Have you measured when the injectors...

[djuosnteisn]

I have to go to my nephew's 4th b-day today. While i'm out i'll pick up the scope and measure the intake cam. Hopefully have some data by tonight.

Didn't make it to the office to grab my scope, but i did wire up the cam, so i'll drive my car tomorrow and measure it during lunch.

[djuosnteisn]

Okay, i logged the cam signal, but i'm not sure how the cam sensor works (it doesn't look like a missing tooth like the crank at all.) I don't have my camera cable, so i'll have to post em up tonight after work.

Does anyone know if a "double post" merge flags the thread as updated? Or does it just edit the last post and not notify anyone?

[Lex]

Cam sensor is also a toothed wheel. But is has a lot less teeth and is therefore slower to update. Don't have the exact pattern in front of me right now, can get it tonight.

[djuosnteisn]

Sweet. Definitely looks like less teeth. One of them teeth looks like it may mimic the opened valve duration, but that's entirely speculation. We'll see what we come up with tonight.

[Darksun280]

Originally Posted by djuosnteisn Okay, i logged the cam signal, but i'm not sure how the cam sensor works (it doesn't look like a missing tooth like the crank at all.) I don't have my camera cable, so i'll have to post em up tonight after work. Does anyone know if a "double post" merge flags the thread as updated? Or does it just edit the last post and not notify anyone?

Edits last post and doesn't notify.

[djuosnteisn]

Okay, so here's the cam measurements.

Again, for clarification, i used PCM pin 2E from this diagram:



Here are two screen shots showing the number of crank tics (6 degrees each) between TDC and the cam sensor (what i guess is the opening point of the valve).

Stopped the scope while the DashHawk was reading VVT of 30:


Stopped teh scope while the DashHawk was reading VVT of 0:



Here's a video with me varying between cruise and part throttle varying the VVT from 0 to 30 to 0 etc on the DH. You can pause the vid and count the ticks.

valvetiming0to30to0.flv video by djuosnteisn - Photobucket

valvetiming0to30to0.flv video by djuosnteisn - Photobucket

Here's a 3rd gear pull and an accompanying log:

3rdgeardatalog1.flv video by djuosnteisn - Photobucket

3rdgeardatalog1.flv video by djuosnteisn - Photobucket

Here's another 3rd gear pull with the scope zoomed in a bit & a log:

3rdgeardatalog2zoomed.flv video by djuosnteisn - Photobucket

3rdgeardatalog2zoomed.flv video by djuosnteisn - Photobucket

And for giggles here's just one other 3rd gear pull since i have it:

3rdgearpull.flv video by djuosnteisn - Photobucket

3rdgearpull.flv video by djuosnteisn - Photobucket

From what i gather, 06speed6 was right about the 16 degree mechanical limit on the VVT mechanism.
-When ever the DH reported 0 degrees, the crank edge was right at TDC, or at least withing 6 degrees (resolution of a crank tic).
-When the the DH reported 30 or so, the crank edge was about 3 ticks BTDC (or 18ish degrees).

Discuss away.

Anyone with any idea how to decipher the cam signal, please offer up some suggestions. I think it's intuitive enough that the falling edge in the vids an pics corresponds to valve opening, but not sure what to make of the other stuff.

[Lex]

DJ, great info so far, do you have a picture showing cam over 2 crank revolutions? (So we get 1 complete cam revolution)?

[superskaterxes]

why does your inj/cyl go up so steeply after say 5500 rpm? is that because of your commanded afr dropping at that point?

btw thats like the sickest log ever. perfect afr, perfect boost, shit looks pristine.

[djuosnteisn]

Originally Posted by Lex DJ, great info so far, do you have a picture showing cam over 2 crank revolutions? (So we get 1 complete cam revolution)?

I thought one of the videos showed it, the zoomed out ish 3rd gear pull. I can definitely take another one though if you need. It looks like the second half of the cam sensor is simply inverted from the first half.

Originally Posted by superskaterxes why does your inj/cyl go up so steeply after say 5500 rpm? is that because of your commanded afr dropping at that point? btw thats like the sickest log ever. perfect afr, perfect boost, shit looks pristine.

injection goes up like that cause my maf curve is exaggerated around that voltage. It's new since i switched over to a MBC, and just haven't had a chance to fix it yet.

