Consult Logger - no play :-(

JCD · Post by **JCD** » Mon Nov 19, 2007 3:30 am

Matt,

Iâ€™ve been driving around today to gather some data about my setup.
Engine is CA18DET, O2 is Innovate LM1, address file is S13_CA18_256_E.adr.

I used â€œDisplay Air/Fuel ratio Traceâ€

JCD · Post by **JCD** » Mon Nov 19, 2007 6:49 am

Matt,
Meanwhile I had a look at those screendumps and there seems to be something wrong there too

I don't know if this is linked to the Player problem, first 4 screendumps were taken before the Log session started.
Iâ€™m not shure but I guess that I started the Log between the 4th and 5th screendump, or just after the 5th.
First three dumps are from one and the same â€œAFR 3D Traceâ€

JCD · Post by **JCD** » Mon Nov 19, 2007 6:51 am

last two

Matt · Post by **Matt** » Mon Nov 19, 2007 1:00 pm

Hi Jean

In regards to your log file, the heading title in the log file saved by NIStune is incorrect.

I had an open problem report on this but could not find any problem with TechEdge... so assumed it was okay. So it must only occur with the Innovate wideband code

It appears that NIStune is saving the display names to the file rather than the field names. That should be a simple fix so I'll get that done quickly...

To fix the current file you posted, open in Excel or notepad and edit the top line

LM1_O2,LM1_1,LM1_2,LM1_3,LM1_4,LM1_5

to

AFR,Aux1,Aux2,Aux3,Aux4,Aux5

This will get your log file recognised again and you should be able to play it back when offline.

Is there another way for saving those results

The AFR playback screen can be obtained from the logger file by doing a fast playback using the >>> button

The AFR playback contains a history of upto 32000 entries for each cell for AFR, Aux1-6 values but there is no save feature on this a such. Hopefully the fast playback is sufficient to get your results back.

Looking at that log trace, it appears that 'current' is actually showing 'average'

If you look at the Curr/Aver fields at the bottom they are always the same. Problem is from the screenshots is the wideband window is minised so I cant see the real AFR reading.

Doesn't matter... I'll have a look through the code and see what is causing that too

ByReaL · Post by **ByReaL** » Mon Nov 19, 2007 5:59 pm

JCD wrote: D4: some 32.00 in Load column 10 â€“ strange at least

regarding this problem, do the flowing test
logging unit connected to TPS and AFR and whatever senzors you want
Nistune board present .

start recording and afther that give a full throtle to the car or play a little with the gas pedal.

stop the recording

load the log file

select the TPS from Nistune board, and the TPS logged with the logging unit, and look at the2 graphs,
I have Type 1 board and Tech edge, and they are shifted a bit,

so AFR is from logging unit, and load and rpm are from nistune boards but since the 2 data chanels are not synchronized, you can experience such stuff. (i belive this is the cause for the probleb)

in my opinion a time shifting between the 2 data channels (nistune board / logging unit) must be introduced, in order to make it more accurate.

look at red and blue line, both lines are for TPS

so the AFR you are seeing in a cell (rpm x load) is not the AFR from that cell because AFR os taken from the logging unit, rpm and load are taken from nistune board and ..data is not synchronized...

Matt · Post by **Matt** » Mon Nov 19, 2007 6:08 pm

The latency issue was addressed in the latest version. However from the log you show above, there is still some latency there. So I must get the latency of techedge data to a minimum

I didn't try it out with the USB consult vs TechEdge so I will give that a run very soon

in my opinion a time shifting between the 2 data channels (nistune board / logging unit) must be introduced, in order to make it more accurate.

This isn't possible I'm afraid... the reason why is that there is two sources of data which are retrieving data from sensors, translating it to display values and then putting it into a data pool

If I delay the USB consult data going into the datapool then this will affect responsiveness of other things. The data pool is what gets logged and displayed on guages, windows, log player, maptracing, AFR logging etc

The data logger simply sits there and reads data every x amount of milliseconds and records to log file

I reduced the latency of the TechEdge wideband so that the timing between the consult data and wideband data was very close

Using the same test as you, putting TPS/AFM inputs into the wideband unit and measuring this

However the above change has caused a problem on slower machines with TechEdge which I need to investigate and released the 0.91122a test release on Friday

The innovate unit is different. It is event based, so we receive data as soon as the Innovate MTS library gives it us and put it into the pool straight away. There should be no delays introduced by NIStune software in this regard

ByReaL · Post by **ByReaL** » Mon Nov 19, 2007 6:22 pm

the only correct solving of this problem is to make 2 data channels, pools,

something like an oscilloscope with channel A and channel B, you see the 2 waves on same screen but you have controls for each channel.

this will be generally valid, each input (nistune type 1 or 2 or 3 board, techedge, LM 1, .....) will be considered as a different data stream, and with a time shifting options between the data channels, the user will be able to sync the data very precise

PS: that log file is old, i'll make a new one to see if the problem was fixed

Matt · Post by **Matt** » Mon Nov 19, 2007 11:43 pm

Best solution is to get both of the data streams at the same latency.

Even if I split the datapools, those both contain realtime data. Only when you get to logging could you then potentially try and put latency in... but how do you do this when the data is logged together realtime also? And in addition to that what are you going to record at start / finish. Its too messy to do it this way and a bit of a hack

With an oscilloscope, there are two separate channels, each with veritcal position etc.... but only one horizontal position. My Techtronix realtime scope has this. You can adjust the h position, h trigger point etc... but all data is captured and recorded in realtime.

Imagine three threads running, two threads reading a stream in and passing to a stream out. Data is valid only in time instances. If one stream has a latency reading actual data when it was received compared to the other stream, then this will affect the output. Make the latencies between all inputs the same by reducing to the minimum then data will show at same rate assuming data fed in has minimal latency

TechEdge has been improved, latency is now shorter than USB consult latency. Might have to speed USB consult wait time up a little. Its very close now compared to that older log

Matt · Post by **Matt** » Mon Nov 19, 2007 11:56 pm

changed USB consult to 30ms polling also... almost spot on now with wideband

Matt · Post by **Matt** » Tue Nov 20, 2007 12:23 am

Innovate code updated to record the correct heading entries now

The 32.00 would probably show up when the innovate has input just air (highest lean reading reached)

I've updated the code to now perform the averaging properly and display the count correctly

Reset button will reset the count and all values on the grid (and in history)

Also the AFR trace dropdown updated to display the external entries

JCD · Post by **JCD** » Tue Nov 20, 2007 7:17 am

It appears that NIStune is saving the display names to the file rather than the field names. That should be a simple fix so I'll get that done quickly...
To fix the current file you posted, open in Excel or notepad and edit the top line

That did the trick for my existing file!

If I understand this well, all other issues are fixed in NIStune 0.91223

Thanks for the good service. Iâ€™ll let you know how the new release behaves here; unfortunately this will not be until next weekend...

Also ByReal, thanks for your input

Matt · Post by **Matt** » Tue Nov 20, 2007 9:47 am

thats fine. keep me posted. I've got plenty to keep me busy in the meantime