binder

iTrader: (8)

hmm, never thought of that. So set my fuel corrections to say 35%.

how would you do this for the open loop in boost? are you saying to run closed loop even in boost? how effective and safe is that?

MR RIZK

iTrader: (7)

Thanks rcdash, it does help a lot as it pretty much answers my confusion. The solutions highlighted (potentially excluding point 2) shows that fail safes are possible however it requires “external” wiring and or changes. I was hoping to see that it was possible to implement a pure software based fail safe where by the trigger and enforcement of the trigger required no wiring.

MR RIZK

iTrader: (7)

Quiet before the proefi owners pike up and disagree with you. Window scaling whats that although it is a nice bit of kit. Only if the vi-pec had a good Z customer base and harness I would switch. The software is pretty nice in the vi-pec with comprehensive documentation. Download it and have a play.

str8dum1

iTrader: (11)

i run my entire map in closed loop. you still setup your map open loop, then turn O2 correction on once its dialed in. You can still setup the PID O2 controller as well to make sure you arent adjusting too fast/slow/too much/ too little. Thats how motecs and proefis are tuned.

The proefi interface is actually not bad once you get a feel for it, and have a good sized monitor. Its not meant really for carpcs, or anything under 16". its a universal box with lots of control, so of course there is lots to see. Haltech is way better of course, but once set, you dont really touch 90% of the settings anymore anyhow.

binder

iTrader: (8)

I'll have to check out the vi-pec system just to be nosy but I doubt i'll be switching to it due to the lack of tuners. I trust injected so i'll probably always use a platform they will tune on.

I also agree with you that it would be nice to have a purely software based fail safe. They have the knock control but it's tough to setup the knock maps perfectly. If i could set a fuel cut based on a/f just with software or even a boost cut based on a/f that would be awesome. Mind you, i mean set it without having to wire up stuff or add another piece of hardware to the system.

ya, i figured most of it was my lack of knowledge of it. I'm pumped about this closed loop thing. I might send you an email on setting some PID settings to optimize my o2 control. I think mine might be slightly off already based on my logs on vac with closed loop. It seems to overshoot the correction a little.

sorry for the derail.

rcdash

iTrader: (18)

binder, I have my knock fail safe (to cut boost back to 10 psi) set up based on engine noise, not true knock, but I don't get many falses.

I tuned the knock table by making sequential data logs. Decide on what your knock level threshold is going to be - I use 150. It doesn't really matter because you are going to change the sensitivity grid to fall below that level under normal driving conditions.
1. Set the entire knock sensitivity grid to one value (like 0).
2. Do a run through the gears at low boost or no boost but through the rpm range and include slow deceleration at the end of a run.
3. Increment entire threshold grid by 10
4. Repeat step 2 and keep repeating until you get no detected knocks throughout the run
5. Now examine the logs and see where for each grid point, the knock level values are below 100 (remember I won't detect knock unless a value exceeds 150).
6. Adjust the knock sensitivity grid points to the value that just prevents knock levels above 100.

This seemed to work pretty well. I did get one or two "false" knocks at the airstrip event that cut my boost on some runs towards the end of the day - but I was running of out meth, thanks to a leaky tank, and the car was really hot, so maybe it wasn't really false - I dunno. When I looked at the logs, it was just barely over my threshold value, so I doubt it was true knock (I would suspect a much larger spike).

binder

iTrader: (8)

wow, thanks for that info raj!

i've been spending hours of time adjusting the threshold map so it doesn't even show any knock spikes on it right now. I put race gas in it with my pump gas tune to make sure it was false knock then i've been adjusting the numbers to read zero.

with me setting them down to read zero do you think that is a bad thing? should i set the knock tuning up from that or start back with an oem threshhold map?

rcdash

iTrader: (18)

You can try to aim for zero, but then your threshold for true knock detection should be set pretty low. I like to see that the knock sensor is reading something and be able to tell how close to threshold daily driving takes me (rather than just being a flat line), so I set it up to allow that.

binder

iTrader: (8)

i'm going to email you to get some further clarification because i read it too fast the first time.

BlinkerFluid

iTrader: (7)

One thing you guys can use for a loss of fuel pressure in boost failsafe is a differential pressure sensor.