Oh and MBC FTW! Don't get me wrong, gear based boost/load control is awesome, and for FWD i'm sure it's an absolute blessing. But with AWD, i just want the full boost when i ask for it, so why not make life simple, plust the benefit of not having to flash a new map to run different boost. And also, not having to use a pill seems to have helped my creep issues slightly.

Here' ya go lex:

[Lex]

Ok, here is the CAM sensor deciphered. Don't mind my handwriting and colors used. You had it right DJ.

So what we learned is that VVT is in terms of crank degrees, not cam degrees.

[djuosnteisn]

Nice non-photoshop skills!

Still though, i don't think the cam advances any more than 16ish crank degrees, at least that's what i see in the pics and vids. I was counting 3 crank tics, and that's still only 18 crank degrees right?

[Lex]

Originally Posted by djuosnteisn Nice non-photoshop skills! Still though, i don't think the cam advances any more than 16ish crank degrees, at least that's what i see in the pics and vids. I was counting 3 crank tics, and that's still only 18 crank degrees right?

Correct. each crank tooth (from rising edge to rising edge) is 6 crank degrees.

Can you send me this log to my email please?


Also Dustin, what is your commanded AFR map look like for WOT?

[jmhinkle]

I told you that MBC rocked. Your boost curved looks wicked on those logs. Not showing any creep there at all.

[djuosnteisn]

Originally Posted by Lex Correct. each crank tooth (from rising edge to rising edge) is 6 crank degrees. Can you send me this log to my email please? Also Dustin, what is your commanded AFR map look like for WOT?

Log emailed.
Commanded afr's are mid 11's to redline. I'm not entirely sure what's going on with the drop near redline. I don't know why the maf cal would change with or with out boost creep, it shouldn't have any effect on a proper maf calibration. Only thing i can think off (and it's a stretch) is that the lines that used to go from the solenoid to the intake are now just blocked off, and i no longer have a pill between the boost tap and the WG, so maybe it acts like a bit of a boost leak. It just needs another calibration and i'll be money.

Originally Posted by jmhinkle I told you that MBC rocked. Your boost curved looks wicked on those logs. Not showing any creep there at all.

Yeah, getting rid of the pill really seems to have helped! Still a bit during colder temps, but much better. Also, creep isn't as noticeable when you keep upping the boost, hahaha. When i'm running 24psi, i'll have no creep at all!

[Lex]

I have a feeling there's another variable at work past 6000RPM in terms of fueling that we don't have access/visibility to.

[superskaterxes]

Originally Posted by Lex I have a feeling there's another variable at work past 6000RPM in terms of fueling that we don't have access/visibility to.

im going with this one as well........

just too much weird random shit going on after that magic number

[06Speed6]

I told yall the VVT only had 18* of movement.

If the ECU table is in crank degrees then we should be able to put 36 in the fields.

[jmhinkle]

Originally Posted by Lex I have a feeling there's another variable at work past 6000RPM in terms of fueling that we don't have access/visibility to.

I agree 100%, because I see a common theme with logs of that dump at the end no matter what the commanded AFR is.

Originally Posted by djuosnteisn Log emailed. Yeah, getting rid of the pill really seems to have helped! Still a bit during colder temps, but much better. Also, creep isn't as noticeable when you keep upping the boost, hahaha. When i'm running 24psi, i'll have no creep at all!

You have no way of measuring 24 so be careful there unless you got the gauge installed. You'll be like the time I was running 24 on an 18 setting on the SB against you and pulling away. LOL

[Haltech]

Im going to move this great thread to the more appropriate fueling section of the gen 1 area. There will be a perma link... Carry on

[djuosnteisn]

Originally Posted by 06Speed6 I told yall the VVT only had 18* of movement. If the ECU table is in crank degrees then we should be able to put 36 in the fields.

But i'm only counting 18* of crank angle, not cam angle. I'll have to do some more pics and stuff to verify i guess. Won't be till later this week, or possibly weekend. I'll get a pic with the DH and the scope in the screen. Actually, i'll use the AP cause i know there is some suspicion with DH accuracy on some parameters. AP is rock solid.

Originally Posted by Lex I have a feeling there's another variable at work past 6000RPM in terms of fueling that we don't have access/visibility to.

What else should i try to measure to help see it? Do you think a direct measurement of the injector line would shed some light? It would at least rule out any possibility of the injector driver module doing something weird.

