Need a little help with Deatschwerks 1000cc injector tune

Attachment 47813Attachment 47814Attachment 47815

I recently installed a set of 1000cc Deatschwerks 1000cc 95lb injectors to replace the 60 lb injectors after they ended up maxing out at 116% duty cycle on my new build. The engine is a 415 ci, with about 9:1 compression, 2 bar speed density - otherwise it ran just fine with the 60s.

So with the new injectors, I received the injector data straight from Deatschwerks, Im assuming it's good. It will run fine and the fuel trims are able to learn enough to keep it around 14.7 afr. The problem is, when I just barely touch the throttle, it will make a sudden change in afr to around 10:1 and just spit raw fuel out the exhaust, let off the gas and it goes completely to the other extreme and leans out completely. Under part throttle cruising the afr looks good and its able to learn the LTFTs and maintain 14.7 average. I haven't gotten around to tuning the boosted region of the VE table yet, but it seems to be working fine in that area too.

So basically, I can tell it's making a transition when I just barely step on the gas and then makes another transition when I let off the gas, but in between during part throttle operation it's fine. At first I thought the throttle cracker and or follower might be playing a role but that would just make it hang a little but I haven't experimented with that yet so maybe Im wrong.

I posted 2 scans, one of them I forgot to connect the WB. The scan shows where it's trying to subtract 15% at 1600 rpm and 20kpa but only 3% at 25 kpa even though the two cells are set to 39 and 43 respectively. This issue is more evident at the lower rpms but happens at whatever rpm Im at when I transition on or off the gas. It's just able to recover under more load for some reason.

Does anybody have any ideas of something Im missing? I just need a little nudge in the right direction I think.

Thanks in advance.

Ok, well pretty sure it's bad injector data. I experimented with changing the PW and Default PW and the adder, seems like the adder is off since I have gotten it to improve quite a bit by playing with that.

Get IDs and make your life easy.

Yup, a little late for that now. Had I known about them prior, maybe I would have. I did look at their data and compared it to deatchwerks and I don't understand why the deatchwerks short pulse adder dives negative after starting at .100 while ID's data doesn't go negative at all.

It's all based on the injector behavior.

I agree with Dave and his assertion that you should start with known good injector data but unfortunately that isn't your case, so you need to try something.

Where did you find the 1000cc/95 lb GM characteristic injector data on their site? I couldn't find it, only the Ford, which I ran through Dave's, Ford to GM injector conversion tool.

Not sure if this will fix your problem but try changing the Transient Fuel Qualifications>Min Fuel milligrams to say .02 and see if it helps or at least makes things better. Also change default pulse width to .725 as well

The correct way to log transient operation is to drive the car at a steady speed for a bit and then increase the pedal quickly by say 20% and hold it there while collecting data for 5 seconds or so. Decelerate is basically the same....attain a constant speed than release the pedal some 20% and hold it there for 5 seconds or so. This will show you how well the injectors are responding to the transients.

See how that goes for you.

Ed M

I have changed the default pw to .725, of course for some reason it gets automatically rounded up a little when I enter it in hp tuners. I have also changed the min fuel to .023 after reading something a few days ago. Does that affect the initial accel only or does it apply to decel as well?

I did get this data emailed to me directly from deatschewerks but after looking at the data they have posted for their other injectors, it looks like they use mostly the same data for all of them except for changes to the pw adder. Im still waiting for their "tuner" to get back to work so I can see what he says. Right now it seems like a crap shoot and im just guessing, hoping I can just get it close enough that it will at least be able to learn the ltfts at idle and part throttle. I do appreciate the advice, maybe I will stop messing with the pw adder and work on the transient instead

Did those changes lower the idle pulse width from the 2.4ms in the above logs?

Ed M

Attachment 47886

So here is my last one, well the last one that I saved anyway. At this point, I was trying so many different combinations that Im not even sure which tune I had loaded at the time I saved this. Im gonna have to go back at it on Sunday and make sure I pair up my logs and tunes. It does look like the inj ms went down to around 2.0 ms on this though.

