What to use for DataLogging
Jun 1, 2006, 05:40 PM
#16
Evolving Member
iTrader: (2)
Join Date: Sep 2004
Location: South Bay
Posts: 287
Likes: 0
Received 0 Likes
on
0 Posts
Once someone figures out the MUT protocol, if should be able to trigger a raw data stream from the sensors that will output at a much faster rate and give us useful data
Jun 1, 2006, 07:40 PM
#17
Originally Posted by PDXEvo
I had problems with the tuner pro, so Im not a big fan of the unit. Specifically, it was difficult to keep the O2 sensor calibrated more then a couple of days, regardless of the O2 sensor. The box, however, was one of the originals (Purchased 3 years ago) and I am certain that they would have fixed the poor electronics used in rev 1 by now.
I was also looking at a similar device from Zeitronix. Anyone use the Zeitronix logger? Or the Innovate logger?
I was also looking at a similar device from Zeitronix. Anyone use the Zeitronix logger? Or the Innovate logger?
With that said, the creators of ECUFlash said their gonna incoperate Inovate LM-1 or LC-1, so that is also gonna be a great choice, when it happens.
Either get Zeitonix or Inovate.
Jun 1, 2006, 07:46 PM
#18
Originally Posted by trinydex
the problem is devices like the zeitronix don't log the timing, but they log everything else useful. if soemthing does then log timing then it does so off the odbii stuff... which is desperately slow.
Use a vishnu Knock Buffer hooked up to the stock knock senosr and the user input on the zeitronix, and presto, you logging knock, the same way Vishnu S.M.A.R.T. system does.
Jun 2, 2006, 06:21 AM
#21
Newbie
iTrader: (1)
Join Date: Dec 2005
Location: Orlando, FL
Posts: 94
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by Evo_Kid
Yes. We had this talk last week (even though it started as a WB thread), Zeitronix is the best, can log every thing @ 74 samples per second per peramerter. Inovate is 12 samples a sec and a OBDII datlogger is like 6 samples per sec, ECU+ only 25 samples per sec.
I am curious about how everyone is coming up with their sample rates. I spoke to someone at Innovate and they said that the data is collected at roughly 1000 s/sec but displayed in logworks at approximately 12 s/sec. The raw data can be viewed if exported to an XML or spreadsheet. I can't find the info right now, but I believe Zeitronix may be collected at the same rate but is displayed at the much higher 74 s/sec.
Originally Posted by trinydex
the problem is devices like the zeitronix don't log the timing, but they log everything else useful. if soemthing does then log timing then it does so off the odbii stuff... which is desperately slow.
[Edit: See Section 8 in the LMA-3 manual.]
With either of those two systems, you should be able to log all the important info like RPM, TPS, AFR, Timing, and Boost. The PocketLogger setup has the advantage of being able to view all the other OBD info and to pull and clear codes but (I think) the disadvantage of a very slow sample rate relative to the standalone Zeitronix software or the Innovate software.
For me, the ideal setup would be something incorporated in ECU Flash that combines the logging and CEL features of the PocketLogger software with a higher sample rate. That way I can keep my current wideband setup (which I prefer for its display and sensor) and not have to screw with having multiple cables and boxes in the car.
Last edited by ScrappyJack; Jun 2, 2006 at 06:25 AM.
Jun 2, 2006, 08:05 AM
#22
The Vishnu knock buffer isn't exactly what I'm looking for. It will show you all the noise the engine sees but it doesn't show you what Mitsu knows the noise floor should look like at different RPM. So sure you might be able to detect a large knock event but that doesn't really tell you how big of a timing hit you took or how many degrees of knock that event was.
I would like to hear how you feed that knock sensor data into the Zeitronix though. More info please.
I would like to hear how you feed that knock sensor data into the Zeitronix though. More info please.
Jun 2, 2006, 02:47 PM
#24
Originally Posted by Jeff_Jeske
The Vishnu knock buffer isn't exactly what I'm looking for. It will show you all the noise the engine sees but it doesn't show you what Mitsu knows the noise floor should look like at different RPM. So sure you might be able to detect a large knock event but that doesn't really tell you how big of a timing hit you took or how many degrees of knock that event was.
I would like to hear how you feed that knock sensor data into the Zeitronix though. More info please.
