When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
'thought i would start a new thread as my initial thread was kind of a start up. with over 240 pulls and a year hard at it sharing my experience with my ECU might be a help to some.
after 22 seasons of road racing for a Nat Championship in SCCA's GT3 class i have done what i wanted to do grinding up tires and brake pads. i have always street driven what i raced and when i switched over to Mazda i bought a succession of RX7s ending with my FD. of course it didn't take long to start engineering it into what i wanted: a true dual purpose car. this was in the Apexi Power FC era. with the addition of the Datalogit i was logging two EGTs, digital fuel pressure and exhaust backpressure. when i ditched the trainwreck OE turbo system there were no medium frame 80 pound per minute turbos. you had to go GT42 for 600 rw rotary hp. the GT42 was a large frame turbo and as such it was not able to be placed near the exhaust port. there was, however, a very nice 42 pound turbo (Garrett TO4e 44 Trim) and 2 times 42 is 84. i designed and built a two turbo setup and ran it for a number of years and over 10,000 miles. two 3 inch downpipes etc. there is a thread somewhere on the board. after a few years turbos evolved and i designed a single turbo manifold to accept Garrett's wonderful compact 80 pound per minute GT4094R.
and became fixated on the Texas Mile and 200 mph on a daily driver closed wastegated setup. this was 2013. i figured 575 with a 6 speed would get it done as someone on the board had done 198 with a 550 V8 FD.
time for a new ECU. i did my research and at the time Haltech had their dreadful E6b, AEM had a theoretical ECU for the rotary that no one could make work, Motec was the high dollar pro option but had all kinds of ala cart add ons and then there was ViPEC-Link. it was an easy decision in 2013. my ViPEC V88 did everything, was easy to work and accepted all my extensive instrumentation.
i moved from Wisconsin to mountains of NE Georgia in 2020 and finally was able to get back to my car in Dec of 2020. apparently something in my motherboard decided to call it quits. ViPEC had merged into Link and Link took a look at it and suggested the newest iteration of the V88, the Link G4X XtremeX.. a super good idea as it spoke the same language as my V88 so the transition was easy. as i was transitioning in early 2021 i discovered Evans Performance Academy. Jeffrey Evans is amazing. he has training videos for most ECUs. there are around SIXTY specific 30-45 minute videos directly relating to the Link G4X XtremeX! he also has a forum so if i have a question i get it answered almost always the same day. Link also has an excellent forum.
i am (still) obsessed with the Texas Mile. it will place a maximum stress on the 13BREW. 30 seconds at 575 rwhp. and it is all about dealing with cumulating heat. actually, the turbo rotary, whether doing the TM or just buzzing around is all about heat. the rotary is a 2 cycle motor and as such does not have an every other cooling TDC stroke. by virtue of the peripheral exhaust port it is easy to load up the motor with charge air and therefore huge combustion chamber pressure and combustion chamber heat.
back in the stone age, 1993, Mazda added a knock sensor. while it merely reported knock (the product of too much heat), it was helpful to those that paid attention to the output.
of course what was needed was a reaction to knock.
at 6000 rpm there are around 40 ignition events per second so all the talk about cutting boost or cutting fuel on a knock event was a laugh.
when you get knock, and you will--- just a tank of crappy gas---- you need a reaction on the next rotor face!
there is only one solution and that is a timing reduction. it works.
i found a module to do it... the "J&S Safeguard Knock System." i ran it from 1991 until 2013.
interestingly, there were so many discrete modules back in the day... knock and boost as examples that are now managed inside of our new ECUs.
given the turbo rotary combustion chamber pressure and heat, knock monitoring and reaction is very close to job one. simply put if you have a single turbo and your ECU does not have an ACTIVE (meaning it will cut timing) knock system i would advise you to swap it for an ecu that does. if you do have an ecu that has active knock... ENABLE IT.
which finally gets me around to being back on point here.
