IceY2K1

iTrader: (3)

Well...it wasn't as easy as I thought, but it does work. It takes a bit of playing with the Add Injector map to compensate especially to handle the load difference in 2nd vs. 3rd gear. This would have been A LOT easier on a dyno instead of tearing up the neighborhoods in 2-states and 3-cities over the holidays, while keeping the cops off my back. Only got stopped, actually I was already stopped, once.

Anyways, here are 2nd vs. 3rd gear snapshots from the same setup/maps with a 6600rpm limit set....

2nd gear:


3rd gear:



As you can see...the ORANGE injector input(ECU output to EU) duty cycle falls at the stock limiter(~6492-6500rpm), while the dark BLUE injector output(EU output to injector) holds fuel all the way until the ~6600rpm EU limit. The other nice thing is that you can control the amount the EU cuts IPW. The stock ECU fuel cut drops to ~1.x ms and stays flat, which feels pretty hard on the engine. The EU fuel cut feels like a slight hesitation or stutter and to me feels racecar ignition cut like.


It is hard to see in this shrunk down cardomain reduced pic, however it shows 15-degrees of timing pulled at 4000rpm on up where the vertical red line crosses and the black line falls to -15degrees. You can see the slope of the RPM red line change and you can definitely feel the car fall on its face:


However, zoomed in shows the RPM line gets rough at ~4000+rpm due to what is either the ECU/EU fighting or the EUs lack of precise ignition timing without the crank/cam inputs:


I'm not 100% sure on the timing until I or someone gets one on a dyno, however it did seem to improve the curve when I added upto 10-degrees(post factory rev-limit though) and it definitely changed the RPM curve slope when I pulled 15-degrees.

Anyways, the maps aren't perfect yet, not even close, and without a WB02 and a dyno, they won't get much better. I'm not sure how much more I'll push once I steal a laptop again, however a few bounces saw 6700-6800+rpm. Higher RPM are definitely possible, I just didn't feel like pushing it further without a WB02 or more like likely less holiday 5-0 patrols.

Kevlo911

iTrader: (36)

For the timing, add some at idle. If the rpms increase, the timing is being advanced.

Velicos

Good to hear this worked! I need to fix a few minor things with my EU and then I'm going to attempt it. Could you post your settings you used for the Rev Limit map please?

IceY2K1

iTrader: (3)

I'm not sure that's a good idea...

You need to log your stock limiter bounce and then we'll go from there. Mines' not finished by any means and you probably don't have the same car/conditions.

Velicos

I fully understand that, I'm just curious to what yours looks like.

I've been talking to DandyMax about it for the past week and we're both very interested in this.

matty

iTrader: (54)

So do you think its time I order it or should I still wait ?? Crank signal work yet ??

I would like to get this installed right away, and get it up and running, maybe on a warm day...if not then in the spring

DandyMax

iTrader: (7)

So it seems timing retard works... that's good, not sure that advance will but who knows.

How did you get a timing value to display? Still using tach input for rpm? The timing value wasn't supposed to display properly without the crank hooked up. Maybe that got fixed for 1.09 (I never got to try that version).

I don't mean to sound negative but I wish you'd gone more than 100 past the limiter. Looking at your graphs it looks like it'd work in theory... although I'd like to see it verified that it can actually hold much farther into the 7000's. Even on a stock ECU it's easy to bounce up a couple hundred past the limiter if you're holding the pedal on the floor.

Now hopefully, there's no difference b/w the 5th gen ECU and 4th gen (ie-that for some reason it wouldn't work on 4th gen). What did you have the hold set at... 6492, and 5ms wait time?

F**K!! I'm extremely pissed at myself right now - and Greddy. They told me that ign inputs would work just as well as tach input. And like an idiot I believed them thinking they should know their product. I shoulda known better. Sometimes I honestly don't know where my brain goes to. I really wish I could have kept playing with this instead of running out of time and having to stick the car away for the winter. In the little time I had at the end of the season I was much more focused on working with Greddy trying to get the crank to work than this stuff. If we'd gotten that signal(s) working all these other issues like timing and rev limit would likely have worked automatically.

Glad to see someone else has been able to carry on though.

Now go and get that thing up to 7200 Alex! Just add a lot of extra DC past 6500 and you'll be safely rich enough.

MardiGrasMax

iTrader: (14)

As long as you dont hold it up their it wont get dangersouly lean beween 6500 and 7000 or so.

Nice work!

