Hi All,
Looking into the data that Excel could not port in seems to be quite a
bit (~8 seconds missed). I therefore counted each break in the
waveform browser out of curiosity.
I counted 998 breaks and the actual total time recording is 12.747
seconds. This works out to 80 bps, so it's still very consistent
(reason I did this is that when I filtered the csv file in Excel, I
used a quick means of doing so [equations and smart filtering to make
quick work of 65000 rows of data]). So, this was a cross check
(there's nothing like physically counting to be sure).
All is good.
Gary is certainly correct that the gap is chaotic. There are times
when a break occurs right after a previous one and times when there is
no break for quite some time. Looks like the average over all balances
out nicely.
If you guys want, you can download the waveform browser (free
download) and I can just view the actual data collection file or save
your own csv or other types of data. That way you can see everything
the way I see it.
If interested, the waveform browser can be downloaded here: (quick
install).
http://www.dataq.com/support/downloads/newwwb.exe
The file is here:
http://www.classictesla.com/temp/gap15-60-NST-3.wdq
You can then just open the file and view it for yourself (if desired).
There is a lot of dead time in the beginning. I think I started the
coil about 28 seconds after I started recording (I had a difficult
time finding the plug in the dark).
This was a good bps test. Had the pc not lost the hardware device, I
could have recorded for a good 5 minute run at this sample rate (which
I still hope to do). I obviously need to lower the sample rate (sorry
it's so high, I just didn't want to miss anything). I may try it next
weekend on a pc with a real chassis and ground connection. If anyone
ever buys one of the DATAQ acquisition devices, I recommend a serial
port (stay away from the usb interface if possible). I'm using model
DI-158/UP/HS.
Take care,
Bart
bartb wrote:
Hi All,
I continued BPS testing today. I paralleled two 15/30's after phasing
for 15/60 power. I didn't want to change any coil parameters other
than power and a single 15/30 would have been just too small for the
cap size. I removed the safety gap and used 120Vac input via switch
to keep as many variables out of it as possible (no variac, etc.).
I did have some problems with the pc loosing the usb hardware device.
It captures data up to the point where the usb hardware device gets
lost. I'm aware of this problem (as is the manufacturer). I recorded
4.5 seconds of run time (actually a little more but Excel maxes out
at 65000 rows of data). I probably should have turned down the sample
rate. Hind site.
Here's a gif image of 15/60 run. Dead time on left of image is the
time it took me to run in the garage and turn on the coil.
http://www.classictesla.com/temp/bps_compress.gif
When I save the csv file, I set the marker just before the first
break (so data to left is not exported). Here is a picture of the
break area expaned with no compression.
http://www.classictesla.com/temp/bps_test_1560.gif
The Excel file identifying time stamps and break voltages:
http://www.classictesla.com/temp/gap15-60-NST-3.xls
The raw csv file:
http://www.classictesla.com/temp/gap15-60-NST-3.CSV
Ok, Javatc calculated 70 bps for this NST, cap size, gap, etc..
What I measured were 373 breaks over a time span of 4.542 seconds. So
82 bps.
Seems about right to me. I don't see a large variation in calculation
to measurement.
Take care,
Bart
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla