J-series Acura/Honda

[Fred]

Logging EGT requires no code support, it's purely a hardware circuit and input to log. It's not really something you can do anything with either, other than maybe "reduce boost target" if you have boost control. It's mostly for observing in real time and logging. It's a human number. It's also the output of all other factors, so is quite predictable and a function of afr, timing, compression, octane combination.

EGT will help you with your cruise tuning, however your in boost tuning is better done by ear, a good ear.

You shouldn't be trying to fine tune anything anyway, you're yet to prove that you've tuned it at all! Get a rough tune on it, without fuss, just by using MTX as it is, and get back to us with a new log with better matching ego and lambda.

By the time you have a decent VE tune on it at 16x16 it'll be time for me to write you a decoder for it so you can run ignition too, then you can tune it properly. Any partial tune with just fuel control is only a temporary thing, keep that in mind. Just because it's infinitely better than the dodgy setup you had before, that doesn't mean it's acceptably good :-)

Fred.

[Dan]

+1 to the above post, well said Fred!

[Fred]

Re the hall sensor, totally lost. What hall sensor? I thought this car had 3 VR sensors? two on one cam wheel and one on the crank? Am I confused?

[Dan]

hey yeah? I thought it had 3 VR sensors too?

[Peter]

I put the same 36-1 trigger wheel onto the end of this cam as I did on my trucks' distributor, and I was planning on running them both off of hall sensors. It was losing the sync during cranking so I was trying to figure out why it doesn't work.

My first try at the VE table has maybe 10-15 points, and a lot of smoothing between the points. Far from complete, but better than the flat60%. RPM on the right, MAP in kPa on the bottom, and VE up the left side. I've only got up to 3200 RPM, and about 120kPa, because everything above that isn't very stable. Right now I'm planning on flattening it off at about 100% VE for the higher pressures, and then make some flat steps down to around 60% for the higher RPMs and lower pressures.

EDIT: On the LosesSync log the only sensors that are hooked up are the BRV and the MAP. The MAP isn't correct, but it's close.

[Peter]

I stepped it a little differently than I was planning, but I think it'll be alright. RPM on the right, MAP in kPa on the bottom, and VE% going up the left side.

[Fred]

Dude, this is a development system, you HAVE to keep in touch with me, or at least disclose plans on the forum, or you'll run into unnecessary issues. I can almost guarantee what was going wrong during your cranking. Check the datalog for sync loss reason ID and post here EDIT doing that now. I bet it's time tolerance based and you need config to loosen that up a bit.

VE map looks like it's starting to take shape. Did you test it on the vehicle? Get your table back to me so it can be included in the build once you're happy that it's an improvement.

Re the fake rpm and boost limiter setups, if you do that, you have to have it with 0.5RPM step and 0.01kpa step so that there is no chance of it going lean. If you have it spread out at all it goes lean as it enters that zone and you melt shit. It's not wise to have a limiter without hysteresis, but it's better than nothing. I should finish the real one later today, so upgrade to that once its available. I'll include a cut at 6k for you.

Re 36-1, i'm unsure about your hall sensors, they may not be a good choice, however you should be able to make them work. You also may need to think about the exact angle you have the missing on, otherwise it could be lost to the variation present in a cranking engine. I just spotted the log, so I guess I should open it up and take a look, one moment...

84 NearlySyncedBackwardNarrowMatchedPair
72 NearlySyncedMatchedPairNarrowBackward
66 TooWideNarrow
65 NarrowTooWide
106 NearlySyncedNarrowWideBackwardNarrowPlusMask
100 VRWiringBackwardNarrowWideMatchedPair
132 VRWiringBackwardNarrowBackwardMatchedPair
71 VRWiringBackwardMatchedPairWideNarrow
88 ExtraToothBackwardNarrowNarrowBackward
69 VRWiringBackwardMatchedPairBackwardNarrow

http://docs.freeems.org/doxygen-html/d5 ... Ds_8h.html

So, to summarise, i was wrong, not time tolerance, never made it to that part of the code: 4 codes for wiring backwards, 3 codes for noise, 3 codes for "got close" which were likely either side of the dual missing + some crank shaft acceleration. It's hard to say what is going on here, if you're using a hall sensor polarity should not matter. Some "hall sensors" are vr sensors with internal drivers, keep that in mind. Still, on those you can't reverse the wiring to the conditioner, because it's not exposed. It's a shame that I've not implemented the uber sweet logging yet, or we could get a visual on what was happening to go with these codes.

It wasn't so much that you "lost sync" sometimes while having it the rest. In actual fact, you didn't have sync very often at all, only 3 times, and never for more than a 0.3 of a second with the majority being about 0.07 of a second.

