[QUOTE=Sackett]Someone correct me if i'm wrong, but AN2 should be the TPS sensor, which the XEDE doesent use, which is the Purple wire in one of the harnesses. Look for the XEDE firmware update instructions, the purple wire that they tell you to cut is the unused TPS input. use that one and see if it works.QUOTE]

I think your right about pin 11 but the purple wire goes back into pin 10 on the grey connector after doing the firmware update and is then connected to the LC-1 AFR sensor(this is AN1). Pin 9 is reserved for the SMART knock buffer and is internally mapped as AN0. I don't know for sure where AN2 is but if it is the TPS sensor then it might be pin 11 (according to Xmap's pin out diagram for Subarus). This would make sense because it is currently mapped as analog MAF in and Vishnu uses frequency MAF in so it should be available, but I'd hate to blow anything up. Can anyone (Vishnu please) confirm pin 11 is an analog input?

I'm going to play with the datalogging this weekend and also try to get the closed loop boost control working so I will post some graphs hopefully this Sunday and a wiring diagram for hooking up the GM sensor.

Has anyone used Xede's built in closed loop boost control? I couldn't find any posts on it. Specifically, what values to use for "gain" and "authority".

 

Linky to a layered XEDE wiring diagram ... click for XEDE wiring diagram

l8r)

 

If the Omori's are anything like the Innovate gauges then they are daisy chained together on an RS232 type of communications bus and not a 0-5 volt signal. In which case you wouldn't get any meaningfull data from the gauge. I could be wrong though. I just T'd in my GM sensor ($54) behind my boost gauge and ran the output into Pin 9 on the Xede. You'll need something like this http://www.dimensionengineering.com/DE-SWADJ.htm to step 12volts down to 5volts for the GM sensor to work (or tap into the 5volt supply to your TPS).

 

Dude, you have WAY too much free time on your hands.

Did your car get sorted out?

d

 

any of you innovate LM-1 guys know if you could run a map sensor with 5v output into the RPM pickup on the LM-1 as one of the inputs?

 

Sure you can. I had the LM-1 for a while, and you can log whatever 0-5v signal you want with it. I'm assuming you're using the RPM converter? If so, you can send the signal from the MAP to any one of the remaining 4 channels (RPM and WB are the first two) and configure logworks for your sensor. There's some good examples on the innovate website that should help.

EDIT: Grammar.

 

yah...i just did some reading...

i wish i would have gotten the auxbox instead of the RPM converter now. I has a built in 3bar map sensor, and temp sensor for either cylinderhead temp, exhaust temp, intake temp...whatever. wasn't much more expensive either.

I understand the knock sensor isn't a 0-5v, but a frequency. Im not really sure what that means, but is there a way to change it to a 0-5v? is that what the knock buffer on the xede does?

 

I don't know exactly how you'd convert the knock sensor into a 0-5v signal, but I'm going to guess it would work something like the RPM converter does. I also don't know what the input frequency range would be either, but I'm going to guess that you are correct in that the knock buffer on the SMART upgrade does exactly what you're looking for.

I actually had the LM-1 and Auxbox. My only complaint with the latter had to do with trying to use an external MAP sensor; the built in one worked perfectly. In the end, I decided to sell them both and get the Xede w/ SMART, mainly because I could actually change things if I didn't like the log results.

 

It's not that I have too much time ... it's just that I'm really **** when it comes to stuff and I'm a slow learner. So to help me understand what was needed and have something legible, I made the diagram.

Car was doing alright (SMART works, 9100rpm redline achieved to Shiv's dismay) up until the clutch decided to take a permanent vacation. So irritating to have a $5 part f-up and end up costing me $$$. New clutch going in next week.

l8r)

 

My understanding is that a knock sensor is nothing more than a microphone stuck on the engine block and therefore only outputs very small voltages (millivolts). I think the SMART buffer takes this signal, amplifies it, smoothes it out a bit and then sends it to the Xede as a 0-5v signal. Just guessing though.

 

well, i need one if it infact works like that.

juyanith, i run the xede in com1 and the LM-1 in com5 at the same time, lol. Works pretty good, log a run on logworks then make changes accordingly on the xede.

now, i guess i just need a 3bar map sensor that sends 0-5v signals so i can log boost.

 

also,

anyone know if the knock sensor sensativity is still adjustable in v3? i know there isn't a table for it like there was in v2.

thanks guys

 

It isn't clear that the closed loop boost control was intended to solve the temperature compensation issue. For reference note that the default screen uses the knock buffer as the "load variable". I'm guessing that someone want to manage boost as a function of knock.

 

I had Omori gauges but I got rid of them. That said the electronic boost gauge is similar to the Defi gauges in that a map sensor is employed with power, ground and an output.

I know for a fact that Defi sensor runs on +5v and its output is in the range of 0 to 5 volts. On the other hand I didn't test the Omori sensor, but it would very simple to probe the I/O's to determine how it functions. I'm guessing it is running off of 5 volts as well, with a 0 to 5 volt range.

I tried playing with the Xede's closed loop boost control with poor results. When I used the MAF as the load variable and the boost control as the output, with increased load I found that the boost duty cycle increased. That isn't what I expected. In fact it is opposite as to what is needed.

Living in a region where the temperature swings are large I have though about implementing closed loop boost control via a 3 bar MAP sensor for some time. I congratulate you for moving this forward.

I'm curious as to your strategy for creating the closed loop map.

 

If someone could figure out how to intregrate a 3.5 bar sensor into the Xede and then use it for closed loop boost control then that would just be pimp. Setting up boost tables through wastegate cycle percentages is just a PIA. Hell, if the Xede would read and log in A/F ratio's instead of a percentage that has to be iterprolated then that would be good!