my ViPEC V88 would generally report knock around 30 on a scale that reached 1000. driving around i would see 5, maybe 10. on a boosted pull it might hit 30. i had one incident where i had what i consider real knock and it reported 350. (i was running a lot of AI through fuel injectors and the meth plugged them).
as i started to get into boosted runs with my new XtremeX i was seeing way different numbers... around 70-170. i posted on the Link board and Link responded that the new ecu had an entirely different/uprated system than my 10 year old V88 so don't sweat it. i guess i did sweat it. the problem was that driving around the numbers were very close to zero. when i went into boost the knock rose in lockstep with the boost.
my primary interest in 2021 was comparing intercooler cores... (again, it is all about HEAT). as i was doing this i couldn't keep being concerned about my knock readings.
i tried everything:
timing advanced retarded... i made a run at 4 advance, 30 psi with my 9180 and it was so retarded that at 30 psi the car felt slooow. zero effect.
AFRs, ditto, all over... zero effect
replaced knock sensors, zero
fuel, 93 pump, 93 pump with 30% ethanol, 40% ethanol and even loaded my Texas Mile fuel which is 75% VP X85L and 25% Q16... zero effect.
all four intercoolers, zero change
removed hood, relocated air filter into ambiant air, zero.
decided to buy a new Mazda block, swapped it in after doing my usual ports etc CPR build... no effect, ran perfectly just like the motor i removed
as i was going through these gymnastics, of course, it emerged in the back of my mind that maybe Link's advice was sound... that the numbers were just a different scale and the ViPEC 30 was similar to the XtremeX 150.
in both my motors the vacuum hasn't changed... so probably no knock, just different scaling.
i am buying a Link Knock Block today. it simply amplifies motor block sounds and adds a set of headphones so if you actually have knock it is easily recognised. it will help me determine the best timing/AFRS for my fuel mix.
once i have a bit of experience with this i will post the results. i split a Link Block with my wonderful tuner back in 2013 and he still uses it as his most primary tuning tool. he has gone on to great fame, and deservedly so. for instance he consults with all of the Toyota cars in the NASCAR Xfinity Series. He does mega hp multi engine offshore boats. owners fly (private jet) him down to Miami to tune their crazy boats. etc.. point is the Link Block is his go to tool.
film at eleven
more to come
Last edited by Howard Coleman; 02-22-22 at 12:36 PM.
6400 rpm is peak torque for my ports. peak torque is created by maximum fill. it is at this point and higher revs where the decreasing amount of time catches up with flow. max fill creates max pressure/heat. this is often where knock occurs.
i had my knock threshold set at 109. note the lower graph which is the timing reaction. the rear knock set off the timing reaction and so the lower graph shows it scaling into as much as 4 degrees adjustment.
i have recently revised my concern as to what is "knock" and what isn't and have moved the knock threshold up to 200 ATM.
it appears that our motors make some sort of noise under boost that they don't while driving around. (note nothing as to readings after boost) just what is this noise(?). my guess is that there may be something that is mini-knock which is inconsequential but with the better monitoring equipment does show up.
perhaps my knock ears will help with this quandary.
perhaps my knock ears will help with this quandary.
they might! i put a screw driver on the knock sensor and my ear on that and was surprised at how much noise there was, even at idle. not a very pleasant noise either.
its also pretty clear that the link is the price performer in the ECU world right now.
once you are able to monitor all the metrics from the front and rear rotors the next objective is to optimise each of them. this is really easy with regard to fuel and timing with the G4X architecture. initially my egts were around 60 to 70 F apart. i simply added 5% more fuel to my rear primary injector. at 6206, 26.6 psi (9180), around 540 rwhp the readings are 13 F apart. elevation 2460 so 8% less air pressure.
don't know whether it was a good move or not but it makes me happy.
there are loads of other situations that can arise that call for an instant (as in ms) RESPONSE. for instance, fuel starvation. i am at 5631 rpm, 100% TPS, at 32.3 psi and my fuel differential pressure dropped from 43.5 to 7.5! i had just installed a Hellcat pump and am running it in boost at 21 volts so i have tons of fuel delivery, except when the pickup is sucking air.
while this dip looks pretty long it is only about .4 of a second and it recovered in the following .3 of a second so no harm.
i thought i had enough fuel in the tank (12 gallons) but the turbo hits pretty hard. i reinstalled my surge tank cover and all is now well. the fuel pressure drop caused a 117 front rotor report which cut timing 4 degrees in the front rotor.