IceY2K1

iTrader: (3)

That's not actual timing just Timing adjustment map.

It will go as far as you want, the reason I chose 6600rpm was because I was having a hard time not hitting 6800+rpm. Due to the weird interpolation(read screwy), it adds fuel past the EU cutoff point and then when the RPM drop, the EU limiter kicks in again bumping RPM back up. It's hard to explain, but basically the RLC map cuts in and out and the Add map fills in the dips(most of the time). I'll try and post a better log to show this.

I have the hold set at 6200rpm and the limiter at 6600rpm. The add injector map columns is the hard part to keep the injector from going static(spike before stock cutoff), yet hold enough fuel to prevent ANY dips. I mean the dips you're looking at are very small 5-10ms in length, but I don't want any. The 4th gen will work just the 5th gen I'm sure. All we are doing here is adding IPW after the cutoff, so I don't see why not. A difference could be the cutoff fuel amount and whether the ignition timing is reduced to match the cutoff IPW. Timing is calcuated based on IPW/RPM per my FSM IIRC, however I didn't know if that was for both closed and open loop maps. IMO, that is why I'm having difficulty with RPM dropping after stock cutoff. I've added up to 10-degrees(5-degs didn't help) and it seemed to improve. I really need to be able to see actual timing to compensate.

Don't be pi$$ed...this took 30-40 rev bounces, one cop license/reg/insurance, and I probably cut my catalytic convertor life in half. Haha. Anyways, we DEFINITELY need the crank input to more precisely track RPM and adjust timing. The ignition inputs are crude and the variability is what was killing me as I was trying to figure out why the interpolation isn't working as expected in the Add Injector map.

The problem isn't going lean, it's going rich. My biggest hurdle at first was keeping the injectors from going static, ie 100% IDC or 20+ms ID. The engine doesn't like that and now I know what it feels like. I can go further, I just didn't have time to push further. I'm sure Mingo or someone will help soon.

IceY2K1

iTrader: (3)

Just for the fuel and Rev-limiter, I think it's worth it to buy now. However, if you MUST pull timing, you may want to wait until someone tests this further.

Sorry, but I don't want to be wrong and have you get pi$$ed, so it is really up to you maaaang.

Nealoc187

iTrader: (23)

^ I agree with what mardi and dandy said, though it is of course your car and you can do with it as you please and feel comfortable. I just don't think that only a 100rpm increase is quite sufficient enough to say that the RPM limiter increase works. It could be accounted for by some sort of lag on the EUs part, or it could be that it indeed works. To put your mind at ease some, advanced timing is less of an issue as you already know at higher RPM, so don't be scared of the timing getting increased too much up there (tq peak is where you need to worry about adv timing the most), if you are worried about retard and high EGTs just remember the crazy retarded timing JWT SC prog has at even higher RPMs and feel safe, and it'd be such a short time that EGTs would barely have time to go anywhere anyways. If you are worried about AFR just increase the IPW some to so that you are sure you are safe. You can calculate how much IPW (fuel flow) you need to maintain a certain AFR (roughly, definately close enough to not be dangerous). At this point what I'm most interested in is seeing if the rev limiter works, definitively. If I saw the EU IPW extending to 6800 or 7000 I'd be convinced, but the 100rpm increase is just so miniscule as to not quite be convincing, but enough such that it appears it indeed works.

If you have the time and inclination, would you mind maybe messing with the 2 step limiter some? Sharif said it was pretty easy to hook up, just requiring a speed input. It's also something you can do of course without having to worry about Smokey chasing you down because you're going 90mph over the speed limit lol.

Awesome thread alex keep us updated on what you find, I know you will.

IceY2K1

iTrader: (3)

You guys need to play with this and you will understand why I say you can hold to 6800 or farther all day long. Just add fuel/timing where you need to. It wasn't that I was "scared" to do anything, I did 15-degrees advance(post limiter) and 100% IDC with no problems, I just had bigger fish to fry then to see what RPM I could get too. I know it can easily make 7000rpm without a sweat and I'm not interested beyond that.

Here is the juggling act between the AddInj and RLC map I was trying to explain Dandy....

AddInj map fuel bump is the 11ms+1.2ms:


RLC map fuel bump is the 12ms:


Also, if you guys need more convincing, here is what the stock limiter log looks like where you can CLEARLY see what is going on with RPM and IPW. Those valleys/dips are down to 6100rpm, that isn't happening with my maps:

IceY2K1

iTrader: (3)

I have catalytics right now Neal, so I'm not touching that 2-step.

