Pump communication problems

[ScottHorn]

I'm having bad luck with pumps lately. I appreciate the help I found here for my flow rate issues, a 1000 psi back pressure regulator did the trick. Now on to the current issue. I'm using the same LabAlliance pumps, and the current problem is with a series I+ (just a series I with a pressure transducer added). The pump is connected to the PC through a USB to serial adapter, which is connected to a serial to RS232 adapter, which plugs into the back of the pump. The software package I'm using (Clarity) polls the pump every 250 milliseconds throughout the run to read the flowrate, pressure, and error status. If the software doesn't receive a response from the pump for a preset amount of time (I've set it to the maximum of 10 seconds), then it stops the run (very inconvenient when you're trying to run overnight). I've checked the log of the communication between the pump and pc, and it seems like at random times (sometimes once a week, sometimes multiple times in one day) the pump will not respond to one of the commands. It is a different command each time, either the command to read flow rate, set flow rate, or check error status. The software waits for 10 seconds, and then gives an error message. When I clear the error message and restart the pump, it responds normally. I've tried increasing the time between commands to 500 milliseconds, and even 1 second, and we still get the same error. I've tried replacing the cables between the pump and pc, reinstalling drivers, and even using a different pump (similar model without pressure transducer, same manufacturer). It doesn't seem to be a software problem, in so much as the software appears to send the appropriate commands, the pump just seems to "miss" them sometimes, or just not respond. Has anyone used these pumps before (also sold as "chromtech" and probably a few other brands)? Any ideas? Similar experiences? I'm starting to suspect that the electricity in our lab is not entirely "clean", as they seem to occur more frequently around the time people leave for the day and start turning off other equipment, but it's just a hunch.

[elsico]

I have such problem before with Gilson software and pumps control. The problem is in USB-RS232 converter you use. Try to find better and expensive one to solve problem. Or best way buy computer or notebook with real RS232.

[ScottHorn]

I suspected that might be the case. We have the same pumps running on another pc with the same software. That PC uses a pci card with two serial ports as opposed to a USB converter, and we have many fewer errors with it. I've slowed the command rate down to 1 command per second, and instructed to software to send a string to clear the pump's command buffer before every command. This *seems* to have eliminated the problem *so far*, but I'm still keeping my fingers crossed.