Last edited by Howard Coleman; 02-22-22 at 04:58 PM.
nice weather so did another pull today. i drained my AI tank and switched back to 100% meth.
it seems like the swap to 11 heat range from 10 helped a bunch as to knock. quite possible the ground strap on the tens acted as a glow plug. for sure the entire strap was clean indicating they were too hot for the climate.knock, finally is not number one on the worry list.
i installed two (additional) Bosch sensors on the rotor housings as i bought a Link KnockBlock which should arrive tomorrow. it will be interesting to hear what is actually happening.
i did a run to 7500 in third and finally got close to a reasonable AFR. nothing but meth as AI. ( rpm limits is due to the roads... nothing straight in the mountains)
553 hp at 7423 rpm, 10.83 AFR 25.8 psi w my 9180. 2560 altitude. add 40 for sea level. 12.09 mph/second uphill.
After a year of reading higher knock numbers with my new ECU... and doing just about everything to lower the readings i finally decided to actually get an answer... back in 2013 as were were tuning for the Texas Mile i split a Link Knock Block with my tuner. it is still his go to instrument for tuning. my very own arrived today.
in anticipation i found my original (2013) Bosch sensors that i had replaced last June and had mounted them yesterday. i mounted them to a ground post on each rotor housing. so i now have four knock sensors, two for my laptop and two for the ears.
the module is in my engine bay and has a volume setting. i set it in the middle and went for a ride using a set of earbuds. i could clearly hear the motor buzzing along doing it's thing. nice. i drove over our mountain (passing the Appalachian Trail) and headed down the other side where there is just enough road for a pull to 109 mph in third.
i didn't hear a peep from my motor other than the usual buzz. NO KNOCK. for some reason i was rich, yesterday i was close to being O K. and i had leaned the fuel table a bit. oh well, i will figure it out. but the big deal was the max knock reading wasn't KNOCK. this squares a bit with what Richard Green suggested... they see 400/600 numbers on their p port 20B. so apparently 160 isn't anything to be concerned about.
being a bit rich the motor made 541 at 7309 and then less as it headed rich. 76.8 to 109.1 mpn in 2.683 seconds so 12.04 mph/second. pretty good at 2460 feet and on a slight uphill grade.
Hi Howard, as another Link user (older G4+ Fury), I'm very interested in this thread. Sounds like since you've got your Link Knock Block, you've confirmed for yourself that the knock thresholds we set in these ECUs are only as good as the data used to calibrate them to detect when actual pinging/knock occurs. For a piston engine, the calibration task is pretty easy - after the motor has been tuned and you're happy with its output, run it on the dyno, and advance timing till it just starts to ping/knock, collect your knock sensor data and use that data to set your knock thresholds. Not as easy (or cheap) for a rotary because inducing knock on purpose with a rotary often results in blown motors...
So what do we do instead to find & set our knock thresholds? Since I'm not chasing that last % of RWHP, my intent is to use the knock control feature as an extra layer of protection to guard against a tank of bad gas or similar issues. Here's the process I'm planning on using to calibrate my knock thresholds, would appreciate your thoughts on this:
(1) Tune the car on the road & dyno to a level of performance that I'm happy with that leaves a conservative factor of safety built into the tune, i.e., it should be tuned to NEVER ping/knock under any normal driving conditions up to max boost/power with available fuel. Though the turbo I've got is capable of more boost, in the interest of reliability, I'll likely limit my max boost to 18psi MGP or less, shooting for high 300's/low400s RWHP goal as that will be plenty.
