compressor vibration & noise analysis

[JK1]

Hai Johnsold,

I read one of your posts how you carried out the task of vib anal.  Here my task is also the similar one to build setup for detecting a noisy compressor.  We are bound to assign some standards or setpoints to reject the comp.  the rig is admidst of highly noisy production line.

You have said abt ground loop.  Give me ideas how to look for the ground loop and also please share ideas regarding this task.

sensor: accelerometer, DSA 4472a, RT 8145.  I have SVToolkit.

I have been monitoring the spectrum and there are specific regions but i couldnt hopefullly select one band or i dont have enough knowledge to justify

help  me

with regards,

J Karthikeyan

[johnsold]

Hi,

The protocol on this Forum is to continue to reply to the original post rather than to start a new thread each time. It makes it easier for anyone who is trying to follow the thread to keep track of all posts.

Ground loops occur when multiple parallel paths to ground (or earth, depending upon language convention in different parts of the world) exist between different parts of a measurement system. In your system the compressor being tested has connections to power and ground (earth). The data acquisition device and the computer also have power and ground connections. In addition the accelerometer case is probably connected to the compressor case. But the accelerometer case is also connected to the shield of the cable connecting it to the DAQ device and then to the computer.

If leakage currents flow from the compressor power to ground (quite common in compressors), that current will divide betweeen the path through the equipment grounding conductor (required by safety codes) and the accelerometer shield connection to the computer. The division of current will be in accord with Ohm's law and the ratio of the impedances along the two paths. If the currents are small (~mA), the only effect will probably be some addtional power line frequency noise or interference on the data. In case of a ground fault in the compressor the currents may be 10s to 100s of amperes and destruction of the accelerometer, DAQ device, and computer could result.

Ideally a measurement system would have a single ground reference point and all devices would be connected to it by separate wires, called star grounding. However, due to electrical safety codes, this is usually not possible. Thus, various types of isolation are used. A nonconductive housing for the accelerometer is one means. Another could be transformer or optical isolation of the sensor signal from the computer.

It is beyond what can be handled in this Forum to completely cover this type of situation. A good test or safety engineer should be able to advise you on the specifics of your case.

Lynn