Hi KNAI,
I think that your problem is in the host code. Converting the U32 from the DMA to the FXP you might have missed an important step.
In the boolean array to number you need to specify what type of representation you require, or supply a data example in FPGA.
You will need to right click on the boolean array to number and select properties, then set the output configuration to the type of FXP that was generated by the FPGA.
In your case this is (deselect adapt to source) and set the FXP to unsigned 24 bits word and 10 bits integer. The VI will then be able to convert the U32 into the correct type of FXP.
Beware that when the FXP value crosses the X-axis there is a discontinuity in the U32 representation and you should beware of this if you use averaging on a U32 with integer maths, for straightforward conversion this is not a problem. You just have to remeber that when converting U32 to FXP you have to set up the output conversion in the boolean array to number properties.
Hope this is of some help.
Cheers
Stephen