Zeitronix WB Settings in v1.5

[Magic_Chicken]

I haven't heard complaints from any other Zeit users.. I think 5 or so requested it.. I wonder how they are doing...

It appears I'm the only one actually using it so far. What are the chances that the user defineable inputs could be used in the histogram? Say have a radio button or check box in the setup for if it's a wideband input or not.

[Windy_City]

I got a chance to check idle A/F with my Zeitronix to compare to HPT. Just like the others, it wa off by 1.00. The display reads 15.6 - 15.8 and HPT reads 14.6 - 14.8.

[Keith@HPTuners]

Do you guys have a formula that works?

Magic Chicken, perhaps send your contact a link to this thread?

[Magic_Chicken]

I've had many discussions with the Zeitronix guy, he basically says his is right and there is nothing he can do to fix it even though I've sent him a few spreadsheets showing that it's not right. I have another data set that I need to try, will update with progress as soon as I can get it done.

[2000Firehawk]

Ok, I did a 3 min idle log today with 2 laptops,
on a warmed up car.
One with Zeitronix Data Logger running.
The other with HPTuner.

I put HPT white & purple wires to the Zeitronix
output white wire to log both WB & raw-voltage
with the HPT scanner. Zeitronix WB is shown in
the Histogram.
I guess raw-voltage isn't native to Histogram.
I reversed the order in the table & couldn't get
raw-voltage to show in Hisrogram ? If I just used
raw-voltage as INPUT 1 with no other inputs, will
it show in the Histogram? I should've tryed that ?

Zeitronix file was 17,000 frames long with 99.9%
of the AFR readings at 14.5 / 14.6 / 14.7

The HPT log was 1000 frames with a Histo avg = 14.3
AFR readings were all over the place from 13.xx - 15.xx

I did record Zeitronix raw-voltage on the HPT log.
Voltages were all over the place.

So Zeitronix software is showing an average of real close to 14.7
HPT average is 14.3

I sent Magic Chicken my logs for his review...


Dave

[Magic_Chicken]

Here are the voltage averages according to 2000Firehawk's HPT log.

13.21 2.642
13.24 2.571
13.27 2.529
13.3 2.57
13.33 2.62825
13.36 2.566333333
13.39 2.512125
13.42 2.55
13.45 2.334
13.48 2.55725
13.51 2.564625
13.54 2.574166667
13.57 2.43625
13.61 2.500833333
13.64 2.5062
13.67 2.487545455
13.7 2.493083333
13.73 2.4265625
13.76 2.463210526
13.79 2.49675
13.82 2.462642857
13.86 2.452947368
13.89 2.448521739
13.92 2.409047619
13.95 2.429740741
13.98 2.4154
14.02 2.410826087
14.05 2.40618
14.08 2.457
14.11 2.406608696
14.14 2.4212
14.17 2.396875
14.2 2.38171875
14.24 2.41205
14.27 2.339444444
14.3 2.351032258
14.33 2.330407407
14.36 2.374888889
14.39 2.340058824
14.42 2.375421053
14.45 2.346870968
14.48 2.317193548
14.51 2.29028
14.55 2.32825
14.58 2.333769231
14.61 2.309142857
14.64 2.292647059
14.67 2.285772727
14.7 2.292621622
14.76 2.2701875
14.82 2.270095238
14.88 2.233055556
14.94 2.269785714
15 2.231875
15.06 2.211777778
15.12 2.224
15.18 2.194125
15.24 2.2862
15.3 2.301857143
15.37 2.266285714
15.45 2.204666667
15.52 2.151666667
15.59 2.1738
15.67 2.1875
15.74 2.1895
15.81 2.1485
15.88 2.3145
15.96 2.16575
16.03 2.234
16.12 2.241

Here is the AFR his Zeit log showed, much different than the HPT log.

14.3
14.4
14.5
14.6
14.7
14.8
14.9
15
15.1
15.2
15.3
15.5
15.6

Data from my HPT and Zeit logs are much different from his. I don't think it will be possible to have native support for Zeitronix and it be accurate. I think our best bet for Zeitronix is for each individual to come up with their own data set and input it manually. If that is the case, then I quote myself from a previous post.

"What are the chances that the user defineable inputs could be used in the histogram? Say have a radio button or check box in the setup for if it's a wideband input or not."

Ideas?

EDIT: Corrected his Zeit log, he didn't have voltages being logged.

[2000Firehawk]

I wonder what the output looks like going to the Zeitronix Data
Logger. Maybe it's a different 'filtered'/'buffered' 0-5V.
The serial connector snaps apart easily, and just 2 pins, 1 & 4
I think, are being used. Just a signal with a return. It would be real
easy to splice into these two wires inside the connector, & use as
the signal/ground to HPT. Only if it's 0-5V.
Eric, maybe you can ask your Zeitronix contact if this can be done ?
Maybe it's a better output than whats on the 'white' wire ?

Hmm.. Maybe not , There's a bunch of signals going to the Data
Logger software...

