 DC89
		
			DC89
		
		
		
		
		
		
		
		
	
			12-03-2012 07:24 PM
Hi,
I am currently using PCI-7350 with the latest firmware to drive a Faulhaber 2237 series motor coupled with a GSI 6000 seires encoder. I have no problem when moving or finding the index when the move is less than 1 revolution (<= 359deg), but anything more than one revolution would throw off the accuracy ( I start seeing trending in position error). I have looked through the knowledge base and haven't found anything relevant. Any feedback as to where to find more information on this or ideas to resolve this would be greatly appreciated.
Regards,
 Highland_Contro
		
			Highland_Contro12-04-2012 05:24 AM
What is the specific model of your encoder? There are speed limitations published for that encoder, and they can be quite low depending on your resolution. There are also limitation on the 7350 concerning encoder encoder input speed. Do you lose counts no matter what speed you run at? Are you sure that it is wired correctly? It almost sounds like you have the index pulse wired to the A or B channel.
12-04-2012 03:59 PM
Thanks for the quick response. I believe the encoder is working as it should since finding the index from any position less than 360deg is very accurate (tested over hundreds of cycles). Also, it definitely reached the desired velocity between these moves so speed doesn't seem to be a concern. To sum it up, the move becomes inaccurate when it is moved in one direction continuously (T>360deg). Is it possible that the board looses some counts when moving in a continuous direction to find the index?
 Highland_Contro
		
			Highland_Contro12-04-2012 09:38 PM
What are the number of counts per rev, and the rate of counts coming in. Yes, the index is always at the same spot, so I would expect it to find it accurately. Are you losing counts in one direction only, or does it happen in both directions? After homing, can you make repeatable moves of less than one rev? Set up a back and forth move, and see(physically) if the positions are changing.