rx72c

iTrader: (8)

So you let a new engine idle for 30 minutes when you first start it?

Have you built any race engines at all?

we build engines that go on the dyno the minute they start for 30psi+ boost.

the7wizard

Some people in this thread are confusing dreaming and virtual world for reality Just cause some kid types it does not mean its true

13brenova

iTrader: (2)

TitaniumTT........to be bluntly honest, I'm getting a little tired of reading about your self-serving posts. It sounds like you are a fairly knowledgeable person, but you are kind of an *** with an enormous ego. I'd rather blow up another engine than have to read another of your childish ranting posts about how everyone else is dumber than you. So.... rant and rave for a moment to repair your bruised ego, and kindly go away.

TitaniumTT

And isn't that what we're talking about here?

Let's see a recent timing map of yours with some logs and dyno charts as well, particularily interested in the EGT's......

Yup, you see a problem with that?

I build engines that don't blow up for no apparant reason.

WOW!!!! REALLY!!!! So because some little 22 y.o. kid on the internet who has been banned from a bunch of forums, blows engines, and doesn't have anything verifiable types it.... it must be true and the only way to do things

None of my posts have been self-serving nor do I have an ego, I know what I know. If you're going to ignore all the info that has been given to you, go right ahead. You'll be looking hard at putting a V8 back in soon after you keep blowing engines. Best of luck, but seriously ask yourself, who are you taking advice from? Do these people have any siteable sources or are they just spewing all over the keyboard?

classicauto

iTrader: (3)

Its possible given the higher boost, but IMO its toward the bottom of the list of possiblities.

Having blown countless engines (well countable, 6) on a HELLtech with a factory CAS, my first inclination is that it was a trigger issue.

enzo250

Well you said they were completely worthless.. I just stated one example where we were able to make 1700hp using them. So in my opinion they can't be completely worthless..


NO thanks..
Why don't you do the R&D yourself and figure out what the engine really needs and then you can post the facts here rather then assuming...

bumpstart

from my view of the trends,, any flat lines showing corruption of the trends is AFTER he has got off the throttle
one of the EGT's spikes fractionally after the data trend restores,, while the other falls
the only trends not showing corruption at any point is the timing,, throttle position and injection ms

i do not believe that this is evidence of CAS signal corruption
as it would be evident in the RPM trend as an unusual spike or dip in conjunction to a similar spike or dip in the injector period and timing trends
indicating the ECU has seen the timing ( noise ) change in its Ne and G signals ,, and reacted to it

as is,, the injector period and the timing plots don't have the irregular jumps indicating the engine has received crazy timing signals and reacted to them

instead,,, since they are after the closed throttle event,, and we have to dissimilar trends for the exhaust EGT
ima calling the corrupted trends to be signal colouration from RFI
AS A RESULT OF the effect on the spark plug as it sees differing chamber resistances now that one of those chambers is at zero compression

its noisy because the chamber has already failed
rather than the chamber has failed cause the plug went noisy and corrupted any CAS signal

and BTW,, reading and analysing PLC trends in an industrial environment , was for a long time,, my employment specialty

arghx

iTrader: (3)

Can the OP please post the actual log file? There's no sense in squinting at a screenshot

classicauto

iTrader: (3)

whats the little ~500RPM jump right before 62sec? Was there any Tire slip or clutch slip in the car?

EGT's seem to wander away from eachother after that.

FWIW, I used to get many types of RPM corruption on my Haltech, not simply spikes, but incorrect readings altogether, not too unlike the little RPM jump in the log. Mine would be making a pull, and the RPM would simply get out of synch with the engine for a second or two, not jumping around, but rising faster then the engine was -very breifly- the nreturning to normal.

It used to spike occasionally though, albeit less often then out of sync syndrome. I blew a handful of engines by having people tell me the signal was fine, or there was RFI noise, or this or that.

