The MID-7604 puts out a lot of noise, where can I get isolated cables for it?

[bravedonxiote]

I'm using a MID-7604 stepper motor driver. I'd like to reduce the ground loop noise in the system that I am measuring on my PXI-5152 (through a series of transducers, etc.) by using an optical or transformer isolator for the 68 pin VHDCI cable that connects the motor driver to the computer. Does such a thing exist or are there other methods to mitigate the noise? I'd prefer to have the MID-7604 and the computer connected to the same circuit/I don't really want to buy a gigantic isolation transformer/surge protector just to plug in the MID-7604.

[ChrisStrykesAgain]

bravedonxiote,

Could you give me a better idea of how your system is wired? What DAQ devices are you using and what kind of connectors/termination? How are your cables run?

If you could post pictures of the configuration in question it might be easiest.

[bravedonxiote]

OlliesCamel,

The devices I'm using are PXI-5152 for signal acquisition, TEGAM 4040B for amplification, MID-7604 to drive stepper motors (which have specifications shown here http://www.sherline.com/stepspec2.pdf) to turn a Sherline rotational stage. The transducers are both Harisonic I30508T 5 MHz 3.5" focus transducers.

I have scanned in a drawing I made up of how everything is physically connected entitled Electrical connections.PDF. Also, the noise I am referring to is the noise that is shown in picture noiseMID7604.jpg. (It is labeled)

For the initial laser noise I'm getting an optically isolated RS232 which should fix that problem. On the MID-7604/7602 stepper power motor drive user guide and specifications it shows on page 13 that a wire should go from the 3 position to the case for grounding. It also shows that it seems to be attached to the shield of the shielded wire. Is this necessary as long as I have the ground connected to the case of the motor or? The way I have it now I just have a separate wire in the bundled/shielded cable that is going to the case but it isn't connected to the shielding itself.

The settings for the amplifier and the PXI are for 1 M ohm impedance.

As you can see, my signals don't have a very large SNR, so any sort of motor noise is problematic, particularly since the noise I've shown doesn't stay in the same place during the test. It tends to move around/be on top of my signals at random times even though it doesn't show that in this particular picture. It is problematic since I'm taking many measurements (20,480 acoustic signals per scan, etc.).

I've tried averaging (256 times) and filtering (10 MHz low pass filter is the lowest I can go), however, the noise still persists a bit since it happens to somewhat fall under the same frequencies, etc. The averaging does help, but I'd like to run the test faster and so eliminating that noise would help me in that regard.

[ChrisStrykesAgain]

bravedonxiote,

Just to be clear, one end of the shield shoudld be connected to ground, but not both. This is illustrated and explained in the capacitive coupling section of the following link: http://www.ni.com/white-paper/3344/en/

If this is the way you already have it configured there are many more suggestions in the white paper linked above that you may find useful. If you need further clarification on anything, let me know.

[bravedonxiote]

OlliesCamel,

Okay, so I should connect the shield on one end to ground, etc. I have a true "floating source" (transducer generating voltage via pressure stimulation) so I have to connect the shield to the ground on the motor driver side, not on the signal generation side like it suggests, right?

Also, I noticed it said to ground any close metal to reduce capacitive coupling, so I'm assuming I should ground the optical table as well?

Does it make sense for me to ground the optical table and motor line shield to the PXI which is grounded to the wall outlet? I think this would be convenient given the easy access to the ground screw, etc.

Or should I ground them directly to the wall outlet, etc.? I'm just wondering what would be easiest, etc.

[ChrisStrykesAgain · Accepted solution]

bravedonxiote,

Connecting to the PXI ground should be fine, provided everything is referenced to the earth ground.

And yes, grounding the table is a good idea as well.

[bravedonxiote]

OlliesCamel,

Great, now the noise is much lower! Thanks a lot!

[ChrisStrykesAgain]

bravedonxiote

I'm glad the solution worked for you; that particular white-paper is worth its bits in gold.