I'm a PhD student at the Institute for Nuclear Physics in Mainz, Germany and we have a KC705 Kintex7 evaluation board with a 4DSP FMC150 r1.2 ADC/DAC extension card. I was able to acquire some ADC samples and send them over Ethernet to my PC, where I simply plot them (see attachment): Channel A shows a nice 10MHz sine (directly from the DAC on the FMC150) as it should. However, Channel B shows a somewhat distorted 1MHz square wave from a signal generator (the output of the generator has been checked to be much more perfect than this plot...). Especially, note the significant slope of the Channel B signal where the square should be constant though. Furthermore, this effect appears for any reasonable frequency (10kHz to 1MHz) of the square wave.
My question is now how the signals are exactly coupled to the ADC chip on the FMC150, since that could be a reason for this distortion due to some "stray" capacitance. Furthermore, can you provide some more detailed documents how the FMC150 is designed? In particular, a schematic diagram showing the connection from Input A/B of the FMC150 to the ADC chip could be very helpful to us (thus detailing the Block "A/C coupling" in the FMC150 user manual). Of course, these documents would be kept in confidence if kindly provided. If you have any further questions, please don't hesitate to ask them.
Hello Andreas, It looks fine to me. It is an AC coupled card. On your square wave you have edge with a given slew rate. This is actually AC, then you have the DC portion. The DC portion is what looks like not flat. I believe that the dB difference between the highest point and the lowest point is the actual transformer insertion loss at this given frequency. The only way to display a square wave flat is to use DC coupling I believe. You could verify that with PSPice or any other simulator. I hope that helps, Arnaud
about 6 years ago
At least I found the schematics in another thread here: [url=http://www.4dsp.com/forum/index.php/topic,911.0.html]http://www.4dsp.com/forum/index.php/topic,911.0.html[/url] But maybe someone can generally comment on this?