Use an analog input on the I/O box, and set it to drop boost if the fuel pressure varies more than -+5 from the differential pressure setting (base fuel pressure).

The difference is a differential pressure sensor will reference manifold pressure and fuel pressure, the reading should be constant if your fuel pumps and regulator are working properly. If your differential stays constant, add that to your boost level and you can also monitor fuel pressure with the same sensor/input. That's my plan.

Hell, you could probably use it to activate a second pump if your comfortable with fuel pressure dropping off a bit to activate it. Add an accumulator for some pressurized system volume to totally overcomplicate something simple. Just brainstorming, you could use the differential pressure sensor and a FueLab fuel pump (reacts to PWM for speed control) and use the Haltech to drive the fuel pump speed based on fuel demand. That would probably be the most innovative setup. Too bad I've had very bad luck with FueLab pumps in my red car.

03/04 Cobras used a setup like this to run their fuel pump controller, I'm sure other vehicles do also.

Jeff, I'm curious, do you know a dude near you with a white EVO, runs 9's. Hangs out with Stoner and goes by the name of Fathouse?

binder

iTrader: (8)

yep, i know fathouse. He lives over in indy. I'm not "friends" with him per se but he's close friends with a lot of my car friends.

str8dum1

iTrader: (11)

i can tell you, a properly functioning fuel pressure compensation is pretty cool. My gas gauge wasnt working right and I ran outta gas on friday. I was at 20psi fuel pressure for a couple miles but was still hitting my AFR targets until the car finally died when I pulled it off the road in a safe location.

hopefully the gas gauge is fixed now

jerryd87

iTrader: (4)

Anyone have any idea how to set up the meth/water injection control a little better? Was looking at it and haltech told me with the i/o box we would be able to set it up to deliver a % meth based on boost, like 25% max flow at 15 lbs of boost. Messing around with it, it doesn't look like i would be able to do anything outside of turn it on with the i/o box, granted i wont claim to be an expert with it.

Cux350z

iTrader: (162)

probably would need an analog output

str8dum1

iTrader: (11)

correctly, it would use a PWM output like a fuel pump or a standalone meth controller.

binder

iTrader: (8)

so doesn't it have a pulsed output to do duty cycle? i thought they said that about the i/o box as well.

If so you just set up a map with axis for boost verses rpm and the cells should be a target percentage for meth injection.

rcdash

iTrader: (18)

Yes it definitely has 2 PWM outputs. It also has a specific output called "water injection" that is triggered at a certain boost, rpm, and temperature but I haven't played with it. Anyway, that is just a on/off output. The 2 digital pulsed outputs exist as "generic duty 1" and "generic duty 2" for which 2 independent 16 x 16 maps are provided. You can change the Hz to any value between 0 and 16 Khz (most solenoids are 30 hz). That's pretty cool. I have an aquamist HFS6 so...

jerryd87

iTrader: (4)

awsome thanks rc, ive been holding off on buying until i take a look at everything i dont wanna spend a *** ton of extra money if i dont have to if i can set up a map for meth that would be perfect that way i dont have a bunch of boxs everywhere =P

str8dum1

iTrader: (11)

i used the normal generic output on my haltech for meth for a while. worked perfectly. I even staged my nozzles using the output and a hobbs switch. you really dont need a PWM controlled meth pump setup. I certainly wouldnt buy a 500$ box just for that. The ability to trim to different boost levels on the fly is nice though.

jerryd87

iTrader: (4)

the only thing im worried about with it is that im going to be running it as a fuel, not really a detonation suppressant as something to lower iat's. So thats why i kinda want something precise or as precise as i can get it. its also direct port and i was figuring probably 10% of my fuel being meth should work well for what i plan on using it for. i also plan on other things as well such as the aux the second fuel pump turning on about 15 psi although that might need to be set higher wont know till i get everything put togeather. trying to get the car shipped at the moment and they are giving me problems because im still making payments, yet they gave me no problem with my wifes focus when we owed about 2.5 times on it what i owe on the 350 i might have to wait until jan/feb to ship it when i pay it off.

and yes i know running pure meth requires a firewall ive got it covered =D meth tank, fuel swirl pot, and battery are all doing inside a sealed stainless box in the trunk