Why does it take more elements to view waveform after Peer-to-Peer than there are samples in the original waveform?

[Trekkie123]

I am generating an arbitrary waveform with the PXI-5412 Arbitrary Waveform Generator and reading them with a PXIe-5122 Oscilloscope. I am then streaming the waveform to a PXIe-7846R FPGA module with Peer-to-Peer Streaming and sending the waveform back to the Host through a DMA FIFO. However, when I graph it, there seem to be gaps in the waveform. Any idea what would cause this? I have attached a screenshot of one graph when reading directly from the waveform generator, and one when passing the waveform to and back from the FPGA where the gaps appear.

[Trekkie123 · Accepted solution]

The issue has been resolved. I created a separate while loop with a wait node set to the acquisition time.

[Trekkie123]

I am generating an arbitrary waveform with the PXI-5412 Arbitrary Waveform Generator and reading them with a PXIe-5122 Oscilloscope. I am then streaming the waveform to a PXIe-7846R FPGA module with Peer-to-Peer Streaming and sending the waveform back to the Host through a DMA FIFO. However, for some reason, after going through the FPGA, it takes about 39,000,000 elements to create the entire waveform whereas the waveform when graphing directly from the oscilloscope is 1,048,576 (the correct amount). Does anyone know what is causing this?

[GerdW]

Hi Trekkie,

---

@Trekkie123 wrote:

However, for some reason, after going through the FPGA, it takes about 39,000,000 elements to create the entire waveform whereas the waveform when graphing directly from the oscilloscope is 1,048,576 (the correct amount). Does anyone know what is causing this?

---

See your own FPGA VI:

(Btw. using AutoCleanup could help here.)

The FPGA is set to read the P2P FIFO as fast as possible (in a SCTL running at 40MHz) without any timeout.

So even when the FIFO is empty it will read a value - most probably the default value "0"…