Think about some other things we could measure to give us some insight, and i'll definitely do it.

One other thought is do you think it could be a misleading O2 reading from un-burnt, un-properly atomized fuel? Like maybe we're drowning the O2 sensor in liquid fuel instead of fuel vapor?

Originally Posted by jmhinkle I agree 100%, because I see a common theme with logs of that dump at the end no matter what the commanded AFR is. You have no way of measuring 24 so be careful there unless you got the gauge installed. You'll be like the time I was running 24 on an 18 setting on the SB against you and pulling away. LOL

I was exaggerating quite a bit on the 24 lol. At least until i get all my hardware installed and dyno tune. Until then i'm gonna stick with 20~21. And i finally have a boost gauge on the way (lol at my lazy ass doing all this without a real boost gauge).

Lex, perhaps i should have logged my throttle position, just in case it was somehow closing (like ecu was infringing on the max load - throttle close, though i highly doubt it i was, nonetheless, we need to be certain). From now on i will be doing all logging with the ap, since it logs so many more parameters at once.

[fjames]

Originally Posted by djuosnteisn Lex, perhaps i should have logged my throttle position, just in case it was somehow closing (like ecu was infringing on the max load - throttle close, though i highly doubt it i was, nonetheless, we need to be certain). From now on i will be doing all logging with the ap, since it logs so many more parameters at once.

I've seen it use both throttle and pedal position (at the same time) to presumably manage load (boost was within limits.) I always log throttle duty cycle and it varies a lot over a small area. Could just be my tune, but it seems obvious it goes to great effort to manage load. You can't talk about this ecu without talking about load, but people still just talk boost ... weird.

[djuosnteisn]

Okay, so today i had the valve cover off for other reasons (one of which was to paint it). I decided to find the cam sensor, to help explain the odd log i saw above, with the weird inverted signal thing.

This is our cam sensor. It's a magnetic hall sensor that pics up the metal gear below it.


This is where it goes in the cam section:



Oh, looky here, what a funny shaped gear:


Here's 2 other views. Kinda hard to see, but the teeth on this gear exactly mimic the traces in the sensor.





In the first gear pic, you can see cam for cylinder #4. It's pointing toward the front of the car, which would mean cylinder #1 is point toward the back.

So this justifies the odd shaped trace in the VVT scope video above. I believe it's accurate, and accordingly, we do only have 16 ish crank degrees of adjustment (even though my dash hawk said 30). Also, a VVT reading of 0 means the intake valve is opening up pretty much right at TDC (within the 6 degree resolution of the crank sensor). A VVT reading of 30 corresponds to the intake valve opening up 16 degrees before TDC.

If i'm wrong in any of this, someone lemme know, and i'll edit this post. Thanks!

.


LOL at me doing things the really hard way. I should have just looked at some of the scans a buddy sent me a while back. Oh well, at least we know it's true.

[Lex]

Good info, so does that mean we were not counting the number of teeth correctly and it truly is 30 crank degrees (15 cam) of advance?

[SpeedSixxx]

sub

[djuosnteisn]

Originally Posted by Lex Good info, so does that mean we were not counting the number of teeth correctly and it truly is 30 crank degrees (15 cam) of advance?

I'm not entirely sure. I know the DH is reporting actual advance, and that's the confusing part.

If you have any other ideas for experiments, i'm game. I may do another scope drive next week sometime and just take a better look at it.

[06Speed6]

That fairly closely matches the 18* of VVT advance that I measured, but it is possible that not all of that is useable in the real world.

[SpeedSixxx]

keep up the research guys....!! ppl like you make this forum stay true and alive!!!

[Richard L]

Originally Posted by djuosnteisn Also here's a video showing the control signal varying with fuel injection amount. I'd prefer to just have to log this control signal from the ecu cause then i wouldn't have to run a cable from the engine bay. My car is starting to look like a fucking time machine, lol. If anyone thinks this is not adequate, lemme know and i'll stop bitching and do what needs to be done. Once i get some decent logs, and we settle on exactly where TDC is in the crank signal, we can start talking about what implications this data has on our platform. From initial looks at it, it looks like the spray starts pretty close to TDC on the intake stroke. So maybe lenny's cobalt buddies were right in saying over lap can rob power by pushing fuel out the exhaust. I'm not sure, but we'll find out.

