0-10V shielding

[RocketTester]

... This is a high-speed oscilloscope discussion that is of little interest for PLC work. If you're concerned about speeds like that then a PLC is not the right choice for what you are doing.

Depends on what "high speed" means to you. In 2005, I selected Beckhoff Automation's EtherCAT PLC hardware to sequence liquid rocket engine tests because it could kill a test in <1 msec. My boss wanted A-B, only because he had heard of it (no hands-on), but best we could manage with it was >10 msec, using special modules. B&R Automation was the only other PLC choice I found and ~2 msec then. Later, NI promoted their (expensive) CompactRIO, claiming in ads "very fast". I asked what that meant in an NI seminar, and found "4 chan at 1 msec sample rate, or 10 msec for >4 chan", which "wasn't fast" for our needs, but NI guy said "you could do faster with custom FPGA code", which I wondered then why didn't the NI experts do so.

 

[NetNathan]

I want to thank you members from answering this question so diligently.
I will be changing the wiring to support shielded twisted to remove all questions.

 

[busarider29]

It may be that the regulation needed for a constant-current source introduces a slight delay for a rapidly changing analog that would not be present for a constant voltage source.

That's where I was attempting to go with it. Thanks for also setting me straight!

This is a high-speed oscilloscope discussion that is of little interest for PLC work. If you're concerned about speeds like that then a PLC is not the right choice for what you are doing.

"High speed" can be subjective. I mean, if you were to talk to an applications engineer from National Instruments and mention 'high speed' to them, then they are most likely thinking of anything north of 1 Mhz. Which in that case, yes, I agree that an O-scope or DAQ hardware (NI) is better suited for that. But if it's anything <= 200 kHz (5 µS between samples), then that is not outside the realm of what a PLC can do. Of course, it all depends on what PLC platform we're talking about too.

 

[busarider29]

Depends on what "high speed" means to you. In 2005, I selected Beckhoff Automation's EtherCAT PLC hardware to sequence liquid rocket engine tests because it could kill a test in <1 msec. My boss wanted A-B, only because he had heard of it (no hands-on), but best we could manage with it was >10 msec, using special modules. B&R Automation was the only other PLC choice I found and ~2 msec then. Later, NI promoted their (expensive) CompactRIO, claiming in ads "very fast". I asked what that meant in an NI seminar, and found "4 chan at 1 msec sample rate, or 10 msec for >4 chan", which "wasn't fast" for our needs, but NI guy said "you could do faster with custom FPGA code", which I wondered then why didn't the NI experts do so.

NI has a cDAQ chassis now that has support for EtherCAT, which they didn't have back in 2005. They also have an EtherCAT module you can use for the CompactRIO. So, you could do now what you were attempting to do back then with NI hardware right now. With that said, a Beckhoff IPC/CX with high speed measurement I/O over EtherCAT is the way to go.

 

[padees]

I want to thank you members from answering this question so diligently.
I will be changing the wiring to support shielded twisted to remove all questions.

It's not only what was said, unless I missed something. But, voltage signals also have a voltage drop like all others and you need to determine what that could be in your situation. You can only run V signals so far and they are greatly susceptible to noise, induced voltage... It's generally best to run current for analog. Think about 1000 feet max, doesn't pick up noise (for the most part) and no voltage adjustments due to drop. Proper cables are required for both.