Engine Monitoring Laptop idea -> have some questions.
#1
Engine Monitoring Laptop idea -> have some questions.
I am going to start working on a plug and play system for my car and I need some help. I haven't looked at the electronic diagrams yet, but that isn't a big deal, I can look at those afterwards.
#1 Is there a spot for a "serial" type connection or something similar that could be derived to a serial connection?
#2 Is there anywhere on the entire internet a mapping for what signals are what, or would I have to directly tap into the ECU's signals (such as speed, RPM's, etc.)
I know I could just buy a Gtech or something, but I want a plug in system that can produce graphs and spreadsheets of power and acceleration over time...in short, create my own power curves.
All I am going to do is write a small program to monitor the signals and then throw them into an Excel spreadsheet to create graphs, but I also want to design it to monitor the engine, EGT's, O2, Airflow/Fuel pressure/oil pressure/speed/rpm etc.
I am sure that there is a spot for mazda people to plug in something to read those signals, If I had the voltages I could rig up something to connect to a serial port on a laptop...hmmm.
If there is already a product out there that can do this then let me know so I can purchase it =)
-Thanks
(That black civic in the bottom of my signature photo didn't stand a chance muahahaha)
#1 Is there a spot for a "serial" type connection or something similar that could be derived to a serial connection?
#2 Is there anywhere on the entire internet a mapping for what signals are what, or would I have to directly tap into the ECU's signals (such as speed, RPM's, etc.)
I know I could just buy a Gtech or something, but I want a plug in system that can produce graphs and spreadsheets of power and acceleration over time...in short, create my own power curves.
All I am going to do is write a small program to monitor the signals and then throw them into an Excel spreadsheet to create graphs, but I also want to design it to monitor the engine, EGT's, O2, Airflow/Fuel pressure/oil pressure/speed/rpm etc.
I am sure that there is a spot for mazda people to plug in something to read those signals, If I had the voltages I could rig up something to connect to a serial port on a laptop...hmmm.
If there is already a product out there that can do this then let me know so I can purchase it =)
-Thanks
(That black civic in the bottom of my signature photo didn't stand a chance muahahaha)
#4
Although I hate to discourage you from an interesting sounding project, you’re talking about trying to monitor a 15yo computer that was basic in its day let alone now. There is no way to plug straight into the ECU and monitor everything in real time, as there was no reason for the factory to provide such a feature. The diagnostic plug in the engine bay is about as fancy as it gets. When Mazda want to monitor all of the ECU’s inputs/outputs directly, they unplug the ECU and plug what is basically a big, fancy voltmeter in between the loom and the ECU. They can then manually select and observe each device’s voltage one at a time. There is also absolutely no way to get acceleration or power data from the ECU, as it has no way to measure or calculate this sort of info. Things like oil pressure and road speed are also not monitored by the ECU.
Like Kahren said, a Haltech, MoTec or similar is quite capable of doing some of what you want, but it is completely beyond the capabilities of the stock ECU. Sorry for the downer.
BTW, there is certainly no parallel port under the rear seat...
Like Kahren said, a Haltech, MoTec or similar is quite capable of doing some of what you want, but it is completely beyond the capabilities of the stock ECU. Sorry for the downer.
BTW, there is certainly no parallel port under the rear seat...
#5
hmmm ok so...
ok soo...
If I were to take voltage readings from all the sensors, pass them through a circuit board that would relay the voltages back in a binary number which could then be interfaced to send a signal through a serial port into a laptop which would be read by my custom program and then displayed on the screen.
Given I would have to calibrate it based on what gauges and such say, but the program could do the math based on all the signals (speed etc.)
Probably would need some extra sensors installed...but big deal really.
DOH maybe some day I will just build a custom ECU and sell it LOL, full display with a touch screen interface...mmmmm, I should wire somthing like that up.
If I were to take voltage readings from all the sensors, pass them through a circuit board that would relay the voltages back in a binary number which could then be interfaced to send a signal through a serial port into a laptop which would be read by my custom program and then displayed on the screen.
Given I would have to calibrate it based on what gauges and such say, but the program could do the math based on all the signals (speed etc.)
Probably would need some extra sensors installed...but big deal really.
DOH maybe some day I will just build a custom ECU and sell it LOL, full display with a touch screen interface...mmmmm, I should wire somthing like that up.
Last edited by PraxRX7; 04-16-02 at 05:02 PM.
#6
Yup, I can't see that being worth the time or effort. You could wire in the haltech 10 times before ever getting that system up and going. Interesting idea I guess. Sounds a little like the greddy Emanage, in respect to using stock sensors to gather information.
Trending Topics
#8
Sounds like a pocketlogger, or any data logger out there.
There are alot of systems like this out there for other cars. I would assume there isnt one for the RX-7 becuase its not economically or physically possiable to build such a system.
