Created Date: January 28,2009

Created by: milldrone

4. From thread 5630 ( http://forum.automationdirect.com/showthread.php?t=5630 ), perhaps fit a drive isolating transformer, with the neutral tapping earthed to yet another ground rod. How far should ground rods be apart?

Hugh

The drive isolation transformer thing is only valid for noise reduction if your incoming power is a floating delta. Or a "wild leg " delta

The isolation was achieved by using short non-conductive hydraulic hoses, but this has too much observer effect at higher frequencies so I am now changing to isolating the pulse generator from the electric motor.

I'm curious. Can you electrically isolate the pump from the motor with an elastomeric coupling and some UHMW sandwiched in the pump mount? Also do not forget about the bolts. Your comment about the non conductive hose leads me to believe the physical electrical link to the motor is the culprit.

Created Date: January 28,2009

Created by: HughBK

Thanks Vaughn,

I am presently implementing the isolation with elastomeric coupling, gasket and bolt insulators, so am on the same path.

Being a simple mechanical person I do not feel the difference between ground and neutral in my bones. I have no neutral connection to the inverter, just 3 power and ground. The ground goes to a ground rod just outside the building, the neutral from the power company, so I guess that this could be a reason for a grounded neutral via a drive isolation transformer.

Hugh

Created Date: January 28,2009

Created by: HughBK

I chased earth loops.

1. Made sure that the transducer earth was only connected at the DAQ end, which did not make any difference. Cannot do this with the piezo as it is only 2 wires.

2. Ran the DAQ and power supply from a 12V battery, unfortunately had to be with a DC/AC inverter for the DAQ. Again no difference.

Thanks, Hugh

Created Date: January 29,2009

Created by: milldrone

Thanks Vaughn,

I am presently implementing the isolation with elastomeric coupling, gasket and bolt insulators, so am on the same path.

Being a simple mechanical person I do not feel the difference between ground and neutral in my bones. I have no neutral connection to the inverter, just 3 power and ground. The ground goes to a ground rod just outside the building, the neutral from the power company, so I guess that this could be a reason for a grounded neutral via a drive isolation transformer.

Hugh

Good you are not supposed to feel the voltage in your bones;);)

I'm attaching a drawing of a floating delta transformer and a grounded wye.

The grounded Wye is what a drive isolation transformer does. Can you see the center point? A VFD makes a lot of noise and that noise wants to go somewhere. In a grounded Wye the noise wants to go to the center point of the transformer, this helps to keep it away from your electronics. If you have a floating delta power supply the noise cannot find an easy route to the transformer, and sometimes your electronics provide a path for the noise to go back "home ".

You had mentioned earlier that the problem went away when you used an insulated hose to connect your system. This is what lead me to think that the noise was traveling in the steel braided hose to your sensor.

I had an application that was suffering from noise generated from VFD's. The manufacturer's grounding scheme was bad they grounded everything but the motors to a "dirty " (building) ground. Then the motors were grounded to the electronics clean ground. As far as I was concerned this was wrong so I insulated the motors from the electronics and connected them to the building ground. After this change the VFD noise was gone. The reason they had done this was because the motors were inside an enclosure the same one the electronics were in. It was a lot of extra work to use the building ground for the motors.

Created Date: January 29,2009

Created by: HughBK

Thanks for that explanation. BTW, the isolation provides a good answer for transducers with low frequency response and low impedance, but I am looking to go a step further for the high frequency and high impedance piezo transducer I am starting to use.

At this stage I plan to put in a separate instrument ground and fully isolate the drive and instrument sides. Then look at the drive isolation transformer.

In your opinion would EMI filter and line reactor be just a way of spending money?

Hugh

Created Date: January 29,2009

Created by: milldrone

Then look at the drive isolation transformer.

Hugh,

I'm only suggesting to use the drive isolation transformer if you have a floating delta power supply. If you have the grounded Wye then I would not use a drive isolation transformer.

In your opinion would EMI filter and line reactor be just a way of spending money?

Hugh

Sigh. Chasing VFD noise is very situation specific. What works in one application does not seem to work in the next application. Sometimes the end user can do everything "wrong " and not have any issues. Sometimes it's just a crap shoot.

My advice is this. If you are an end user do the things that do not cost any money first. Then ratchet up from there.

If you are an systems integrator or an installer do the things that you will feel comfortable with so you do not have any call backs.

You mentioned that you were a mechanical guy, (so am I) and you did not seem to balk at insulating the pump from the motor. So I'm presuming that was a not too difficult thing for you to do.

By the way ADC App Assist is correct, it is a grounding issue. But sometimes it is easier to move the mountain than to move Mohammad.

Created Date: February 02,2009

Created by: HughBK

Vaughn,

Thanks for the sigh. One always hopes that one of the electronic gee-whizzers could come up with a magic bullet, but I did realise it was wishful thinking.

I have a week or so wait while the isolation plastic bits are made up, and I will instal a separate instrument earth in the meantime. See what happens next.

I believe that I have grounded WYE at the powerboard, but then only delta connection and earth to the VSD, so that the VSD neutral is effectively connected to the instrument ground.

I am a lone explorer, so I just need to see how well I can make it work. I am certainly, to mix a few metaphors, trying to find knowledge in the jungle of noise, with very low signal levels. The time to develop the test gear is always a frustration.

Hugh

Created Date: February 03,2009

Created by: gkaufman

You might try using an RF filter on the output of the drive. Automationdirect sells one with part number RF220X00A.

You could also try putting the RF filter on the output signal of your pressure transducer and even the supply input for the transducer - wind the output wires through the RF window as many times as you can.

Good luck.

Dimensions (In/mm) 3.54/90 L x 0.98/25 W x 2.70/68.5 H

Zero phase reactors, (aka RF noise filters) help reduce radiated noise from the inverter wiring. The wiring must go through the opening to reduce the RF component of the electrical noise. Loop the wires three times (four turns) to attain the full RF filtering effect. For larger wire sizes, place multiple zero-phase reactors (up to four) side by side for a greater filtering effect. These are effective for noise reduction on both the input and output sides of the inverter. Attenuation quality is good in a wide range from AM band to 10Mhz.

The reactor must be put at inverter side as closely as possible.

Created Date: February 25,2009

Created by: HughBK

Thanks for the help.

I have at this stage two ground rods, in the ground about 20ft apart.

The first is connected to the incoming 3 phase neutral, the VFD ground, and the drive electric motor ground.

The test machine, which is effectively a rotary hydraulic valve, is electrically isolated from the drive motor by plastic mounting bits and an insulating drive coupling. The test machine is earthed to ground rod #2 but only through the instrument screening.

The hydraulic power pack that provides the pressure and flow to the test machine is connected to the test machine with hydraulic hoses that are conductive, so the electric motor on the power pack is grounded to #2. This motor is 3-phase, driven from the same supply as the VFD.

The pressure transducers are direct mounted onto the test machine. The DAQ is powered as single phase from the same supply as the VSD, but is earthed to #2.

This provides a good enough result at this stage with some increase in noise as the VFD loads up. It seems to me that the weak link is that the DAQ power is contaminated, both active and neutral, particularly the neutral as this is connected to VSD ground. So the next stage would be to use RF filters.

These could be on the incoming side of the VFD, or on the active and neutral to the DAQ. All my wires are short so that I do not think that radio pick-up is a problem, so I favour the idea of filtering the DAQ supply. Comments welcome.

Is not life a learning experience?

Best, Hugh