It does let me know that my previous dead time testing right did work fairly well overall. The dead time data I had measured myself for my old 380CC Audi injectors wasn't half bad.

 

The Gen 1 JCW and Gen 2 “S” brakes are the same. It is Gen 2 JCW brakes that are the Brembos. As that goes, the 12.2” Wilwood kit is basically the same as the Gen 2 JCW Brembos. I have those on my R56. Could be an option.

As for issues bleeding the brakes, is it possible that the fiasco you had allowed air to get into the ABS motors? If air did get in those, it is hard to get out. Unless you have a way to cycle the ABS motors, the next best thing is to take the car out onto a dirt road or the like and hit the brakes hard enough to engage the ABS. This will work the air out of them and into the line so it can be bled out through the caliper. It may take a few cycles of doing this before all of the air gets out…

 

I have the R56S upgrade, as it was a very cost-effective way to better front brakes. Add EBC Blue Stuff pads all 'round, and it stops very well. I have lifted the rear wheels in a couple panic stops.... Will think avout the Z4 master cylinder and such when the current master cylinder needs replacement, as my usual MO is to upgrade when repairing to amortize the bulk of the upgrade in the cost of the repair....

Will peep these after work.

I may upgrade to Bosch 650cc injectors to keep the duty cycle in a good range if the final tune at any given step along the way puts the WOT redline duty cycle above 75-80%. After I fixed the small vacuum leak a few weeks ago, logs were showing duty cycle above 100%, which, coupled with AFRs in the 8.8-9.0 range, told me the extra 2 PSI was really going to beat on the injectors hard if left with the map dialed in to "safe rich starter tune" on the now repaired boost leak. Swapped in injector info from my Bytetronik 550-scaled original tune, the WOT fuel table from the GP tune, and the first couple columns from the RMW JCW example tune in the closed loop fuel table. Have used that as a new starting point and added a little bit more timing and tweaked some fuel table cells to get down to about 84% duty cycle at WOT at redline and about 10:1 AFR (so, some room to try more timing at WOT or lean it just a tad) for now, until I get through the thermostat change and airbox rollback. The brand new factory MINI plastics are sooooooo clean.

Next, after that is AdrianCL extracting my tune and ROM data to exorcise the Bytetronikfrom my ECU, and possibly swap in the Siemens-VDO facelift ECU I picked up used a few weeks ago....

Then, drive it for a couple oil changes, and contemplate BVH and cam, along with 55% pulley and maybe a TMAP that can read past 21 PSI if needed....

Then, ECUMaster. By then, I should be past inspection day, so I'll get the biggest part of the next inspection interval on the stand-alone....

 

Nice. You just helped me identify a table in Bytetronik which was labeled in such a way that I didn't link it with the concept of "dead time".


They have presets as well, but for whatever reason, the way the proggy is set up, you have to enter the values, and check the "Easy Choice" setting to see what those settings match. Same with the scaling.

"Easy" <dr-evil-air-quotes-easy.jpg>


AdrianCL's 15 minute on-the-spot quick starter tune was pretty darned good - better than Bytetronik's "base 550 tune" - even when working remotely through their program on my computer and not knowing the dead time or injector scaling numbers for my injectors. Turns out, only the lowest two table values differ from their settings for the JCW380 injectors, so the dead time was close enough, though the numbers you posted in your table differ slightly, so maybe more like <dr-evil-air-quotes-good-enough.jpg>. With the adjustmentsI have made, I at least won'tbe beating up my injectorsat WOT quite so badly. I guess the next step will show if the tune dials in nicely and drops the duty cycle enough to leave a bit of leeway....

 

I think you're right, I used ISTA to pulse the ABS module but there was definitely some air trapped somewhere, its out now, which makes me happy That trick with the 2x4 in the last photo helped me immensely this go round, idk where I learned it but someone taught me that if you keep the pedal held down you wont drain out the master cylinder when removing a caliper. A total lifesaver. I only had to bleed out the rebuilt caliper and the hose.

these willwoods look nice for the money, I’m not so sure I want a 2 piece rotor though. I’d get nervous about those screws backing out…though I supposed as long as they’re torqued to spec it should be fine.