(2) Turn on knock control in the Link, but set knock thresholds everywhere in the table (RPM vs. MAP) to the maximum 1000 level so the knock controls should never intervene to trim timing. This will allow me to collect knock sensor data (knock levels for each rotor.) on every drive so I can characterize what "normal" sounds like to the knock sensors for all driving conditions on the road. Initially may need to play with the knock sensor gain and filter (i.e., narrow or wide band, various frequency bandwidths) settings to optimize your signal to noise ratio. Ideally, you should see knock levels in the 10s, say 0 to 100 from idle to red line while off boost, perhaps in the low 100s when in boost, say 50s~400s or so.
(3) To make sense of all the knock sensor data collected over multiple road sessions, I'll be using the histogram feature of the MegaLogViewer (MLV) analysis software. Basically I'll replicate my knock threshold table in MLV (RPM vs MAP vs Knock Threshold), and for the Z-axis, and it will histogram the knock level statistics for rotor 1 & rotor 2. Then you can drill down to each table cell and see you mean average, minimum and maximum values of all data collected within that cell - I would initially set the knock threshold for any given cell at about 5% less than its MAX value. If that results in too many knock false alarms, I would raise the threshold of the offending cell in perhaps +1% increments.
"use the knock control feature as an extra layer of protection to guard against a tank of bad gas"
sub par gas will happen to most of us. in addition to gas quality there is a long list of other culprits. none of these many factors should cause you to lose an engine.
the solution of course is an active knock system.
setting it up, whether for 300 or 600 hp, is fairly simple. it isn't enough to have a knock system in your modern ecu, you have to enable it
use 3500 cps narrow band for your 2 rotor. turn it on and go for a drive. log, log, log. after reading the knock output you will soon zero in on a top of your readings. Link suggests use that top and add 20%. that's your threshold.
Pete has taken a more disciplinedd approach and i am sure it will work well. when i was trying to understand the output from my new ecu i was getting high readings around 67/70 so i set my threshold at 100. as i raised the boost and better set my AFRs i noticed that my knock rose a bit over the threshold. (didn't feel anything at the time). but noted in my logs that the ecu had lowered my timing in either the front or rear rotor. instantaneously, which is good.
as i have continued raising the boost etc i have raised my threshold. currently 200 on a scale of 0-1000. i am seeing max readings around 160. as mentioned, i bought a Link KnockBlock and my last run had a 162 and two 120s. i heard zero knock thru my earbuds so i conlude it is just some sort of noise. i am very happy to have my KnockBlock.
Link has something new within the knock system which they call "normalised." they are really excited about it. i think it does sort of what Pete is doing. i haven't gotten there yet.
it hasn't helped that the different ECU's have different scales...
this is the relevant page from the Service highlights
in the older engines Mazda determined a "Knock Zone", and the 12A-T having no knock sensor, retards timing. the 13B-T adds the knock sensor, but i think they still only tune in a specific zone (i couldn't find it, it also mentions the frequency)
the FD being newer would probably expand this zone, but in stock form would probably still not listen to the knock sensor much over a certain RPM.
this is the Rx8, it explains the system nicely, but isn't very helpful
use 3500 cps narrow band for your 2 rotor. turn it on and go for a drive. log, log, log. after reading the knock output you will soon zero in on a top of your readings. Link suggests use that top and add 20%. that's your threshold.
Interesting choice for the frequency band pass setting - Are you using the typical Bosch "doughnut" sensor? Those are wide band sensors which are basically piezo microphones that cover most of the audible spectrum (20-20KHz); I thought I read somewhere in the Link help files that for that type of sensor you need to use one of the wide band settings & band pass options for one of those sensors? And use one of the narrow band options for one of the frequency resonant type piezo sensors like the OEM FD sensor?
Anyway, here's a screen shot of my Link G4+ knock settings that I'm currently using. Note that I just have it setup to log knock data to characterize sensor response, so the thresholds are set to 1000's across the board for now. I'm using the 4KHz - 10KHz wide-band frequency filter, and a gain of 4. I might try switching to the 3.5KHz narrow band setting and possibly tweak the gain a bit lower (more on that below...)
