Problem reading the frequency of more than one signal
3 replies to this topic
Posted 08 February 2013 - 03:02 AM
Hello, I am willing to use 3 humidity sensors (File n°1) using the frequency signal output circuit recommanded by the manufacturer (File n°2) and a U6 module with Labview. When I use one circuit, with one sensor and the "U6 Four 32-bit Frequency Measurements.vi" (configured as in file n°3), using only FIO0 and Ground, I get a very good reading (reaching the sensitivity of 0.6pF/%HR specified in the data sheet). File n°4 shows the frequency measure and File n°5 the capacity calculation from the frequency measures. The strange fluctuations in the middle of the test are linked to a problem with our humidity chamber over 70%HR (this sensor has the same behaviour as our other kind of sensors that were running at the same time on another system (not connected to the same computer)). So everything is correct, and the signal has no noise either. My problem accurs when I want to read more than 1 sensor. As you can see on File n°6 (program on File n°7), when 3 sensors are connected with three diffrent circuits (but same 5V power unit) on to FIO0, FIO1, and FIO2 I get this kind of result : - most of the time I get a lot of noise - some times (mainely on the first half of the test and at the end of the test) the noise disappears and the 3 signals are all exactly the same as you can see, just as if the Labjack unit was reading three times channel 0 for instance... PS : each step represents a 5%HR humidty change PS : I am measuring and recording the data every 10 seconds Would you have an idea of is going on? Am I missing something in the way the configure the 3 FIO channels? Thanks a lot, User
Posted 08 February 2013 - 10:39 AM
What if you run your 3-timer VI, but only have 1 circuit connected? Does the 1 look good? What if you have 2 connected? Do they both look bad, or is it only when you connect 3? What if you use a fixed 6000 Hz frequency rather than your actual signal? Do you have a frequency generator you could use to created a fixed square wave that you connect to all 3 channels? An external frequency generator would be the best test, but you could also use the 4th timer on the U6 in Frequency Output mode to create a fixed square wave. How about if you connect 2 of your signals and 1 fixed frequency? These tests are trying to isolate whether there is a problem with your VI or the U6 timer measurement, or if the problem is that the signals actually get noisy for some reason.
Posted 12 February 2013 - 09:14 AM
Hello, Thansk for your answer. I have made some more tests regarding your advice. The recording rate of the data is the default one set in the VI (every second). Pic 1 : FIO0,1, and 2 are connected to a function generator output (square) set to 6315 Hz. The value is correct despite some small negative spikes periodically Pic 2 : FIO0,1, and 2 are connected to circuit n°1. The measure is correct, all inputs measure nearly exactly the same value Pic 2Bis : Zoom out of Pic 2 Pic 7 : FIO0 Sensor1 with circuit 1 FIO1 Output of the sqaure wave generator (different value than for Pic 1 but I forgot to wright down the value I set it to, but I believe the measure is correct) FIO2 Sensor2 with circuit 2 Circuit 1 and 2 are connected to the same 5V power source I blowed of sensor 2 (decrease of the frequency as you can see). Sensor 1 was protected in a small bag so it wasn't affected by my blowing Pic 9 : FIO0 Sensor2 with circuit 2 FIO1 Output of the sqaure wave generator FIO2 nothing connected Circuit 2 connected to 5V power. Circuit 1 completly removed Pic 10 : FIO0 Sensor2 with circuit 2 FIO1 Output of the sqaure wave generator FIO2 nothing connected Both circuit connected to 5V power (but circuit 1 not measured by Labjack) It seems that when one circuit is connected to the 5V power supply, the labjack reads a correct value. Blowing over the sensor makes the frequency decrease and the slow move back up is smooth and nice. The value measured form the output signal of the square wave generator is correct to. THe problem occurs when 2 circuits are connected to the 5V power supply. In this case, I see exactly what happend in the my last post (in the humidity chamber) : When the humidity is stable both channels measure nearly exactly the same value (I doubt that the 2 sensors and 2 circuits are similar to that exctent so that is strange). When humidity changes at least for one of them (blowing) both signal get noisy. When the sensor recovers form the change and gets close to the second one, it jumps (up or down) just as if the U6 module couldn't make the difference between the 2 channels when the difference gets small. PS : I made some other tests with another sensor (same principle, other manufacturer, but working around 40 kHz this time) and this is what I have observed : When I set the frequency of the wave generator close to the frequency measured with those 2 new sensors (that show like above exactly the same value....that is strange again) and that I move around this particlar value, the signals of the 2 sensors "stick" to the frequency of generator. If the frequency of the generator gets to far from the "real value" of the sensors, they "unstick and the frequency generator doesn't affect the 2 signals anymore. I can't make this phenomen accur whith the sensors we have been talking in this conversation. Thanks User
Posted 13 February 2013 - 05:55 PM
We tested a U6 with three frequency inputs and used a waveform generator as one input to try and add noise and get the frequencies to be skewed by changing its output frequency and were not able to reproduce your issue. You may want to look at the power being supplied to your 555 timer's. They produce a decent amount of noise which could be why you are able to reproduce your issue when having a second circuit powered but not being read. Check the datasheet for your devices first, but you can use a DAC output to power your sensing circuit if it can be properly run at less than 5V. If you power the timer at a 4.5V-4.75V level the DAC outputs will be able to cancel out some of the noise to deliver steady voltages. You may also want to look into powering your sensors from an external power supply to debug your problems. A simple solution may be to add decoupling capacitor's between VS and GND right next to your 555 timers if you haven't already done that. Its a fairly basic circuits concept but many people forget to do this so it is worth mentioning.
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users