I went to a 60-1 motronic trigger and haven't had any failure like I used to.

flaco

iTrader: (7)

TitaniumTT hey dick relax!!! post somewhere else if your not helping

bumpstart

i do not disagree that going motronic is the solution for eliminating that tricky Vr zero crossing point analysis
( as the haltech was always intended for motronic or hall effect timing signal anyhows,, and the added Vr conditioning is just a baindaid at the front door so to speak )
in fact at the high end level,,
with high boost, high revs,, and fat coils with lots of noise
--where that high potential for timing scatter is very engine fatal
then i would go as far as saying your a mug if you dont set up the motronic or hall effect senders

however,, there is also one thing many here dont consider when they blame the ECU
anything with a dizzy hole mounted CAS is subject to its own wondering signal at RPM and high horsepower due to shaft flex and corresponding runout at the distributor drive
which makes for acceleration and decelerations of the CAS all within individual rotations
-- which particularly good and speedy processor ECU will see,, and react to

and unless you've got your trend sampling many times a second,, your never going to see it clearly

however,, i don't see the evidence of timing corruption in this example,, i see no corresponding data trend that indicate that the injection ms ,, or the timing made any significant moves at the detonation point
all isee is some minor changes in the RPM plot ( not supported in the injector MS plot )
that more likely indicates the odd miss,, and recovery

what everybody can see,, is that differing EGR trends at that off throttle point
-- now assuming there is some lag in the recording,, i think its definitely indicating that chamber temperatures rose drastically on one rotor at time of death
if you look at the boost creep trend,, it gets rather erratic in the moments before death
--indicating to me,, series of non combustion events in one chamber-- a missfire

now,, if you all read howards peak pressure thread,, he makes interesting anecdotes about the effect of timing split,, recycled EGR's and auto ignition

-- and i believe that is exactly what has happened here at point of destruction

i believe the misfire events have lead to the recycling not of burned EGR,, but of recycled,, and heated ,, unburned mixtures following a series of misfires

i believe ( further reading in howards thread of some mazda/nasa research )
that there is some evidence to suggest the preignition has much to do with trailing plug timing , temperature and the final peak pressure spike over the trailing chamber
i believe the hot plug,, and the late split ( 15 ) and the recycled unburned gasses have amounted to too high a peak pressure over that hot plug,, and auto ignition

having the trailing gap closer at say 10 degrees ,, will have initiated the combustion earlier before that pressure rose to the peak of self ignition
( my 350 HP engines all run 8-10 split at the top )

culprits-- too large a gap for the ignition voltage avail
- too hot a trailing plug
-- too wide at gap on the trailing spark split

BDC

GREAT post, Bump! This thing with the Haltech trigger has been a thorn in my side for several years. More lately I've just blamed the ECU but not knowing exactly why. On a few cars I've fixed trigger/bouncy tacho issues by installing the MSD 8509 boxes. Still don't fully understand the electronics part of it but it's seemed to fix things. /shrug

Maybe it's got something to do with the internal reluctor adapters on the older ECU's? The newer ones don't seem to have any issue from what I can tell.

B

enzo250

There's absolutely nothing wrong with using stock mazda cas.

t-von

iTrader: (8)

That's assuming the original 450 reading was correct. I don't think it was.

13brenova

iTrader: (2)

Here is the data log.....

t-von

iTrader: (8)

Do you or anyone here recall having this issue with the Fd crank set-up with their Haltechs?

bumpstart

certainly having many different evolutions of firmware updates
across many ECU platforms,, some old,, and some using different wiring to the rest ( e6x )
and then having the filter and gain settings user definable and subject to 100's of permutations at that users discretions

has amounted to an overload of misinformation regards to what will actually work for you and your unique vehicle



when the simple microtech gives the user no options to filter,, and hence all are on the same playing field
and thus it has a better rep for just getting the ignition right
( and it was always intended to be a Vr machine right from the outset )

its a matter of being ESO,, equipment superior to operator ,, as there is simply too many options and too much mis- info around with the haltech


so many blame the machine,,, when its much more likely its your unique trigger settings not quite being par


when you bring back in the MSD,, or the RA-10 reluctr adapter boxes ,, its not user adjustable,, and is firmware locked
thus your in the camp where your provided with filter settings that just work,, without the fiddle
- much like the microtechs

and yes,, some of that adaptive software in the MSD and ra-10 external reluctor adapter boxes is probably a newer update ,, and better sorted,, than what is in the machine

however,, if you have the time,, tweaking the internal filters so it is just one incremental setting away from seeing nothing whilst cranking
and re-establishing the CAS air gaps to the closed end of factory tolerances

then you should have maximum resistance against timing signal coloration
- and done this way ,, when at revs and generating as much as 100 V at the Vr signal,, it will have minimal issues at high revs with signal interpretations

its fiddly,, it doesnt work in every example due to the variations of one car to another
and many opt for the external boxes

me,, i used to go for the haltech box,,( actually back then we used a pair of RA-8's )
and haven't had the experience with the MSD one,, am keen to see how trouble free they can be

at any rate,, if your shooting for reliability for past 400 rwhp at revs ,, then you would be better jumping entirely over to the motronic trigger or hall effect crank wheels
-- at least with a haltech,, you have that user end option !

