QuickSilver2002

I have a wideband o2 setup and a logging device that allows me to log the wideband voltage reference at 240 samples/sec. The logging device has several open channels and I’m trying to figure out if there is a way that I can get RMP data onto the same graph. The device can measure between 0-14v, so I would need something in that range.

Does anyone know of a way to get RMP data via a voltage reference?

vette.se

I do not think it will work. The rpm signal are a frequeny up to (I think 6500hz) so you can not connect it directly to the logger. You will probably need a frequency to voltage converter.

:cheers:

kp

Curious who's wideband you have. I have an aussie DIY WB with the display and the display has a single output for logging. I was wondering what other options there was besides the FJO out there with some extra channels. Hard to complain about the DIY, I have 247.00 tied up in including the sensor - lots of soldering though...

QuickSilver2002

Kp,

I have the same DIY unit, but have added this device and softwarehttp://www.dataq.com/194.htm for data logging. RPM data would make this a really nice/cheap setup.


[Modified by QuickSilver2002, 12:12 AM 8/28/2002]

kp

thats pretty cool, I'll have to look into that - seems like a cheap way to get the data on a pc. They just released logging software for the diy 5300 LED display (has serial out) for the pc and palm. I have been using my palm m105 (100.00) for auto tap and I'll use it for the diy wb as well. Its not as nice as using my laptop but a whole lot less to lug around. It would be pretty nice if I could just figure out how to get the atap and wideband synced pretty close. I guess I could just start the logging on the palm (wideband) when going into wot and comparing it time-wise to the atap data if I could get the sampling rates between the two close. It would sure be nice to have rpm along with the wb readings all at once though.

easiest way to get rpm would be from an inductive pickup like FJO uses, I'm sure the signal would have to be amplified/converted but there are a lot of cheap inductive pickup handheld tachs I have seen - wouldnt be too difficult if someone (like me) knew what they were doing :)

And btw, if anyone is wondering what a DIY WB is.. Its a cheap fairly accurate wideband O2 do it yourself kit. heres a couple links..
http://www.techedge.com.au/vehicle/wbo2/default.htm
http://www.diy-wb.com/





[Modified by kp, 6:50 PM 8/27/2002]

QuickSilver2002

Kp,

It seems like you would need a frequency reading device to utilize an inductive pickup (I thought those were expensive). Let me know if there is a cheap solution.

As far as other ways of doing this go, I think I may have a solution to integrate with autotap data.

I'm pretty sure (although I have not yet verified) that autotap writes its timestamp based on the system clock of the PC (it looks like the format is based on seconds since midnight with 10ms accuracy).

I'm in the process of writing some software that will log the a/f ratio data from the wideband device with a similar timestamp. This should allow me to run autotap and the a/f logging software at the same time on different serial ports. I will then run a merge process that will find the best a/f record for each autotap record and combine the two together for analysis (the frequency of the a/f data should be much higher than atap).

I'm going to be busy the rest of the week and weekend with real work, but I'll try to post some results next week.

kp

My project has kinda came to a halt, I made the mistake of waiting until I was done building my kit to buy the O2 sensor and now no one has the darn things :(

Oh well, guess while I'm waiting i can play with LS1 edit and actually install the supercharger :)

David Lively

Thought about just getting the data from the OBD-II port? If your device has a serial port, Multiplex Engineering (Paul Bowen's company, great guy) makes an interface for about $70. RPM is one of the things that's actually EASY to get from the port.

I also have the Dataq 194 - nice board, works great.

DOCTOR J

To get a Tach reading for the Dataq logger, I used an LM2917 frequency-to-voltage converter chip. In the C4 I just tied into the coil to get the RPM signal (BTW, 6000 engine RPM is a 400 Hz signal to the tach). This is what I was doing with the tach & DIY-WB signals for AFR tuning (jpg is AFR volts & RPM vs time):
http://temp.corvetteforum.net/c4/doc..._a_10feb_s.jpg

If you go to the NatSemiconductor site http://www.national.com/ you can get a copy of the application notes for the LM2907/2917. They have example circuits to use for Tach apps. The chips are like $3 each from DigiKey.

Here is a pic of the last tach sender I put together (the box I had on hand was a little bigger than I needed, and I played around with a couple of different configurations before I was done, hence the wire mess on the board):
http://temp.corvetteforum.net/c4/doc...ach_box07s.jpg

Note that I used a regulated power supply (LM317) to give the board a stable voltage source. I also used Vbatt conditioning (clamp, caps, & zener) left over from the WB board to take the spikes out of the battery supply line (standard automotive practice).

FWIW, to get an RPM signal directly on the LS1, I think you could either tap into the 5v trigger to one of the coils, or MSD sells an opto-isolator to get a low v sync signal directly off a spark plug.

Feeding my tach circuit from the coil-driver side of the ignition module, I used a high-voltage mylar cap to isolate the input. I used ceramic trim pots to set the output to 1v=1000rpm.

I used a second 2917 chip (and a couple of high-brightness LEDs) to make a shift light - that's why I wanted a regulated voltage supply. It needs a stable reference if you want to turn on the LEDs at a specific RPM - again I used a trim pot to adjust it where I wanted.

One of these days I'll have to make a diagram of this thing & post it as a tech note. Hope this helps.

DrJ



[Modified by DOCTOR J, 4:21 PM 8/29/2002]

kp

Nice :)
The OBDII thing is another idea..
It would be great if atap had an additional voltage input channel :)

QuickSilver2002

Great info Dr J.

I'll have to look into building a circut if I can't get what I need from atap/odbII.