There are alot of systems like this out there for other cars. I would assume there isnt one for the RX-7 becuase its not economically or physically possiable to build such a system.
#9
Au contrare...
I believe Jameco offers datalog devices between 4 to 8 channels.  Now, you're not going to be able to monitor EVERY ECU channel short of a multi-million dollar equipment (last time I checked, the ECU had over 50 leads), but you should be able to monitor the "important" ones...
These datalogging packages are a few hundred dollars, but it's a generic kit that needs to be tailored to the FC wiring.  Good thing is you're normally monitoring voltage signals, so you don't need to cut&splice any wiring - just tap into them.
-Ted
I believe Jameco offers datalog devices between 4 to 8 channels.  Now, you're not going to be able to monitor EVERY ECU channel short of a multi-million dollar equipment (last time I checked, the ECU had over 50 leads), but you should be able to monitor the "important" ones...
These datalogging packages are a few hundred dollars, but it's a generic kit that needs to be tailored to the FC wiring.  Good thing is you're normally monitoring voltage signals, so you don't need to cut&splice any wiring - just tap into them.
-Ted
#10
I've been thinking about doing this for a while now.. I was planning on using the parallel port, tho.. A few thoughts:
You can get a simple A->D converter chip pretty cheap... But that takes up 8 data lines on the parallel port (there are only 8 REAL data lines, but you can sorta cheat with other signal lines to get a few extras).. However, it DOES allow for like 6 different analog inputs, which you could select via output lines on the parallel port (more cheating.. ).. This would allow polling of o2 readings, tps levels, maybe oil pressure, coolant temp, etc... Just be aware that you only get to set the reference +/- one time, and all the inputs must use it.. so if your TPS is varying between 1-5 volts, and your o2 sensor is somewhere between 0-1, then you'll need at least 0-5 to cover both... and that makes you lose accuracy for both.
In order to calculate power, you need to know acceleration and weight (and a few other things to get it even more accurate.) In order to get acceleration, you can get the rate of change in RPM (assuming you know the gear you're in).
Assuming a max rpm of 8000, that turns out to be ~133 pulses/second from the coil...so if you can hook it up to a parallel port input line (watch out for too much current...and i think the coil tach line is 12V...), and poll it at at least 266hz, then you should be able to determine roughly your RPM. The fasterl you can poll it, the more accurate your results. Ideally, have it trigger an interrupt that logs the current time minus last interrupt time... that gives you the inverse of your RPM.
Now....calculate the CHANGE in RPM over time, and you've found something proportional to your acceleration... If you want to do it "right" and have it accurate much above like 50mph, you'll need to know your speed, and factor in wind resistance. Do something with weight, and you can plot yourself a lil dyno chart. A lil tuning will make it more accurate, but it would DEFINITELY work for comparison purposes (before-after mods)
I have a 486 dx 50 laptop thing that I plan on using for it.. Hopefully I'll have some time over the summer to actually work on it..
-Tesla
You can get a simple A->D converter chip pretty cheap... But that takes up 8 data lines on the parallel port (there are only 8 REAL data lines, but you can sorta cheat with other signal lines to get a few extras).. However, it DOES allow for like 6 different analog inputs, which you could select via output lines on the parallel port (more cheating.. ).. This would allow polling of o2 readings, tps levels, maybe oil pressure, coolant temp, etc... Just be aware that you only get to set the reference +/- one time, and all the inputs must use it.. so if your TPS is varying between 1-5 volts, and your o2 sensor is somewhere between 0-1, then you'll need at least 0-5 to cover both... and that makes you lose accuracy for both.
In order to calculate power, you need to know acceleration and weight (and a few other things to get it even more accurate.) In order to get acceleration, you can get the rate of change in RPM (assuming you know the gear you're in).
Assuming a max rpm of 8000, that turns out to be ~133 pulses/second from the coil...so if you can hook it up to a parallel port input line (watch out for too much current...and i think the coil tach line is 12V...), and poll it at at least 266hz, then you should be able to determine roughly your RPM. The fasterl you can poll it, the more accurate your results. Ideally, have it trigger an interrupt that logs the current time minus last interrupt time... that gives you the inverse of your RPM.
Now....calculate the CHANGE in RPM over time, and you've found something proportional to your acceleration... If you want to do it "right" and have it accurate much above like 50mph, you'll need to know your speed, and factor in wind resistance. Do something with weight, and you can plot yourself a lil dyno chart. A lil tuning will make it more accurate, but it would DEFINITELY work for comparison purposes (before-after mods)
I have a 486 dx 50 laptop thing that I plan on using for it.. Hopefully I'll have some time over the summer to actually work on it..
-Tesla
Thread
Thread Starter
Forum
Replies
Last Post
Claudio RX-7
Haltech Forum
5
04-23-19 02:50 PM
82streetracer
Haltech Forum
11
03-11-19 05:34 PM
befarrer
Microtech
3
08-22-15 05:52 PM