02-21-2011 07:31 AM
I need to measure a frequency of 1 Hz with accuracy less than 1 ppm
icant do it with a PXI 6608 its normal ?
1 Hz is measured 1,000007 Hz by a PXI 6608
in this case, the error is 7ppm but i need less than 1ppm, i should measure 1,0000007 to have 0,7 ppm
please help its so urgent !!!!!!!!!
02-22-2011 07:34 PM
Hello,
The signal you are connecting to the device, what is the accuracy of the signal? Does it have accuracy much better than 1ppm? What device is creating this signal?
As for setting up the PXI-6608, the specifications manual for the device states that the PXI-6608 has an 10MHz OCXO with accuracy 5ppb so I would use this as the source of the counter. The user manual for the device on page 20 states that this is possible. When looking at the counter input frequency measurement, it sends your 1Hz signal to the gate of the counter and then measures how many pulses from the timebase that occur in 1 period. Let's change the timebase using the channel property node in the counter input section called CounterTimebase.Source to the 10MHz OCXO clock.
02-23-2011 09:22 AM
Hello Jim
the device that we used is a board based an STM32 µC that generate a 1Hz signal pulse width is 30µs (Ton = 30µs , Toff = 1 - 30µs)
we tested the accuracy of the signal with a calibrated Agilent and we have 0.7 PPM
we use Labwindows CVI, so how can we change the timebase source for the Low frequency to the 10MHz OCXO to use as a source of the counter ?
thks
02-24-2011 10:21 AM
hich1109 a écrit :
Hello Jim
the device that we used is a board based an STM32 µC that generate a 1Hz signal pulse width is 30µs (Ton = 30µs , Toff = 1 - 30µs)
we tested the accuracy of the signal with a calibrated Agilent and we have 0.7 PPM
we use Labwindows CVI, so how can we change the timebase source for the Low frequency to the 10MHz OCXO to use as a source of the counter ?
thks
hich1109 a écrit :
Hello Jim
the device that we used is a board based an STM32 µC that generate a 1Hz signal pulse width is 30µs (Ton = 30µs , Toff = 1 - 30µs)
we tested the accuracy of the signal with a calibrated Agilent and we have 0.7 PPM
we use Labwindows CVI, so how can we change the timebase source for the Low frequency to the 10MHz OCXO to use as a source of the counter ?
thks
we tested the accuracy of PXI 6608, so we connected the CTR0 source (PFI 34) with the CTR 1 gate (PFI 36) to generate a signal by 6608 like our Board based STM32 generate, and we measured the frequency we had good results, so OK for the PXI 6608
After that we connected our board to CTR1, we had bad results >1ppm
May be a bad impedance adaptation of signal between our board that generate 1 Hz and the PXI 6608?
thks
02-24-2011 05:47 PM
Hello,
In order to set the counter source to the timebase, you can use the the get/set/reset CI_CtrTimebaseSrc in order to set this in CVI. Please consult the DAQmx C reference Help for more information. Here are the functions if you can't find it in the help.
int32 __CFUNC DAQmxGetCICtrTimebaseSrc(TaskHandle taskHandle, const char channel[], char *data, uInt32 bufferSize);
int32 __CFUNC DAQmxSetCICtrTimebaseSrc(TaskHandle taskHandle, const char channel[], const char *data);
int32 __CFUNC DAQmxResetCICtrTimebaseSrc(TaskHandle taskHandle, const char channel[]);
As for performing a loopback test to test its accuracy, this isn't the best way to do this because the source is creating a signal with relatively the same accuracy as you are reading in back on the same device. Knowing that this signal is accurate down to .7ppm, we need to apply a source for the counter that is much more accurate than this. First way to do this is to change the timebase source to the 10MHz clock like I mentioned above. The other way is to place the PXI-6608 in slot 2 so that it overwrites the 10MHz clock on the backplane. This 10MHz clock is used with PLLing the 80MHz clock that is normally used in counter tasks. Which slot do you have the PXI-6608 in? Could it be placed in this slot?
02-25-2011 02:11 AM
Hello Jim,
The PXI 6608 is plugged in slot 2 of Chassis backplane and the jumper S1 is in slot 2
but, i like to change the measurment time in the low frequency to have more accuracy is it possible ?
thks
02-25-2011 09:04 AM
Hello,
Since you have placed NI-6608 in slot 2, the accuracy of the counter is in the 5ppb range because the 80MHz clock used for the counters is PLLed to the 10MHz OCXO timebase. You can change the counter timebase as I mentioned above to the 10MHz OCXO timebase using the function above. I am not sure if I have answered your question though so if I haven't, could you explain your question in a little more detail?