DandyMax

iTrader: (7)

Oops. My bad. Somehow when I read it my brain interpreted it as "timing" not "timing adjustment".


Hmm... looking at those last logs I think I'm getting an idea of what you're talking about. I do see the difference you point out between the 12.2ms and 12.0 ms durations. So let me see if I understand what you're trying to say: You're thinking that once the rpms get up to the new EU limit (6600), the RLC map releases control and it reverts back to the AddInj map. And depending on how much PW/DC you have added (the net PW would in all likelyhood be less than the last point on the RLC map), then the rpms will fall back again below the 6600 limit, at which point the RLC map kicks in again and up the rpms go until the limit is hit, the RLC map kicks out again and the AddInj map kicks back in and so on and so forth we go around the merry-go-round.

Is that more or less what you were getting at? If this is indeed the case, then it's not really a hard limit.. ie-not like an ign cut, just a very soft cut. In fact, theoretically, you could almost just keep climbing then couldn't you if the AddInj map continued past the RLC map with sufficient PW additions (assuming you don't hit 100% DC)? In reality though why would you do that, it's far better to extend the RLC map far enough.

But there's another thing here and maybe it's not a big deal but it just occurred to me: if the new limit on the RLC map is hit/exceeded and subsequently control reverts back to AddInj mapping, then wouldn't you momentarily be running pretty lean right up at high rpms if you didn't crank the adjustment way up (since the stock ECU is providing virtually no fuel)? Could that cause problems on FI cars? You'd have to add an awful lot of PW to stay on the right A/F but in that case then you're pretty much defeating the new limit anyways being prone to "climibing" like you mentioned and it wouldn't feel like much of a cut/stutter.

Maybe I haven't wrapped my brain around this completely yet -it would be a lot easier if I could actually play with this on my car, and all my statements are based on the assumption that the above theorizing is reality- but I'm thinking that you would not want to use the RLC map as the absolute highest rev cut on the car. The best thing would be to use it to go past the stock limit, but set it farther than you actually want the new limit to be, and then set the REAL new limit with the IGN cut feature. That way you have a hard cut that's safe (except not so healthy for your cat), and you'd never "climb" high enough to have the 2 fuel maps switching control back and forth, thus eliminating the need to screw around so much up high on the AddInj map. Does that make sense at all?


I'm not sure I understand exactly why you're getting those spikes right before the stock cutoff. Can you explain? Judging by your rpms shouldn't you already be into RLC map control at that point if your hold is 6200? Am I missing something obvious?


Well I guess I was just kinda ticked that I could have maybe done a bit more with the rev limit stuff before stowing the car in storage. But I will still follow through and get that crank update for sure. Also, it appears there is another 4th gen out in CA now with the EU and he lives near Greddy too! When the update is available maybe we can get him to do some testing too...

Oh sheesh look how much I wrote. Sorry for the long post. It's sometimes hard to put things clearly without getting wordy.

DandyMax

iTrader: (7)

Neal my understanding is that it only requires the speed sensor input (which is a cinch to get). Then all you do is tell the EU the conditions for the limiter to kick on/off (and the desired rpm of course).

It will only come ON if it does not detect any speed sensor pulses (no wheel rotation) AND if the throttle position is greater than some % you set (range is 50-100).

It will kick OFF once it detects a certain number of pulses again (how many is up to you - the range is 0-40). I assume this is meant to account for some wheel spin off the line, but I haven't yet taken time to find out know how many rotations 40 pulses would correspond to (ie-how many pulses per rotation).

IceY2K1

iTrader: (3)

No, the RLC map by itself, ie with no Add map, would do the same "square wave" from the stock cutoff to whenever you set the RLC limit point. I believe this is do to the EU 5ms WaitTime limit, however I'm actually using 0.8ms since it seemed more consistent. Basically, when you see the 12ms bumps, that's the RLC map adding fuel and when you see the 12.2ms bumps that's the Add map + small duration the stock ECU is still outputting.

See above.

Yeah, kind off, I've been trying to deal with this issue, ie once RPM drop, you land in a lower cell of the fuel map and it adds more fuel so you keep going. However, if you set your say 6600-7000rpm Add map cells with a 6600 RLC limit to 5ms, you can feel it and it's much better/safer then the stock ECUs 1.2ms->12ms->1.2ms..... At least in my opinion.