David L,

Do you have more info on the odbII interface? I went to the website, but could not find anything. Is there any sample code/libraries.



[Modified by QuickSilver2002, 3:36 AM 8/30/2002]

DOCTOR J

Thank you. Here are couple of other notes on data logging with a Dataq board, FWIW:

I used a 0-5 v channel for the DIY WB O2 signal. The RPM signal is on a 0-10 v channel.

(The idea of using the smaller ranges was to maximize the resolution of the Dataq 8-bit ADC for the signal being measured. I have an older Dataq board, and the Website had info about altering the circuit for different input levels. Haven't used their new kit yet, but the old board is great.)

I built the RPM signal & WB O2 signal as stand-alone boxes (not tied to the ALDL) for a particular reason. My plan was to tune the WOT fuel curve in the C4, then move the whole test rig over to my boat (non computer engine) and tune it the same way.

That was the plan. Then I got to looking at the effect of headers & different intakes on the C4. Suddenly I notice it's Labor Day Weekend, and I still haven't tuned the boat yet ...

The other point to note is on the RPM plot: I use it for a 'dyno' evaluator.

If you consider the slope of the RPM vs. time curve (that's a WOT 2nd gear pull in the previous jpg), it can be related back to acceleration or kinetic energy. I use the slope of the curve (on a common test basis) to evaluate changes to the ECM programming. It's a quick & dirty indicator of how changes affect power output - at least for acceleration. To make that work, I had to get an RPM signal that had not been 'filtered' through the ECM. Hence the stand-alone RPM transducer.

DrJ

David Lively

The manual for the interface is here: http://www.multiplex-engineering.com/tech/manual/

That describes the protocol pretty well. The actual message definitions are in a couple of SAE docs, which I have. If ALL you're after is RPM, then the interface may be a bit of overkill (ie, you're going to need a laptop or a BASIC STAMP or other such device with a serial port). I have a pretty extensive database of GM proprietary messages and response formats including pretty much everything engine-related. (I can't post this, obviously, but if you have a question...)

The Multiplex Engineering interface also supports a "Monitor mode," which lets you just sit back and watch the network traffic. Quite educational at times, if your a codemonkey.

QuickSilver2002

Great info guys.:yesnod:

It turns out I was looking at the wrong Multiplex Eng website before. This is really intresting, I'm going to have to play with this stuff a bit more.

For now it looks like I have a workable solution that merges with an autotap log. I have the code written, I just need to go gather some data and see what the results look like (that will have to wait until Monday though)

I'm just running autotap on one com port and my code and the dataq board on the other. My code simply uses the ultimateserial active x control and writes another log file that looks just like an exported autotap file. I then merge the 02 data into the autotap log by finding the closest o2 reading that is within 10ms of a defined autotap parameter. I’m hoping this will result in some useful data.

I'll post the results after I do some testing on Monday.

David Lively

Sounds like a great approach. What are you working in? VC++, VB, Delphi?

QuickSilver2002

Ok, got some preliminary results yesterday and it seems to work very well. I had one small bug (aren’t they all) with the format of the timestamp, so I still don't have any graphs or anything, but it looks like the data will be very useful.

I will hopefully have time to do some logging tomorrow night and post a few graphs...

I actually wrote the code in c#/.Net. It's what I use most often these days in my work and has good support for ocx controls. It was really easy to do, and only took me about 3 hrs from start to finish. It's nothing fancy, but it does use an internal queue and a low priority background thread to do the file IO (this frees up the foreground thread to gather the results at a very high sampling rate).

David Lively

Nice - I haven't got to play with C-sharp yet, though I hear great things about it. Wave of the future.. again.

QuickSilver2002

Ok, this is really strange.

I finally got some logging done, but noticed that the wideband was reading a little lean while just cruising around as compared to before. It was reading exactly 14.7 before, but it was now up to around 15. After further research, I realized that hooking up the Autotap cable to the laptop caused the voltage readings from the dataQ board to jump up a bit (from around 2.5 to around 2.6+). It is really strange. I’m using a serial to USB converter on one of the cables, so I tried moving that between the two and the results were exactly the same. I can reproduce it every time I hook up the autotap cable. Besides this problem, the solution works great (I never imagined that this would be a problem).

I checked the voltage at the actual WB vout with a multi meter and it was correct there. I’m not sure if the DataQ board is sending out the wrong voltage in the packet, or if some other bug is causing the reading of the packet to be thrown off.

If anyone has any ideas on why the voltage reading would jump, please post.

I’ve got one of the displays to build that has rs232 output, so I guess I’ll give that a try after I get around to building it. It could be a while since I’m working on my SC install now.

DOCTOR J

QuickSilver2002

Sounds like it might be a ground reference problem among the ALDL port, the Oz WB, the Dataq, and the PC.

FWIW, check the circuit diagram & fabrication notes of the original DIY WB. It has an isolated power input available, to separate the output signal from the vehicle ground - thus preventing ground loops between it & the measuring device.

Maybe you need to power the WB output buffer from the PC (or carefully try to tie the grounds together somehow ?).

I haven't run into the problem, but I only ground the PC through the Dataq board or the ALDL input at one time - haven't tried to run both at once.

DrJ

QuickSilver2002

Dr J,

I think you are right, but I'm not sure how to solve it. I'm not an electronics expert by any means. My limited knowledge makes me think the current is simply finding another path to ground. The readings on the input side were correct so it must be within the Dataq board. I’m definitely over my head at this point.

I wonder if the increase is consistent and that I could just calibrate against it.

Perhaps I could also run the ground on the dataq board directly to the chassis and see what happens.