Dave

[Badassbird]

I am also having issues here as i finally got it to work and they are way off base. Any word here guys?

[Magic_Chicken]

I wonder what the output looks like going to the Zeitronix Data
Logger. Maybe it's a different 'filtered'/'buffered' 0-5V.
The serial connector snaps apart easily, and just 2 pins, 1 & 4
I think, are being used. Just a signal with a return. It would be real
easy to splice into these two wires inside the connector, & use as
the signal/ground to HPT. Only if it's 0-5V.
Eric, maybe you can ask your Zeitronix contact if this can be done ?
Maybe it's a better output than whats on the 'white' wire ?

Hmm.. Maybe not , There's a bunch of signals going to the Data
Logger software...

Dave

Short answer, it can't be done.

Long answer (quote from Zeitronix):

The O2 sensor doesn't output a voltage which corresponds to the AFR. The Zt-2 box drives the sensor and the bidirectional current from the sensor is an equivalent of the AFR. It's more complicated than splicing into a sensor wire. On a top of it the Zt-2 regulates the O2 heater keeping a constant temperature of the sensing elements regardless of the EGTs based on internal parameters for the sense cells. Tapping into the sensor wires will disturb this information and will make the heater loose the regulation. As a result you will have false AFR readings.

I also talked to Zeitronix about other engine management software and how it is done with them. The only one I researched was AEM's EMS and they have an adjustable gain setting to dial in the WB voltage differences.

After logging both the voltage and AFR with my Zeitronix, I don't see how even the Zeitronix is accurate, but he says this:

The actual logged AFR is correct. It's being processed internally by the microcontroller. The User input is heavily filtered. Since we cannot control customers installation, grounding and type of signal it's going to be connected into the User input it requires filtering, otherwise you could see a data fluctuation of logged User input. This introduces a delay between the internally processed AFR number and logged User input. Moreover the conversion of those two A/Ds happens at different time.

I also asked him why he isn't using a linear voltage conversion like the rest of the companies that use the same sensor and this was his answer:

We didn't choose non linear AFR/V polynomial. It's a natural Bosch wideband O2 curve. It "bends" around 14.7 due to a different chemicals in sensors cells to capture both lean and rich regions of the AFR.
The wideband controllers you are referring to artificially linearise the Bosch O2 sensor output.

Our only choice that I can see is each user make their own custom table for their Zeitronix. Which kind of sucks because we wouldn't be able to use the histogram, but I don't really see an alternative at this point. (unless my idea up above is incorporated into the scanner, hint, hint )



I'm open for suggestions, anyone else have any ideas? It sure does suck if we all just wasted $300-500+ on a wideband that won't work for the main purpose that we bought them (at least my main purpose was HP Tuners integration).

[2000Firehawk]

Analog Wideband Output (for your favorite Standalone Engine Management System)

0.1 AFR accuracy


CONTACT US
The most efficient way to contact us is via e-mail [email protected] . We check our mailbox several times a day and answer your questions promptly.

You can also use a snail mail:
Zeitronix
P.O. Box 9125
San Pedro CA 90734

WARRANTY
Our products carry 6 months manufacturing warranty.
All returns and warranty repairs need to be authorized. If you have problems with our product contact us, 99% issues are being resolved remotely.
We do not refund sensors.

[Magic_Chicken]

Unfortunately, it does have the analog wideband output, so you can't get him there. When I first got my Zeitronix I was going to send it back for a refund because it wasn't working right (turned out to be a faulty wiring harness) and he said he can't give refunds on installed sensors, so I would've had to eat the cost of the O2 and EGT sensors. Needless to say, I didn't send it back and it's probably too late now anyway (for me).

[2000Firehawk]

This time I did a very controlled test.
Car was warmed up 1st.
Laptops were started up with the car running.
Car at idle, in park, for 3 minutes, with ALL AFR data in
just one cell of the Histogram this time.
30 MAP @ 800 RPM = 14.3 AFR [ Histogram average }
Results of approx: 1000 frames.

HPTuner user inputs 1 & 2 wired to Zetronix output wire.
User input 1 = Zeitronix WB
User input 2 = raw voltage
Excel averaged exported HPTuner log.
AFR = 14.2938 RAW VOLTAGE = 2.3582

Zetronix analog output table shows
AFR 14.2 @ 2.34 Volts
So the HPTuner #'s of AFR 14.2938 @ 2.3582 Volts,
falls right in line , on the 'curve' , with the Zeitronix
WB Analog Output Volage vs AFR table on the 'green'
installation sheet.

At the same time I logged on my other laptop the
Zeitronix Data Logger figures.
The same 3 minute time period. Excel averaged.
AFR = 14.6214

Dave

[Magic_Chicken]

This time I did a very controlled test.
Warm car at idle in park for 3 minutes with ALL data in
just one cell of the Histogram this time.
Results of 1000 frames. Excel averaged.

AFR = 14.2938 RAW VOLTAGE = 2.3582

How were you logging AFR and Raw Voltage at the same time? From the sounds of the above you did both in 1 cell? I'm not quite understanding that.