These 550's top out at about 70% at redline maintaining a .78 lambda (11:1 AFR) my 380's were hitting 90%. I never really was too concerned about it, but its nice to have the headroom, additionally these have a better spray pattern so the startup behavior is better and the fueling is more consistent while driving. Hows WOT fuel handled on the stock ECU? I assume its kind of a "spray and pray" type deal relying on the VE model being good.

Glad I could help The dead time is unbearably confusing, difficult to measure, and you can find different answers all over the place for what the right values should be for a given injector. The data I used for the 550's was sourced from Bosch directly. It seems to be right. Even worse? Its virtually impossible to know if you have it right, because you'll just end up baking any error in the dead time into the fuel table when you tune for fuel. The only real way to know if its right is to swap in another set of injectors that you are 100% sure have the right dead time data and see if the fueling is still accurate. The effects are most pronounced at idle and low fuel flow, at high fuel flow dead time isn't a huge consideration. I spent a lot of time trying to measure dead time, and found good success using the EMU black to drive the injectors and measuring their flow at different duty cycles, but that method doesn't give you easy control over the drive voltage. You can't run the engine, so you can only really get a test at ~12V. At the end of the day i've come to the same conclusion that many other thousands of tuners across countless platforms have come to:

Never use injectors that don't include data.

 

ALSO!

These Bosch Injectors I used (0280158117) are sold by Ford Performance for S197 Mustangs. Ford gives very detailed data on them. This is really close to the data i used, it may be more correct but I haven't tried it yet. They do give a lot more info though.

https://performanceparts.ford.com/pa...-9593-g302.pdf

A fun blending of my two cars My other project is a 2019 S550 mustang

 

The ECU runs open loop for full load, but I think, since mass airflow is calculated from manifold pressure and air temperature, that the variable of air temperature is thusly baked in, and is a key factor in the choice of cells in the "spray & pray" table. Pretty sure the knock sensor has a say in the timing as well.

100%. Found some Bosch 630cc ones cheap, but no data. Hopefully, data will be available for the ones I get, if I have to get them. Might need to with BVH/cam/55mm TVS pulley. Probably will be OK on 550s with stock head and 60mm.

The same part # is on the ones Bytetronik sells. I guess we know how they behave, then, and I can relax a bit about the duty cycle, given the AFR and the likelihood of it decreasing at least a little further as it is dialed in....

 

I think if you’re running e10 gasoline 550’s will be perfect, I wouldn’t worry about high duty cycle too much.

 

installed these LED Archaic headlights. My factory Xenon lights were in really sorry shape, headlights full of dirt and dead bugs, cracked deteriorated internal wiring, and faded lenses.

feels really wrong to remove what used to be a really high end headlight option, but the performance speaks for itself. Amazing what 20 years of headlight tech will do.

these have a goofy startup animation which I immediately turned off. I did wire up the DRLs as well.

 

You got a vid of these with and without the cheesy animation? EDIT - found the website. Does switching off the startup animation also kill the turn signal animation? The site is not 100% clear in its wording on that....

My driver's side headlight lens is a mess, and this set probably costs less than one Xenon light. They were expensive before, but have really shot up a lot since I replaced the passenger side one.

Also, do these have leveling, and does ir work with MINI's set-up?

 

turning off the startup animation does also kill the turn signal animation. No auto leveling on these, just standard manual leveling. I would prefer a clean set of factory Xenon’s but like you said one side costs more than this whole set

 

....If you can even get them.
Lotsa NLAs these days....

 

A lot of race cars use the Wilwoods. I have them on my R56 S that I use on the track. I didn’t like the cap screws that they came with and replaced them with Grade 8 bolts. To top it off, Wilwood recommends that they be installed with Red LocTight. I got my kit from TCE Performance Products. Actually, I should say “kits” as I upgrade to a custom 1” thick rotor that he markets based on the Wilwood products.

 

😬

maybe I should try and source a set whenever come across some. This definitely makes me think I should keep the ones I have. They’re in pretty bad shape but knowing they’re so hard to source means the threshold of what’s considered “trash” is a bit lower.