LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Error -52000 undefined error

I'm getting error -52000 in a loop where the only thing I'm doing is reading from an FPGA FIFO (invoke method node). The description says this is an undefined error. Anyone have any idea what it could be?

 

I don't see the error listed here

0 Kudos
Message 1 of 5
(4,098 Views)

Looks like somehow I'm overflowing my FIFO (elements remaining keeps jumping from 0 to 18000 then back to 0), sending everything out to lunch. I'm guessing the fix is something in my software logic, but I'm pretty sure the error relates to overflowing.

0 Kudos
Message 2 of 5
(4,092 Views)

for(imstuck):

 

You're definitely right about the overflowing part (see screen shot below). How big is the chunk of data you're sending to the FIFO? Would you be willing to post any screen shots of the code?

 

My guess is we'll need to increase loop rates and decrease the chunk size that we're writing to the FIFO.

 

error 52000.png

Caleb Harris

National Instruments | http://www.ni.com/support
0 Kudos
Message 3 of 5
(4,068 Views)

it's fixed point data and I am putting 14,336 elements (7 cylinders * 512 samples * 4 channels/cylinder) in the FIFO that I am grabbing. The problem is, I can't put less than that because I am taking data from an engine with multiple cylinders which could each be offset. Those offsets may change. So, I don't see another way to know in what order the data is being put in my DMA FIFO. For example, let's say cylinder 2 is offset from cylinder 1 by 10 encoder ticks on one engine and cylinder 3 by 20 ticks. Now on another engine cylinder 2 is offset from cylinder 1 by 20 ticks and cylinder 3 by 10 ticks. That means data will be put in my FIFO in a different order depending on the engine. I then have no idea how to decimate and process it at the other end. Therefore, I'm waiting until I know I have all the data, then queuing it up in a specified order to be decimated. My FPGA code is posted here. On the RT side, I have posted a screenshot of the loop in which I am getting an error and deqeueing elements. The fixed point data being read is signed 20 bits word length, 5 bit integer length, which is the default data type returned from the cRIO modules. I believe this is a LOT of decimal places that I don't need so suggestions appreciated.

 

 

0 Kudos
Message 4 of 5
(4,029 Views)

Could you upload a simple project that reproduces this issue?

If you can figure out a way to use smaller FIFO's it might solve it but just in case I would like to reproduce the issue to pass it along to R&D.

Ben J.
National Instruments
Applications Engineer
0 Kudos
Message 5 of 5
(3,996 Views)