- From: John Conner <conner_at_empiredi.com>
- Date: Sat, 01 Jun 2002 16:31:12 -0600
Looking for ideas..... I have a system collecting data at 1000 scans per second (16000 samples/sec) using a PC104-DAS16jr/16. It has been working great since March. (last check on uptime showed 42 days). I'm trying to get the same set of programs to run on a different computer with a CIO-DAS16jr. The system even has a close clone of the disk on the other system. This system does not work. I've changed the comedi_config command to reflect the change in the board. If I input say a 30Hz sine wave on channel 0 and ground the remaining 15 inputs, I do not get the 30Hz signal on channel 0 with the CIO-DAS16jr. The signal is scattered over at least 3 of the 15 other channels. If I put a PC104-DAS16jr/16 on an ISA adapter and do the same test with the same cables things are fine. If I try a CIO-DAS16jr/16, the collection hangs during the setup. ( This board only has a 10MHz master clock going to the 8254. The Clock parameter is required wiht comedi_config.) This looks a lot like a problem I had with the PC104-DAS16jr/16 years ago. With a long cable (5+ feet) the sample clock (cntr 2 out) would pick up enough ringing to false trigger the MUX counter. The quick fix was to cut the wire at the board end of the cable. ComputerBoards/MeasurementComputing claims that it was fixed in an later revision of the board. The cables I'm using have this line (pin 20 on the DB37) cut. So, anyone have any ideas as to why the boards and programs don't want to work together? John Conner Empire Digital Instruments
Received on 2002-06-01Z21:31:12