So should I work on the transients more and run off the data they gave me?

Heres the PW they gave me, I just think it's weird how it goes negative. But if at .0000 is where it idles, then it has a problem there and if .0608 is around 15% TPS, then it also has a problem there, but seems ok everywhere else. Just not sure how to log that to determine what cell it's in.

limit 0.0000 0.0608 0.1216 0.1824 0.2432 0.3040 0.3648 0.4256 0.4864 0.5472
2.8576 0.1000 0.0696 0.0392 0.0088 -0.0108 -0.0258 -0.0372 -0.0478 -0.0571 -0.0677

0.6080 0.6688 0.7296 0.7904 0.8512 0.9120 0.9728 1.0336 1.0944 1.1552 1.2160
-0.0780 -0.0836 -0.0872 -0.0886 -0.0884 -0.0861 -0.0769 -0.0693 -0.0631 -0.0587 -0.0554

1.2768 1.3376 1.3984 1.4592 1.5200 1.5808 1.6416 1.7024 1.7632 1.8240 1.8848 1.9456 2.0064 2.0672
-0.0543 -0.0519 -0.0487 -0.0451 -0.0406 -0.0340 -0.0305 -0.0285 -0.0283 -0.0282 -0.0281 -0.0231 -0.0171 -0.0163

2.1280 2.1888 2.2496 2.3104 2.3712 2.4320 2.4928 2.5536 2.6144 2.6752 2.7360 2.7968 2.8576
-0.0147 -0.0122 -0.0086 -0.0055 -0.0049 -0.0029 0.0004 0.0041 0.0073 0.0092 0.0100 0.0102 0.0121

So that looks like a positive. Should have also leaned out the idle, correct? I would lower again to say .018 and see if it is better or worse....a lot like going to the eye doctor :-) I guess at this point I would stick with the supplied data and try all the normal tuning techniques to see how good or bad it gets. See what the transient test shows as there may be some other transient adjustments which to me like black magic.

If I am about to state the obvious then I apologize....

To tune the LS1 PCM idle, there are several steps I take. It is important to remember there are two main idle modes....1) Standing still (Adaptive idle) and 2) moving (Rolling idle). within those 2 main elements there are sub-functions like cold start open loop, RAF airflow control (minimum air), adaptive idle air (slow correction -- based on idle LT/ST trims), spark overspeed/underspeed (fast correction -- based on desired idle speed error), in gear/PN, AC on/off etc.

Normally I raise the idle speed so it will idle on its own and tune the VE and MAF (understand your setup is SD) first. I move down in the rpm levels as low as I can to baseline the airflow models. This is important as it becomes the base for the idle and rolling idle control.

Adaptive idle (non-moving) is next. First order of business is make sure that either the IAC table (DBC) or the ETC scalar (DBW) are compatible with your throttle body. Tuning is done using RussK's .cfg idle technique. You search and find the procedure for this on this section. This will use the LT/STIT values to zero in the RAF (Base Running Airflow) table. Should include the Dynamic Airflow (VE) and calc airflow (MAF) pids in this .cfg as well. Will take several attempts at it to get it fine tuned but by this time all should be in pretty good shape, at rest as well as rolling.

The rolling idle also includes the Cracker/Follower tables, but in a vette of your year, they are disabled when you depress the clutch.

Hope this helps,

Ed M

Attachment 47950Attachment 47951

Ok, so I did a lot of experimenting with the pulse width and adder. No matter what I do or how much I take out of the pulse width, it will still go super rich at just off idle - so it idles fine at 9%tps but up to 12% tps it will pretty much make the 02s go flat and run between 10:1 and 13:1 afr. I did clean up the idle air flow a bit but that didn't help anything since it idles ok anyway. What did seem to make a difference was disabling the closed loop proportional field. Im not sure if that disables the closed loop proportional as well though.