Add Injector map interpolation. The imprecise RPM I think is the culprit. Basically, the EU is dumping fuel too early or too late depending on if I put my AddInj column at 6450, 6500, or 6550. This has been my headache. I don't know why it does interpolate correctly sometimes and others it jumps to the next cell when it shouldn't. I'm hoping it's just that the logging isn't showing a fine enough resolution and with the crank input this will be more accurate/reliable.

The RLC map doesn't react fast enough, so there is a small dip sometimes. It's probably not a big deal, but inorder to make sure the inj duration output never drops below inj duration input, I have the AddInj ramp fuel.

DandyMax

iTrader: (7)

OK. Finally, I think I understand what's going on now. And I also have a bit of news. I just had a conversation with Kenji.

He said that on many other cars, the stock ECU will cut fuel at the limit point but only for a short time (during which the rpms would fall obviously). However if by some other means fuel was provided, at rpms higher than that limit the ECU would resume fueling. Therefore the RLC map was only designed to operate (ie-wait out) that short time of fuel cut.. to bridge the gap until you're past the stock limiter and the INJ map can be used again with the resumed ECU fuel signal.

So on a such a car, the variation in values in the INJ map would be much smaller, and you wouldn't have as wide an interpolation gap. However our ECU's provide a continuous fuel cut post-limit, so it would seem we're forced to add tons of fuel into the INJ map to keep the PW's up there.

Can you post your RLC and INJ maps Alex (or email them to me)? I understand which maps the 12.0 and 12.2 ms PW's correspond to but I'd also like to see the maps themselves to correlate better with what I'm seeing on your log snapshots. I still think it would be better to go quite a bit higher than the stock limit to be able to more clearly see how the EU responds. I'm wondering if the RLC map is kicking in and out (square wave effect) continuously past the HOLD point because you're too close to the stock limiter and/or the inaccuracy of the rpm pickup. Let's say with a crank signal, and a LIMIT value set quite a bit higher than the stock value, the RLC should only kick in at the HOLD rpm, and again at the LIMIT rpm shouldn't it? In between those points the INJ map should be controlling. That seemed to be the way Kenji described it...

Now you also said there was a bit of a delay when the RLC kicked in? Perhaps that also has something to do with not having precise rpm info from the crank? I dunno, but I mentioned some of these things to Kenji and he will talk to Japan and see what they say. I didn't get the impression Greddy USA is fully knowledgeable about this feature either; Kenji and Benny are going over to Japan in the next little while to get some training from the EU designers & programmers themselves.

Now for the good news: apparently Japan has finished both the software/firmware update and the add-on hardware module for the VQ30DE/DET. It has either already arrived at Greddy's offices in CA or it is in transit. Kenji mentioned there had been some packages arrive from Japan recently but he was away and needs to check on where exactly the module is currently. I should know more in a couple days but it's looking like I will have this in my hands soon.

Additionally, their NJ location is now operational and, if for some unforeseen reason it should become necessary (I don't think so but you never know), this is within driving distance for me to personally go down there and meet with Kenji, do testing, etc.

IceY2K1

iTrader: (3)

Exactly....

The EU seems to interpolate between the HOLD and LIMIT pulse widths, so I think it is actually trying to control the fuel BETWEEN the hold/limit points not just at them.

I really suspect this is because of the imprecise RPM ignition input, especially since it varies between 2nd/3rd and even between runs with the exact same maps. The loading of the engine due to gears and/or the road causes the stock fuel cutoff to vary a few RPM per the EU logging. Sometimes the EU catches it in time, but sometimes it is late. That is why I ramp fuel just before the stock fuel cutoff, ie 6450-6500.

JOo better get TWO and send one my way~! Can you please CC me on your guys emails?

IceY2K1

iTrader: (3)

Also, ask Kenji why the hell the EU displays anything that is x.50 in the maps as x.4 in the logging, ie 11.50->11.4 and 12.50->12.4.

Don't even want to know what happened when I put "-0.5" in my AddInj map...OMFG. Pulsewidth logged as 40-50ms, however that's not possible and didn't seem to be actually occuring. Yikes~!

DandyMax

iTrader: (7)

Well I gotta say this is all making much more sense now. Some days it takes me a while for the gears to get turning. lol. You really have to study things to see what's going on, and that's always easier to do having gone through the whole in-depth exercise physically with the car rather than looking at pictures afer the fact (as I'm sure you'd attest to with the 30-40 rev bounces). lol.

