|
| View previous topic :: View next topic |
| Author |
Message |
hraftery
Joined: 14 Apr 2008 Posts: 6
|
Posted: Thu Jun 05, 2008 10:46 pm Post subject: Discontinuous buffers of data |
|
|
I've been trying hard to track down an issue with a few of our data collection units. They use a ChicoPlus with an SD16 to sample data. The software side uses a Caliente component, married to a SD16, Clock and ChicoPlus components. A GenericFilter then accepts input data from the ADC on the SD16. We use the OnFilterData event of those GenericFilter's to memcpy the data into our own buffers.
Recently we've begun to see discontinuous data appearing in the InBuffer of a OnFilterData event. When collecting a 150Hz sine wave, the data will sometimes contain a few discontinuous events - the sine wave appears to shift phase immediately.
This occurs regardless of sample rate, processor loading, memory usage, ReserveMemDSP settings and BufferPeriod, but only occurs when the card is installed in certain machines. While I go bananas trying to figure out where the fault could possibly be, can anyone suggest how the InBuffer data could have these glitches?
The SD16->Overrun flag never appears to assert, but I suspect there is a problem in the processing we do in the OnFilterData, since the faults generally appear near the time where we actually start incrementing our own buffers. |
|
| Back to top |
|
 |
gmiller
Joined: 07 Mar 2006 Posts: 105 Location: Simi Valley, CA.
|
Posted: Fri Jun 06, 2008 3:15 pm Post subject: Discontinuous buffers of data |
|
|
| The majority of the time this is the lack of enought buffering for the driver. With these faster computers now your older code should finish faster. Make sure your reserve mem is set up to meet the needs of the driver. |
|
| Back to top |
|
 |
hraftery
Joined: 14 Apr 2008 Posts: 6
|
Posted: Mon Jun 09, 2008 4:50 pm Post subject: |
|
|
Well there you go, fiddling with ReserveMemDsp values appears to have kicked it into life. Here's what I found (note this is on an old 600Mhz machine with 128MB of RAM):
BM: <= 2048KB
Rsv: <= 4096KB
Result: Glitches as reported in first post.
BM: 16348KB
Rsv: 32768KB
Result: Hanging. That is, no feedback from hardware.
BM: 1024KB
Rsv: 16348KB
Result: Glitches as reported in first post.
BM: 32768KB
Rsv: 1024KB
Result: Working!
Historically we've never given the particular values much consideration - we generally accept the defaults or set it for about 2048KB and 4096KB. Now (several years after this machine first went into service), as reported above, the particular values are critical!
Are there any guidelines for how we should set these values?
Also, there are two entries for ChicoPlus in the "Type" drop down. It doesn't seem to matter which one is selected. Does this sound right? |
|
| Back to top |
|
 |
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You cannot download files in this forum
|
|
© Copyright 2006-2012 Innovative Integration
Powered by phpBB © 2001, 2002 phpBB Group
Based on iCGstation v1.0 Template By Ray © 2003, 2004 iOptional
|
|
|