Help.....please !

[dvmike]

I have been getting communication errors intermittenly for about a year. Most times it starts back up after a reboot and I don't have the problem for a month.
I am now getting these communication errors excessively . I have had 5 this morning !
I have replaced the USB cable, and put the filter from the old one on it . I have changed USB ports . I have wiggled the cables . I have checked the connection in the control box where it interfaces with the communication board.
I have found one thing . It seems like the machine makes a bump sound when this happens.
Could it be that on a communication error the machine is instantly losing and finding communication, and trying to make up for the "free wheel" distance, causing the bump sound.
I have an Acer laptop running Vista that has 1 gig memory that I can hook up if it sounds like the communication error / Lag might be coming from the computer.
Any suggestions?

[bleeth]

That Bump sound could be static electricity discharge and that will blow com. Check your grounding, not only from your table, and gantry (s) but also dust collection.

[dvmike]

Thanks Dave . I'm working on adding extra grounds now. If you guys have other ideas let me know. This is our busy time, and I need this machine bad !
Again thanks !

[dhunt]

Seen this thread?

http://www.talkshopbot.com/forum/showthread.php?t=13184

.

[bleeth]

Mike:

Make sure your grounding is in series or parallel and you do not create a "loop"

[Gary Campbell]

Mike....
I would go with 90%+ on the static as cause for the "bump" The only other cause I have experienced was an inductance spike caused by current collapse on a solenoid valve that had a bad shunt diode.

Can you make it happen when the machine is not moving? Does it happen more or less when the dust collector is on?

Take a long (25') stranded wire from the case of the control box (alpha or RBK) and with SB3 running, the machine on, but still, and touch it to any/all metal items. see if any touches return any error messages.

I have done, but dont recommend, similar tests during controlled moves with the dust collector running. Its dangerous to get tangled in loose wires.

I am going with conventional logic and say that a majority of comm errors are due to the control computer. Not that something is wrong, just that everything is not right for CNC use. Make absolutely sure that no backround processes are running. Windows is full of them. Try a replacement computer first.

That said, a good number of folks, myself included have reduced the number of these and other errors by a comprehensive ground/bonding system. The powder coating on the steel and the anodic coating on the aluminum are both insulators. I am sure that you can get continuity thru a bolted connection, but dont think an Elect. Eng. would guarantee it without additional steps taken. The balance of the connections are bearings, yet another insulator with possible incedental contact. Bearings could possibly generate static. If your testing shows inconsistant grounds, then you need to bridge each connection with a low resistance bypass. More here: http://shopbotwiki.com/index.php?title=Grounding_your_ShopBot

[dhunt]

I am going with conventional logic and say that a majority of comm errors are due to the control computer.
Not that something is wrong, just that everything is not right for CNC use.
Make absolutely sure that no background processes are running. Windows is full of them.So many people overlook this vital part of the whole equation..