Originally Posted by Howard Coleman CPR
Link has something new within the knock system which they call "normalised." they are really excited about it. i think it does sort of what Pete is doing. i haven't gotten there yet.
The G4X has significantly better processing power and more internal memory than the previous G4+ generation, so they are probably tapping that to do some additional processing of the knock sensor signal - so perhaps its no longer just a simple amplitude based threshold break; maybe they are doing some limited frequency & time domain processing on the sensor signal to help differentiate random noise from likely knock sound characteristics. Every engine when it pings/knocks will produce a very specific set of sound characteristics (dominant frequencies, rise/fall times, etc.) that can be easily sifted through noise if you know what to look for.
Originally Posted by j9fd3s
it hasn't helped that the different ECU's have different scales...
That's for sure, also doesn't help that all the different ECUs just use simple/stupid amplitude based threshold measurements to flag knock, making it hard to avoid false alarms.
Originally Posted by j9fd3s
this is the relevant page from the Service highlights
<picture snipped for brevity>
in the older engines Mazda determined a "Knock Zone", and the 12A-T having no knock sensor, retards timing. the 13B-T adds the knock sensor, but i think they still only tune in a specific zone (i couldn't find it, it also mentions the frequency)
the FD being newer would probably expand this zone, but in stock form would probably still not listen to the knock sensor much over a certain RPM.
this is the Rx8, it explains the system nicely, but isn't very helpful
<snipped again for brevity>
Good description in the RX8 info, except it's missing the resonant frequency/band pass. Guess Mazda doesn't like us re-engineering their cars with aftermarket EFI systems. Anyway, I recall reading from the the series 5 FC FSM/service highlights manual that the OEM sensor fro the S5 FC was a resonant type with a 3.5KHz center frequency. The operational test of it in the FSM consisted of hooking up a timing light, running the engine at idle and tapping the engine lift bracket with a ball peen hammer to verify that the timing retards a bit when you do.
So here's why I think I may have to change some of my knock control settings. As noted in the previous screen shot, I'm using the 4-10KHz wide band frequency filter because the Bosch sensors are wide band sensors, unlike Howard's setup, which is using the 3.5KHz narrow band setting, and IIRC the same Bosch sensors. The log screen shot below shows why it's really hard to set a reliable knock threshold for these things. If you look at where the yellow cursor is, you can see the rotor #1 knock level shoot up from about the 70~90 level up to 640 in an instant. So was rotor #1 knocking or was it random noise? At the time, MAP was 19.3 psi (~4.6 psi boost), at ~3489 RPMs, leading timing was a pretty conservative 15.7* BTDC and the car was smoothly accelerating without issue. This spike lasted less than 1/2 a second, and I didn't feel any hiccups in power delivery or hear anything with my unaided ears, so I'm guessing it's just random noise.
Good description in the RX8 info, except it's missing the resonant frequency/band pass. Guess Mazda doesn't like us re-engineering their cars with aftermarket EFI systems. Anyway, I recall reading from the the series 5 FC FSM/service highlights manual that the OEM sensor fro the S5 FC was a resonant type with a 3.5KHz center frequency. The operational test of it in the FSM consisted of hooking up a timing light, running the engine at idle and tapping the engine lift bracket with a ball peen hammer to verify that the timing retards a bit when you do.
So here's why I think I may have to change some of my knock control settings. As noted in the previous screen shot, I'm using the 4-10KHz wide band frequency filter because the Bosch sensors are wide band sensors, unlike Howard's setup, which is using the 3.5KHz narrow band setting, and IIRC the same Bosch sensors. The log screen shot below shows why it's really hard to set a reliable knock threshold for these things. If you look at where the yellow cursor is, you can see the rotor #1 knock level shoot up from about the 70~90 level up to 640 in an instant. So was rotor #1 knocking or was it random noise? At the time, MAP was 19.3 psi (~4.6 psi boost), at ~3489 RPMs, leading timing was a pretty conservative 15.7* BTDC and the car was smoothly accelerating without issue. This spike lasted less than 1/2 a second, and I didn't feel any hiccups in power delivery or hear anything with my unaided ears, so I'm guessing it's just random noise.