Great information!

I have a question please. The above video shown a nice clean ECU control signal for the fuel injector, can you tell me which ECU pin you pick this signal from.

[JumpingJackson]

Richard L, your new here are you working on something for us ECU wise? Oh wait you work for like Aquamist or something! i recognize your name from the NASIOC forums haha

[djuosnteisn]

Originally Posted by Richard L Great information! I have a question please. The above video shown a nice clean ECU control signal for the fuel injector, can you tell me which ECU pin you pick this signal from.

All the pinout info is in the first post , here you go:

ECU injector control line to injector driver
cyl 1 - 3B
cyl 2 & cyl 3 - 3E & 3H (haven't confirmed exactly which is which)
cyl 4 - 3K

And you can use this screen shot for color of each wire.

http://www.npcompleteperformance.com...full;init:.jpg


PM me if you have any other questions or need me to do anything!

[Richard L]

Originally Posted by JumpingJackson Richard L, your new here are you working on something for us ECU wise? Oh wait you work for like Aquamist or something! i recognize your name from the NASIOC forums haha

I don't think I have the ability to work on the ECU. I wanted to know how the MS3 manages its fuel delivery. So many new cars are moving towards direction injection. The MS3 is an odd ball with the paralleled injector(+) between two adjacent cylinders.

I am working for Aquamist. My main job here is to look after the end users and support our dealers on new or old installations. I do not sell directly except during a new product launch.

I am here to learn as much as I can so that I can contribute later. This forum is the most technical one I have ever come across.

[Richard L]

Originally Posted by djuosnteisn All the pinout info is in the first post , here you go: ECU injector control line to injector driver cyl 1 - 3B cyl 2 & cyl 3 - 3E & 3H (haven't confirmed exactly which is which) cyl 4 - 3K And you can use this screen shot for color of each wire. http://www.npcompleteperformance.com...full;init:.jpg PM me if you have any other questions or need me to do anything!

Thanks for the image. I have done a screen shot for future reference.

Ny question was about the fuel inejctor signal. It appeared the previous image shown a positive pulse but the latter iamge shown a switched negative pulse, am I missing something?

My interest on this is work related. A few of our MS3 customer wanted to use FIDC to meter water/methanol flow so I am here to gether information so that I can help them.

Richard

[Richard L]

I think I know what was happening, I have mistaken the cam sensor (green trace) as fuel injector signal. I should have paid more attention. The injection signal was used for the trigger input.

OK, back to the injector deciphering work.

[SpeedSixxx]

Originally Posted by Richard L I think I know what was happening, I have mistaken the cam sensor (green trace) as fuel injector signal. I should have paid more attention. The injection signal was used for the trigger input. OK, back to the injector deciphering work.

u making parts for us? injectors?...

explain

[Richard L]

If you are referring to fuel injectors, no we cannot make those parts.

I work for aquamist, a company that makes water/alsohol injection system. I am very new to Mazdaspeed.

[SpeedSixxx]

Originally Posted by Richard L If you are referring to fuel injectors, no we cannot make those parts. I work for aquamist, a company that makes water/alsohol injection system. I am very new to Mazdaspeed.

welcum

[Richard L]

I can help with any water/methanol questions but I will start up another thread in a few days time.

[ptperformance]

Originally Posted by Lex This is unfortunately incorrect. The double signals are from the fact that the positive signal lines on inj 1,4 are tied together as are 2,3. There is no specific signal to close injectors, they simple close once the hold period is over. The negative signal lines are independent on all the injectors. This is shown in the FSM section attached. DJ, when measuring injectors, please measure both the signal at the injector and the ECU signal as well. The injectors are also around a 70V signal, not 100V.

I hope that this has been addressed but the injectors are an open and then close fire. The open volts are 65 and close is 4x something. The FSM states this clearly, so FYI.

[SpeedSixxx]

Originally Posted by Burn69 Good Information...Thanks

another loser trying to rack up his post #'s...

fuck off and donate. you keep this nonsense up and ur gonna get spanked.

[SSinstaller]

Originally Posted by SpeedSixxx another loser trying to rack up his post #'s... fuck off and donate. you keep this nonsense up and ur gonna get spanked.

+1

If I see it much more, I'm going back through his post history and deleting all the BS..