I would like to hear how you feed that knock sensor data into the Zeitronix though. More info please.
I dont see why people would buy an ECU+ for just datalogging. Its a piggyback and buying it just for datalogging seems a little, umm, weird.
The Vishnu knock buffer works just like the knock circuit in the ECU+, but the logging is faster, so you see more.
Here is how to hook up the Vishnu knock buffer to the Zeitronix (or any other dataloggin system with a 0-5V user input.
Pink - to the knock sensor wire (use a ECU pinout to find out what one is the knock sensor- road/race has the VIII pinout on the website)
yellow - to 0-5V user input on zeitronix or other logging system
black - ground
red - 12V power
Blue and white wires are for a built-in O2 simulator, I didnt use it so I dont know how to hook it up
Jun 2, 2006, 02:51 PM
#25
Originally Posted by ScrappyJack
I linked to that thread in my post above. I also mentioned that I regularly see 18 s/sec with the PocketLogger. As I mentioned elsewhere, you have to play with the Byte Timing.
I am curious about how everyone is coming up with their sample rates. I spoke to someone at Innovate and they said that the data is collected at roughly 1000 s/sec but displayed in logworks at approximately 12 s/sec. The raw data can be viewed if exported to an XML or spreadsheet. I can't find the info right now, but I believe Zeitronix may be collected at the same rate but is displayed at the much higher 74 s/sec.
I am curious about how everyone is coming up with their sample rates. I spoke to someone at Innovate and they said that the data is collected at roughly 1000 s/sec but displayed in logworks at approximately 12 s/sec. The raw data can be viewed if exported to an XML or spreadsheet. I can't find the info right now, but I believe Zeitronix may be collected at the same rate but is displayed at the much higher 74 s/sec.
I've never heard that Innovate product log at 1000 s/sec. Also, if you look every where on Innovate website and look though all the manuals, they all say 12 s/sec.
Jun 2, 2006, 06:28 PM
#26
But how does the Zeitronix interpret the voltage ..... a knock hit at 4k would be less than the noise floor at 7k.....you would have to watch it continuously and know what the value should be at the RPM you are at. This could happen in a blink of an eye. Reading timing would be the best way.
Jun 2, 2006, 08:01 PM
#27
I don't think it interprets anything.. you watch the voltage and make a judgement call.. Unfortunately everything that reads the sensor outside of the ECU will end up working the same way..
The knock tables in the ECU obviously are some sort of discriminatory value that it uses based on several conditions, from there it qualifies what it considers knock..
I was doing logging today and would see knock voltages of around 2v, which I consider high, but the ECU didn't always treat it as knock.. (stock ECU map for what its worth)
When I listen to the engine with earphones plugged into another knock sensor device on the engine, you don't hear much if any knock activity except a rare ping.. in other words, its doing what your hearing might, discriminating background noises to find what it considers real knock, I assume its based on the volume (voltage), amount of existing background noise, the frequency of the knock (both in signal frequency, and how many times it may occur)
This just isn't something most logging devices can do, you need to watch the ECU's reaction to those values very closely to determine if it thinks its knock or not.
The knock tables in the ECU obviously are some sort of discriminatory value that it uses based on several conditions, from there it qualifies what it considers knock..
I was doing logging today and would see knock voltages of around 2v, which I consider high, but the ECU didn't always treat it as knock.. (stock ECU map for what its worth)
When I listen to the engine with earphones plugged into another knock sensor device on the engine, you don't hear much if any knock activity except a rare ping.. in other words, its doing what your hearing might, discriminating background noises to find what it considers real knock, I assume its based on the volume (voltage), amount of existing background noise, the frequency of the knock (both in signal frequency, and how many times it may occur)
This just isn't something most logging devices can do, you need to watch the ECU's reaction to those values very closely to determine if it thinks its knock or not.
Jun 2, 2006, 10:45 PM
#28
Here is some screenshots of Zeitronix datalogs using the knock buffer on the user input.
This is the overlapped view, which I perfer, but look a little confusing in the begining.
The first screenshot has no knock, just the sound of the engine or nosie floor
This next scrren shot has an event of knock around 6800RPMs
This is the overlapped view, which I perfer, but look a little confusing in the begining.
The first screenshot has no knock, just the sound of the engine or nosie floor
This next scrren shot has an event of knock around 6800RPMs