[2000Firehawk]

Sorry about that. All my AFR in the Histogram was with one cell
only as the car was warmed up & both laptop started at the same
time with the car running in park. Only one cell was 'hit',
[ 30 MAP at .8 RPM ]. The 'raw-voltage' showing on 'table' display.

You can log a WB + 'raw-voltage' at the same time with HPTuner.
WB shows up in the Histogram & 'raw' is table/ect. displayed.
I wanted to see both, to see if the HPTuner Zeitronix 'native' WB
pre-defined 'setup' was ok.

I have both HPTuner inputs 1 & 2 connected to WB 0-5V output wire.

When I 'export' the log to a Excel file, both AFR & RAW #'s can be
averaged with Excel 'tools'.
Excel AFR average = 14.2938, HPTuner Histogram average = 14.3 ,
in my 'test' .


You know, it looks like the Zetronix 0-5V is off by exactly .4 AFR ??

Maybe they are just plain stupid, or maybe it just happens to be off
by exactly .4 AFR at this point, [14.7 AFR], on the 'curve'.

My next step is 13.0 AFR 'MafLess' ..

I'll do the same 3 minute 'test' with the car warm & running.
If it's off by exactly .4 AFR at that point on the 'curve' , then we
have a 'good' 'native' formula.
We just have to add .4 AFR to all our data...

I do know in my case, that 14.3 AFR in the Histogram average or
with Excel 'data' averaged is 14.7 AFR for me... :

Dave

[Magic_Chicken]

Did you pipe the Zeit's analog output to both HPT user inputs 1 and 2 to log AFR and voltage? Each cell in the scanner's table corresponds to a different input on the EIO pigtails.

[2000Firehawk]

Yes..

[Magic_Chicken]

I'm so sick of rewiring my EIO to try different things! lol

[Keith@HPTuners]

Dave, please e-mail me your spreadsheets so I can put them into the software and let some other users test.

That is a GREAT method for testing the accuracy of the data. Kudos.

[2000Firehawk]

Hi Keith,
I don't have any spreadsheets. I just did a 'controlled' test to see if I can make any sense, of the Zeitronix WB Analog 0-5V output.
Both HPTuner inputs 1 & 2 wired to Zeitronix Analog 0-5V ouput, plus the 'proper' HPTuner & Zeitronix grounds, grounded. This way I can log AFR in the Histogram + 'raw-voltage' into the 'table'.

1st 3 minute log at STOICH 14.62857, The Zeitronix Data Logger software gave a 14.6214 average. WoW, super, DEAD on the money....
Zeitronix Analog output averages via HPTuner software = 14.2938 AFR , 2.3582 volts.
Histogram average = 14.3 So, 14.3 analog vs 14.6 True AFR reading.

2nd 3 minute log at a 'mafless' idle . The Zeitronix Data Logger software gave a 13.6976 average. Zeitronix Analog output average via HPTuner software = 13.4347 AFR, 2.0923 volts.
Histogram average = 13.4 So, 13.4 analog vs 13.7 True AFR reading.

In both cases, Zeitronix analog reads appox: .3 AFR lower than True AFR reading.
In both cases, these analog readings appear to 'fall' on the Zeitronix Analog supplied table 'curve'. Maybe, instead of ,[ + 9.7758 ] ,at the end of the Zeitronix Analog 'formula', maybe it should be, [ + 10 ] ??

AFR = 0.627*V ^2 + 0.457*V + 10

Could it be something this simple ??

If not, maybe you guys can 'make' us a user-defined 'native' to Histogram, function, so that anyone with an 'off-the-wall' brand 'X' WB, can input their own AFR 'formula', so we can use the Histogram AFR feature.

[2000Firehawk]

AFR =1.3127*V^2 - 1.2163*V + 10.428

Hay,
I was just re-reading this whole thread, & started 'playing' with this 'formula' , that Magic Chicken got from Zeitronix.

I like it...

1st 3 minute log at STOICH 14.62857, The Zeitronix Data Logger software gave a 14.6214 average. WoW, super, DEAD on the money....
Zeitronix Analog output averages via HPTuner software = 14.2938 AFR , 2.3582 volts.
Histogram average = 14.3 So, 14.3 analog vs 14.6 True AFR reading.

*** 2.3582 is 'raw' averaged Volts.. ***
*** 2.31 in above 'formula' = 14.6230 AFR ***

2nd 3 minute log at a 'mafless' idle . The Zeitronix Data Logger software gave a 13.6976 average. Zeitronix Analog output average via HPTuner software = 13.4347 AFR, 2.0923 volts.
Histogram average = 13.4 So, 13.4 analog vs 13.7 True AFR reading.

*** 2.0923 is 'raw' averaged Volts.. ***
*** 2.107 in above 'formula' = 13.6929 AFR ***

My 'controlled' testing must have a plus/minus 'error' factor ??
So maybe this is the 'formula' to use ?
What do you think ??