Thanks for the maps. They look pretty much like I expected but just wanted to verify.

Hmm well if this is the case, then it would suggest that the design is such that the INJ map is always active, but when the RLC map activates, it overrules and takes precedence. Aside from not having the crank signal, do you think the issue is just the wait time then - it's not staying active long enough? It'll be interesting to see what Japan says. I have asked Kenji to ask them specifically about this case (ie-continual fuel cutting ECU).

Haha, well I don't know how many they've sent this first time. They probably only made one or two, but I'll do what I can to get a few more out to the guys who've jumped on this early like you and Mingo, Jime etc. I think Japan is providing Greddy USA with the schematic so they'd build any future modules in CA.

I have for the most part been talking to Kenji on the phone, I find I get much better response this way but sure any emails I'll CC you.


Yikes! Sounds like just a programming bug (ie -the EU is functioning properly just not displaying and/or logging the value correctly). I will pass it on.

IceY2K1

iTrader: (3)

I'm not sure I agree that's a solution, unless the Add Inj map overrides the OEM pulsewidth that way I won't get that HUGE spike right before the OEM cutoff when my AddInj value gets added to the stock duration. Basically, I should be able to set the HOLD/LIMIT RPM in the RLC map and then control the output duration from the HOLD to the LIMIT with the columns in the AddInj map.

So you're planning on waiting until you get your car out of storage to try their fix?

Yeah, the logging is the problem, since IDC stays consistent.

Also, it looks like Brandons' IS3 logs only went up to 6700ish not including his unloading the engine extra bump. Have you seen any other EU Rev-Limit logs besides Brandons that only went from 6400rpm to 6700ish?

DandyMax

iTrader: (7)

What I was thinking was that it shouldn't require setting the Inj map up so high like that if the RLC map was able to override and maintain control right at that hold point and further. And if things were synced up properly with the rpm pickup. It's pretty clear where your spike comes from looking at your Inj map columns. I guess we'll have to see how this plays out once we have the crank signal but even then it may still require some reworking on Greddy's part. It seems the fuel functions right now are still meant to piggyback the ECU for the most part not replace the signal, even though that would be better with our ECU type.


Haha not if I can help it. My plan is to go up the highway and bring my EU up to Jime's place and use his car (I'm not sure if he's got it installed yet and haven't even talked to him about this yet but I have a feeling it'll be fine). But I will try to get you a module as soon as possible too. If anyone deserves one you do for all the work you've done, not to mention being the 1st to get an EU after I did.


No. In fact, other than you and Brandon I don't know of anyone who's actually tried and got a working Rev Limit extension with the EU. I posted a few times trying to get the Z guys to try it but no one has that I'm aware of.

Do you know of any others?

IceY2K1

iTrader: (3)

As long as we can still use the AddInj map to add/subtract fuel from the HOLD-to-LIMIT, then we'll have the multiple fuel load points(columns) to work with and the RLC map will keep the base fuel constant, then it sounds fine.

If not, mail it my way and I'll give it a try. Just have to steal a laptop again.

Exactly, just wondering why everyone was so concerned about my 6600rpm limit, since even Brandon didn't go much past 6700(6733 IIRC) and some of my not so pretty logs go that far. None of Brandons' logs are pretty/smooth, since he didn't have it setup correctly and was running pretty scary in the beginning. My feeling is that it is CRUCIAL to get the injector output duration to stay flat from stock cutoff to 6600 the way the EU works now in different gears/under varied engine loads. Once I have that, all I have to do is change the RLC map 6600 LIMIT column to say 7000(new limiter) and set the AddInj map 6600 load column to 7000(new limiter), ie simple 2sec. change/no biggie.

Anyways, the EU Rev-Limiter does exactly what I want it to do MOST of the time. Just every now an then it drops the ball especially when the loading of the engine changes between runs even with the SAME maps. I believe with more tweak time, I can dial in fuel/timing to keep the RPM linearly increasing, which should prevent falling into the lower map cells and my only real remaining issue. A dyno with consistent loaded runs would have probably solved this quickly, however on the street things vary alot making it way more difficult to nail down exactly what is causing my issue.

If you look at Brandons' maps, his RPM line log is way more smooth and has even finer resolution, ie 0.02ms vs. my 0.06 or so, due to his crank input. So, that will definitely help the interpolation between the cells, especially large jumps like what we need at the OEM cutoff unless Greddy changes things.