07-23-2014 03:23 PM
The terminal interface is just a card with screw terminals to interface with the 68 pin connector. It is true that we are using the 5752 in a single-ended mode by connecting the - to the ground. As for the impedance mismatch I can get a signal directly with the array that has as an impedance of 12 Ohm in the resistance and -130.5 Ohm in the reactance. I would think that this would be more of a mismatch than the 50 Ohm to 100 Ohm. We have also tried matching by using an inductor resistor circuit in addition to a 100 Ohm resistor between the + and ground (as the draw down resistor). Of which each was tested separately, but with no improvement to the signal. As for the harmonics in the signal I have attached an FFT of the signal. Looking at the spectrum it shows the 6.5 MHZ as the dominate frequency, with the 2.25 MHz more of a sub harmonic. Is it not true if the 6.5 MHz was a harmonic then the 2.25 MHz would be the dominate frequency.
I apperciate your input and help.
07-23-2014 05:45 PM
Are you sure that plot is correct, because if it is, then you don't have a 6.5MHz resonant signal, its a ~650kHz resonant signal...
07-23-2014 05:59 PM
I am sure it is MHz there is an error in my Matlab code on the scaling of the x-axis, I had just missed it.
07-23-2014 06:07 PM - edited 07-23-2014 06:08 PM
Just to clarify, the image you attached indicates:
The amplitude flatness of the 5752 below about 1MHz degrades, and this changes somewhat with the configuration of the ADC. If you have a fundamental resonance at ~335kHz (e.g. due to cable lengths from your source, through switches, to the FlexRIO Adapter Module), then it's feasible that measured amplitude of harmonics vary from (are greater or less than) the ~335kHz fundamental.
General questions:
From your signal source to your 5752, can you estimate for the total cable length (cables/wires/switches/etc)?
Do you have access to a time domain reflectometer? I believe that would be the best instrument to use in order to determine if there are impedance mismatches along any of the path length of the signal to your 5752 that are causing a ~335kHz resonance. A TDR will provide you a plot of characteristic impedance (in ohms) vs. distance, so you can characterize what your signal sees as it goes through your system's cables, switches, etc.). This will help you determine at which point in your signal path you may have a temporary shift in impedance (an impedance mismatch), which would cause reflections at a particular frequency.
-Andrew
07-23-2014 06:36 PM
Sorry for any confusion but there is an error in the scaling of the x-axis
the fundamental would then be 3.35 Mhz
the 2nd 6.71 MHz. I know because I counted 6 cycles in 1 msec of the signal in the time domain.
I can estimate that there approximately 1.5 meters of distance from the source.
Unfortunately I don’t have a domain reflectometer. Is there another way to test this?
Thanks for your help
07-31-2014 01:13 PM
6 cycles in 1ms...
6 cycles/1ms * 1000ms/1s = 6000 cycles/sec = 6kHz.
Are you sure your scales and math are correct?
07-31-2014 01:45 PM
It looks like I can't type as fast has I has thinking and got a little frustrated because the pin drive with the data took a dump so I could not correct the scaling of the FFT properly. I counted 6 cycles in 1 micro second
6 cycles/1 microsecond * 1000000 microseconds/1sec= 6MHz
Again I apologize for the confusion. I am trying to get a hold of the TDR to check the cables.
10-14-2014 03:51 AM - edited 10-14-2014 04:01 AM
It is possible that your preamp (30bd gain!) is oscillating. Scope with input 50Ohm impedance is fine but 100Ohm from the 5752 can cause reflections leading to oscillation.
I would test with differential input shunted with an additional 100Ohm SMD resistor (rigth at the input) and a 1Meg (to 10k) from AI- to AIGND to provide a bias path.
Some amplifiers a sensible to capacitive loads , thats why impedance matching and short cables are important.
What will happen on the scope if you add another 50Ohm (maybe build from two 100Ohm SMD or a 49Ohm) in series at the scope input leading to a 100Ohm missmatch input impedance?
And use SMD resistors