Yeah it is looking good. I don't like the **** there but that can easily be worked around.

 

After I saw how much fuel economy dropped with the turbo installed I got creative with my waste-gate plumbing. Need another solenoid and a couple check valves, but it will be able to open the WG at idle/cruise and then force it closed all the way up.

 

Typically fuel economy improves slightly with addition of a turbo ...... so I'm thinking you have an issue .

 

Pfft, not in my case,

 

Well yes I have several issues most are because it's barely tuned and it goes pig rich in boost. I didn't change anything in the cruise range and I've lost a couple mpg. I'm assuming it's because of back-pressure from the turbo, so I've failed the wastegate open to find out.

I thought most people lost fuel economy when adding a turbo, especially highway mpg. Am I mistaken?

 

The only way you will lose mpg at cruise speeds is if something is wrong . Normally you would expect to get ever so slightly better MPGs in all ranges of acceleration bar those in boost .

 

Yeah I get equal highway fuel mileage after going FI. Definitely use more fuel in the city though, that's just a given...

 

Yeah it is ...but only because it's almost impossible for anyone that has a turbo to not use boost.
If you put an NA car and a turboed car (with identical engines) on a test course and accelerated them at exactly the same rate I'm pretty sure the turboed car would use fractionally less fuel.

 

Tracking down my issues now. I think I have a coil pack going out, and I also have a water leak at the injector bung. If I can get a few days off work without being sick I might even get some stuff done.

I'm playing with the TI knock sensing chip. Communicating with it with no problems, but i'm not having any luck getting the A/D to work with an external crystal. I think I may have burnt out something when I soldered it, it's my first time playing with a surface mount IC so it's not unexpected. Should have a second chip in a few days. Also I think I found the fault in my knock sensing code so something should work out of all this.

Have everything for the intake, just don't have the time and energy to redo it right now. I'll post more when I have a real update.

 

Found out about and bought a MS3 Knock module. It's the same TI chip with all the auxiliary components including the crystal and caps already installed. A bit of a rip off considering the cost of the parts and the lack of any added value from the design, but it beats having a custom board etched and then trying to soldier all the surface mount components.

 

I've redone the intake , this should be close to my final version. No steel in any of the intake/wi plumbing now. It has one 90 deg coupler which needs to be replaced, the rest looks good. Need to plumb in the BOV recirc but that's for another day.

Meanwhile I've redone the electronics and am currently installing the dash mounted knock/WI/boost control. It's a bit of a mess right now but it is also approaching the final version. I think I'll redo it again for fitment probably have a circuit board etched so it won't be a spaghetti system of jumpers.

The TI knock chip works. It has a bit wider frequency band compared to the code I'd written, but I don't think it will really matter. The reason why I'm sticking with it is that it doesn't need anything else to work, it can take the signal directly and report the amplitude of the target frequency. No tricky coding to make the math work quick enough, no dead spots when the chip runs other code, and no other hardware needed. I'll post more on the electronics when I get it all installed.

 

Here is the display installed. It's working and reads out IAT, MAF, as well as pre and post TB MAP as well as knock and current water injection rate. I haven't had time to put it through its paces yet, at this point I'm just glad none of the wires are crossed and the indications are close to reality. MAF is reading 3-4 at idle, whereas actual from obd2 is reading 4-5. I do have a vacuum leak somewhere, but the display should be closer to the obd2 value... odd.

The code does a linear interpolation from a table in case anyone is wondering. So the values displayed are from the actual sensor curves for MAF and IAT. MAP is calculated from sensor voltage since the response is linear.

Under the hood wiring is a mess, I need to start working on an actual harness since everything is firming up. Also my small plumbing needs to be redone since it's a mess and I'm happy with how everything is right now. The project progresses.

Right now the 55lb ev12 injector is installed preturbo. It should support 25% Water/fuel to about 500g/sec air. Also I have a flow meter on order so I can add another level of failsafe.

 

Lot's of wires. I would surely mess that up, lol.

 

Mines a bit better than that for now. How much did that knock chip cost?

 

The chip is about $7, make it an even $15 with resistors caps and a development board. The other option is a ms3 knock board for around $100. I figured out my initial problem after I bought the ms3 board. Let me know if you want a parts list or other help the ti documentation is very light on application.

 

Figured out a few of my problems. I pulled the main pulley and trigger wheel when I installed the turbo manifold and then I never reset the KAM. It all snugged up and nothing was changed, but after resetting the KAM no more misfires. Still crossing my fingers on that one because it's only been a couple days. After I'm sure that's cleared up I'm connecting the RPM signal and knock lockout to the e-shaft sensor and we'll see if that causes any problems. Then everything will be together and time to start putting the pressure on.

 

good news, I am interested in the knock chip. Where did you get it. A Parts list would be great. I can figure the electronics out but it takes me time. Do you have any pics of the board. I'd feel okay with soldering up a chip like a arduino uno, but a chip like on the mega is a bit too small for me.

 

Its an SOIC-20 chip, so you need a breakout board and some patience to solider it. Here is the intersil datasheet for the chip that TI copied it includes how to wire an external crystal.
http://www.intersil.com/content/dam/...p9/hip9011.pdf
And here is the TI datasheet.
http://www.ti.com/lit/ds/symlink/tpic8101.pdf

Attached is a basic version of the code I used for interfacing it on a arduino mega. For an uno you need to change the SPI pins. Pin 49 is the CS pin, and pin 4 tells the chip when to detect the knock.

In the morning I'll post some pictures of the electronics so you can get an idea for yourself. If I end up redoing this again I'll have a custom board etched and eliminate a bunch of the mess. Right now it's just a messy prototype.

 

Those will take me some time to process, If you do go with a custom board let me know. I might be intrested in one also.

Were their any other knock processors you found? Why did you choose this one?

 

I've still got the breakout board with the first chip. PM me if you want it. I chose this one because I found other people using it with some (very little) documentation. If you can find another chip worth trying, let me know.

 

Trying to decide what to use as a water injection tank. Currently have a gallon antifreeze jug, but I don't like it for long term. I'm looking in the 4-6gal range, possibly a fuel cell. Devils own has nice tanks, but nothing that fits quite right in the trunk. I'm trying to find something that will fit up against the backseats and keep the weight central without loosing much trunk space. Anybody have ideas?

 

Figured out a way to test the whole water injection system without adding any hardware. The pump has a pressure switch that cycles it on and off, and under low demand its cycles are short and far between while under high demand it has long cycles close together. This gives me pump duty cycle.

So I can have a pump duty cycle check both at start up and when injecting. At start up it sprays for a brief moment then checks how long the pump takes to re-pressurize. If the duration is too short the nozzle is clogged. When at power it checks for duty cycle being to short (again nozzle clogged) and too long (loss of suction, or leak). If either is detected it cuts boost to the safe non WI level. Between that, IAT boost cut, and knock boost cut it should be reasonably safe.

 

I just fitted an alarm if water doesn't reach a certain pressure ....... ever heard of the KISS principle ?

 

Yeah, but if the fuel injector plugs (kinda likely) then pressure never drops. Compared to what other solutions I've thought about this is simple.

Remember this isn't just for safety margin, if it fails (and nothing else catches it) the engine will fail soon after.

Just flow tested my injector at 800ml/min. May need to go bigger for my goals, but it's good enough for now.

 

Only if :
*you use injectors not designed for the purpose they were intended for
*don't use distilled water
*don't have any filtration

You really should be using the right tool for the job rather than complicating it as a compensation for using the wrong thing .

This would suit your needs perfectly ...actually designed for the job .......... solenoid nozzles
http://www.alamomotorsports.com/FJO/waterinjection.html