is that where it was? i own the thing on paper! lol
i'm early in my tuning, but at ~3500rpm and 5psi i see knock too, 16 degrees of timing is too much for mine, similar AFR. back of the envelope says i'm making stock FD power but at 5psi instead of 10, it helped a lot to take the 10psi timing values and move them to the 5psi row, or 10 degrees. if it was mine i'd pull a degree of timing right there, and see if anything changes. or maybe that is exactly the scenario where you'd want the ecu to pull timing for you?
since your gain is 4 and my gain is 1 your 640 is my 160. 160 is not an unusual toppish reading for me. like yours, my high reading is always an outlier. not representative of most of the readings. also being a spike, it then disappears and we are back to normal readings. i have yet to have any high reading continue to print top numbers.
i have found nothing in any of the metrics that could have triggered the reading other than my engine has mostly been too rich on fuel. i am really happy to have the ears and i resolve to believe whatever i hear and don't hear. i am tuning with the ears being the deciding factor. my tuner has used them since 2013.
" if it was mine i'd pull a degree of timing right there, and see if anything changes."
did that. i had timing down to 4 IGL/11 split at 30 psi. the car was a dog and the knock readings, or whatever is going on, did not change.
my XtremeX offers a narrow 3500 cps specifically the rotary. the closest i could get with my prior ecu (ViPEC V88) was 5.5.
Last edited by Howard Coleman; 03-02-22 at 05:46 PM.
since your gain is 4 and my gain is 1 your 640 is my 160. 160 is not an unusual toppish reading for me. like yours, my high reading is always an outlier. not representative of most of the readings. also being a spike, it then disappears and we are back to normal readings. i have yet to have any high reading continue to print top numbers.
Makes sense, and this really illustrates one of the inherent limitations of knock controls that rely solely on amplitude based measurements - it's easy to have spurious noise screw up your measurement and result in false alarms.
Originally Posted by Howard Coleman CPR
i have found nothing in any of the metrics that could have triggered the reading other than my engine has mostly been too rich on fuel. i am really happy to have the ears and i resolve to believe whatever i hear and don't hear. i am tuning with the ears being the deciding factor. my tuner has used them since 2013.
" if it was mine i'd pull a degree of timing right there, and see if anything changes."
did that. i had timing down to 4 IGL/11 split at 30 psi. the car was a dog and the knock readings, or whatever is going on, did not change.
my XtremeX offers a narrow 3500 cps specifically the rotary. the closest i could get with my prior ecu (ViPEC V88) was 5.5.
I may have to buy a Link Knock Block... sure would come in handy! The lowest narrow band filter available in my G4+ is a 4KHz bandpass, which may help me. So as a quick experiment, I switched from my 4-10KHz wide band filter to the 4KHz narrow band, and reduced my gain from 4 down to 2; initially tried setting gain to 1, but it seemed deaf below ~3000 RPMs. Here's a log snippet at the gain = 2 and narrow band 4KHz filter setting. I just took an easy drive and stayed off the boost. Knock values throughout the log never went above 60 or so, what you see here is typical - cruising in vacuum at about 4600 ~ 4700, then lifted throttle/coasting where overrun fuel cut engaged just to the right of the yellow cursor. Will see where the log goes when I give it some boost; may need to increase the gain a bit if there's not much difference.
Originally Posted by j9fd3s
is that where it was? i own the thing on paper! lol
i'm early in my tuning, but at ~3500rpm and 5psi i see knock too, 16 degrees of timing is too much for mine, similar AFR. back of the envelope says i'm making stock FD power but at 5psi instead of 10, it helped a lot to take the 10psi timing values and move them to the 5psi row, or 10 degrees. if it was mine i'd pull a degree of timing right there, and see if anything changes. or maybe that is exactly the scenario where you'd want the ecu to pull timing for you?