I also set the min fuel milligrams to .015. Now it still decelerates pretty rich and that seems to be the main issue now. By looking at the log, the ve ltfts are calling for adding fuel when it is already pig rich. So if anyone has an idea how to get the decal to lean out I would appreciate it as well.

So you are throwing darts at a big wall. Make them give you the correct data or your money back. You can get your SPA to go negative, the factory LS3/7 goes negative in a couple of spots. I wish I had a big enough injector for you, we have our own and we have GM data for them. I am experimenting with a 82lb right now, just put them in a 02 Z with a A&A setup on a LS3. Right now we have a 57lb which would be too small.
The problem you are having is that the data is telling the PCM what the injector is, you really can't guess at this because you could be close in one spot and way off in another. Also are you even sure that injector is pulsing @.725ms. I can't seem to think of the last on I had on the bench that was actually flowing fuel right there. What is their SPA at .725 ms. When the engine is idling what are you at? I could run data on these for you but that is really something the manufacture should be doing properly.

are they ev14 style injectors? if so just use ID ev14 injector data, thats what i always do. Same bosch core, idc what anyone says tbh it is always spot on. I do a couple of changes on transient fueling but thats about it.

Quote:

---

Originally Posted by cesar

are they ev14 style injectors? if so just use ID ev14 injector data, thats what i always do. Same bosch core, idc what anyone says tbh it is always spot on. I do a couple of changes on transient fueling but thats about it.

---

That couldn't be more wrong and it has been shown by ID to be extremely wrong. I can build my own data and I can tell you that the data from a stock 42lb, to a 58lb, to a 82lb is very different and they are all different than a ID 850. The only common thing in all those injector is the ohms reading of the coil. Not only that but EV14 is the style plug that is on it and Bosch makes allot of different injectors with the EV14 or really the uscar connector. The core has nothing to do with, fuel flow has everything to do with it.

How long does it take you to get good injector data? I assume you've got some pretty good equipment to do that? Problems don't really become apparent until you start getting into bigger injectors. Are you planning on getting 100+ lb injector data for certain injectors or are you marketing these specifically as your own injectors?

Quote:

---

Originally Posted by PICNIC_GEORGE

How long does it take you to get good injector data? I assume you've got some pretty good equipment to do that? Problems don't really become apparent until you start getting into bigger injectors. Are you planning on getting 100+ lb injector data for certain injectors or are you marketing these specifically as your own injectors?

---

It takes us a day to run the injectors for the raw data. I did not plan on going bigger than 82 because that is ID's business model and honestly most of my business would be smaller injectors. I am starting with new LS3/7 injectors and we are machining them to the correct size. I explain a little bit of it on our website, but I am working on a video that will really put it into perspective and it shows the process. Well it shows the machining process, I don't show much about the data collection, but Howard from Redline has a video on his site that has him, Greg Banish and myself collecting the raw data. So far we have been testing them and several sets have been sent out, so far all results have been great. We flow every injector to match injectors from 2.5 ms up to 8. All injectors are matched to within 1%. It is allot of work and we are still in the early stages of getting injectors on the shelf. We have been working on this project now for about 3 years. My website is zip-corvette.com.
Thanks
Justin

Ok, so I am still waiting on their "tuner" to get back from wherever he is to talk about the data. But I did get part of the problem to go away. I went back to my original thoughts about the follower and cracker. I could still be wrong about that, but I did have those tables previously modified for use with the old injectors to help out with the bigger cam. So putting the cracker and follower back to stock settings and then problem where it went really rich at light acceleration - under 13% tps- went away. Now with the follower, it still goes rich on the decel but it has the noticeable cruise control effect. Im thinking that maybe the bigger injectors just skewed the cracker and follower beyond its ability to maintain a 14.7 afr. If this is correct, and Im hoping it is, then the only thing I will have to fix is the hunting idle coming down from a decel, other than that it idles fine once it settles.