13brenova

iTrader: (2)

Does someone currently offer this for the 13b, or does it need to be a custom piece?

bumpstart

i think the rx8 has a similar to motronic missing tooth crank pulley, single pickup system, ,, that later haltech revisions will run
( unsure if it will be an update for e8 ,,but it is an option for any platinum )
- i am unsure if you could seperate this trigger wheel at the pulley hub,, not to many dead rx8 engines at my house ( none )

and fairly sure you can get the 60-2 or 36-1 wheel and pickup via autronic
but this begs the question if your haltech will run in sequential rotary modes
all whilst providing ignition split
as i just have the reservation that it may only be with a 4cyl mode
( and thus force you to run in two igniter mode,,ie wastesparked 4 cyl )

i think claudio will have the best answer,, and i have a feeling he is going to recommend the rx8 semi motronic missing tooth trigger setup
( provided your ECU will update to it )

FrankV702

iTrader: (14)

https://www.rx7club.com/haltech-forum-62/e6k-hall-sensor-timing-lock-914332/

Is this what you guys are talking about??

bumpstart

actually,, with a quick look around,, i found this thread
https://www.rx7club.com/haltech-forum-62/finally-no-more-factory-cas-861592/

in there is discussed several options ,, and the OP there has bought a BMW sourced 12 tooth wheel for the crank
and i assume is still using the CAS for the not so vital home G signal
( which doesn't need massive fidelity,,its just a flag )

of which i believe the crank trigger is hall effect,, and the CAS is Vr,, and likely going via the MSD reluctor adapter to give the haltech a 5V hall/ TTL signal without zero analysis smudges

-- a unique solution,, but one with potentially very good Ne signal accuracy
that is still giving the ECU the 12/1 it needs to run sequential rotary mode

reading it further,, claudio also hints that older machines will be limited to wastespark /4cyl non sequential operation with the 36-1 and 60-2

( which is what i elude to in post above )

PS

in that post,, the OP links the shop for said 12 tooth hall effect trigger wheel

http://store.nexternal.com/shared/St...unt2=973960518

arghx

iTrader: (3)

Here are the alleged points of data loss, boxed in orange:



Now take a look at the actual data (Note that I didn't show the EGT cells because in the raw datalog they were logged in analog millivolts):



Yeah there was around half a second missing (cells 4498-4499). It may or may not be a problem with the ECU itself, and I am skeptical on this. You can see he let off the gas around 7000rpm. Now here's the longer "flat spot"



The boxed cells are the data that showed up before and after about a two second gap. But before the data was lost here the motor had already fallen out of boost. The high sample rate and scaling of the graph is exaggerating the data loss shown in the OP's screenshot. I'm not convince that this isn't just a laptop-related issue. For the rest of the log I saw no significant anomalies in any of the major engine sensors used to calculate fuel and timing (IAT, TPS, MAP, crank signal). There's no "smoking gun" here IMO. A loss of .6 seconds as he was letting off the gas pedal just doesn't convince me, and the other loss of data isn't very irrelevent to the pull itself. I don't see an obvious cause-effect relationship between the loss of the motor and the loss of data near redline.

Besides the lack of overboost fuel cut, I still think the motor died from too much timing and too lean of an AFR for the boost levels. The plugs didn't help though, I'll say that.

j9fd3s

iTrader: (3)

be interesting to just turn the key on, engine off, and start a datalog, see if it still misses time...

13brenova

iTrader: (2)

Just finished the compression check on the engine. The front rotor was good with about 95 psi on every face. The rear rotor had about 1 psi on each face. Even if you totaly blew apart all of the apex seals, I would expect a little more than 1 psi! What the **** happened?

13brenova

iTrader: (2)

Just got the turbo out. The turbine wheel looks fine, I think? If you run your finger along the edge of each blade, you notice a fine lip. This seems like normal wear to me, but I'm no expert. There are no chunks missing, dings, or pepper marks visible. With 1 psi on every face of rotor #2, I expected massive carnage to the wheel. I'll post pics tonight.