I'm early in my tuning too, and this is my 1st time tuning solo, though at some point I'm probably going to consult with a pro and get it on a dyno. My FC was tuned by a pro on a dyno, and at 5psi boost it runs about 21* BTDC timing and never knocks, so I thought my 16* at that point was pretty conservative; I'm also going by what I picked up from the rotary tuning courses I took on HPA. Anyway, similar to yours, my FD with a BW SXE364 turbo has about the same punch to it at 5psi boost as my FC with a BNR stage 3 turbo does at around 10psi boost, based on butt dyno.
if you have disabled the external oil pump and are premixing you do not want overrun fuel cut. the engine will be spinning with no fuel/premix and in short order you will have warped apex seals.
caveat: i no little as to when the EOP operates as in 1999 i threw mine in the trash where it belongs.
IF you have a working EOP and it adds crankcase oil on decel (don't know that it does) then MAYBE you could enable the overrun fuel cut as the motor would be getting some lube.
I'm early in my tuning too, and this is my 1st time tuning solo, though at some point I'm probably going to consult with a pro and get it on a dyno. My FC was tuned by a pro on a dyno, and at 5psi boost it runs about 21* BTDC timing and never knocks, so I thought my 16* at that point was pretty conservative; I'm also going by what I picked up from the rotary tuning courses I took on HPA. Anyway, similar to yours, my FD with a BW SXE364 turbo has about the same punch to it at 5psi boost as my FC with a BNR stage 3 turbo does at around 10psi boost, based on butt dyno.
just as a clarification, i'd literally pull 1-2 degrees and see if there is a change. based on the screen shot (which isn't the whole run) it looks like one rotor may have had something happen. what that is, we aren't sure.
point two, if we're trying to figure out what the ECU is trying to tell us, the last thing we want to do is make assumptions.
if you guys are feeling adventurous, if you run ~18ish degrees of timing and 14.7:1 afr at around 2000rpm and part load you might get some knock activity, the Rx8 does it and my P port also, it sounds like birds chirping, and you're not making enough power to hurt anything.
if you have disabled the external oil pump and are premixing you do not want overrun fuel cut. the engine will be spinning with no fuel/premix and in short order you will have warped apex seals.
caveat: i no little as to when the EOP operates as in 1999 i threw mine in the trash where it belongs.
IF you have a working EOP and it adds crankcase oil on decel (don't know that it does) then MAYBE you could enable the overrun fuel cut as the motor would be getting some lube.
No worries for providing apex seal lube during overrun fuel cut conditions. I still have a working FD electronic OMP, with the RA OMP adapter so I can run premix lube directly to it from a separate gravity feed tank in lieu of burning sump oil. It's all managed by my Link G4+, very similar to how Mazda does it - 3D table for load (MAP or MGP) vs. RPM controls the OMP output %, and the Link includes fail safes to can detect various problems in the system and can impose "limp home" protections. The only downside is driving the OMP eats up 4 auxiliary outputs for the OMP stepper motor, and 1 analog input for the OMP position sensor on my G4+.
Originally Posted by j9fd3s
just as a clarification, i'd literally pull 1-2 degrees and see if there is a change. based on the screen shot (which isn't the whole run) it looks like one rotor may have had something happen. what that is, we aren't sure.
point two, if we're trying to figure out what the ECU is trying to tell us, the last thing we want to do is make assumptions.
if you guys are feeling adventurous, if you run ~18ish degrees of timing and 14.7:1 afr at around 2000rpm and part load you might get some knock activity, the Rx8 does it and my P port also, it sounds like birds chirping, and you're not making enough power to hurt anything.
I actually did pull some timing out in that region, about 4 degrees in all, basically blending the ~10psi boost and ~5psi boost load areas together. Don't have enough log data to see if it made any difference yet WRT knock #'s, though throttle response seems a bit laggy now.
02-22-22, 08:56 AM