It's probably better if we just stick with the 6-2 setup until you've run out of stuff to improve with the setup, and then go to the native decoder.

Fred.

[Peter]

Dude, this is a development system, you HAVE to keep in touch with me, or at least disclose plans on the forum, or you'll run into unnecessary issues.

Well, I'm pretty sure if I would've told you that I was planning on putting a finer toothed trigger wheel onto the end of the cam you would've told me something like: You don't need that! Get it tuned, and then we can work on setting up a better decoder. And you're probably right, but this tuning business is rather tedious. Plus I kinda figured it was going to have a problem, so I didn't touch my current setup. I setup a few circuits on a bread board just to see if it would sync.

Did you test it on the vehicle? Get your table back to me so it can be included in the build once you're happy that it's an improvement.

No, I was a little too intoxicated for driving last night, and I just got up this morning. I'll have to setup the wastegate so that it doesn't start opening right away, and tune it for more boost before I send you the table.

Re the fake rpm and boost limiter setups, if you do that, you have to have it with 0.5RPM step and 0.01kpa step so that there is no chance of it going lean.

The only reason it says 6001 and 226 is because that's basically what your VE map has. In the Load axis I have 22501 for the MAP, and 12001 for the RPM axis.

You also may need to think about the exact angle you have the missing on, otherwise it could be lost to the variation present in a cranking engine.

I set the first tooth after the missing tooth on TDC Cylinder#1, because that's how the current trigger wheel is.

Some "hall sensors" are vr sensors with internal drivers, keep that in mind. Still, on those you can't reverse the wiring to the conditioner, because it's not exposed.

http://www.diyautotune.com/catalog/hall ... p-479.html I have no idea, but it gives me a really nice square wave on my oscilloscope. I thought about putting one of my extra VR sensors on this trigger wheel, but based on the shape and distance between the teeth I doubt it would be happy.

It wasn't so much that you "lost sync" sometimes while having it the rest. In actual fact, you didn't have sync very often at all, only 3 times, and never for more than a 0.3 of a second with the majority being about 0.07 of a second.

I named it that because Megatunix would go back and forth from a reasonable cranking RPM to 0.

It's probably better if we just stick with the 6-2 setup until you've run out of stuff to improve with the setup, and then go to the native decoder.

And back to my first statement.

[Fred]

you would've told me something like: You don't need that! Get it tuned, and then we can work on setting up a better decoder. And you're probably right

It's up to you if you want to fiddle around or make solid progress, however unless you like pulling and replacing engines or have amazing perfect self control, you should get it into a reliable state first, before spending time on anything else.

but this tuning business is rather tedious.

Because you're doing it in excel! Get in the car, get it free revving, get the AFRs stable and smoothish, do it with a friend, get them to drive while you tune, or vice versa. It's fun when you can see the changes taking effect and the car driving better and better as you do it.

I'll have to setup the wastegate so that it doesn't start opening right away, and tune it for more boost before I send you the table.

Minimum variables, please. Don't change that until you've got a rough tune on as it is. Then you can do that as another step. Each step should have only one thing change, whatever it is, hw or sw.

You also may need to think about the exact angle you have the missing on, otherwise it could be lost to the variation present in a cranking engine.

I set the first tooth after the missing tooth on TDC Cylinder#1, because that's how the current trigger wheel is.

Hmmm, still, if you can adjust it, you may get better results at other angles. Somewhere I have a screen shot and log of an arctic snow cat cranking and I couldn't visually make out where the missing tooth was due to the compression effects.

I have no idea, but it gives me a really nice square wave on my oscilloscope.

Hmm, is the missing clearly visible?

I thought about putting one of my extra VR sensors on this trigger wheel, but based on the shape and distance between the teeth I doubt it would be happy.

Likely correct.

Fred.

[Peter]

It's up to you if you want to fiddle around or make solid progress

I feel like a better trigger wheel is an accomplish-able goal. Tuning is more of an art, and I have no real way of knowing it's done.

Because you're doing it in excel!

Well, in eclipse using OLV. I was only using excel to graph it, but you're right Megatunix makes it a little easier.

Hmmm, still, if you can adjust it

Highly adjustable until I weld the trigger wheel onto the cam wheel.

Hmm, is the missing clearly visible?

Yep

The idle area was to lean to run so I took that area back up to a flat 60% for the time being. It seemed to run a little better in most areas, and runs way better when the turbos spool up. Megatunix didn't give me any crap. It loses sync for reason 65 when I send some new variables to the ecu, but it idles through it.