Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Barry Singleton]

  Hi All;

  I have a real problem with AC line noise in our building. It is severe enough that I can hear the high frequency buzz in all the transformers in the building and the service transformer outside. I can even hear it in the power supplies of the PC's in the office. The worst part is that the noise is caused by us. We have two 55 horse power direct drive spindle CNC lathes that are inverter controlled. On acceleration one can observe some bending of the AC waveform but on braking they generate the noise that sounds like arcing in all the transformers. It is severe enough that the UPS/AVR's jump off line for a second or two.

  These two machines are on their own feed from the service transformer with their own meter running on 480V. The rest of the building has one 480V panel and three 120/240V panels each with their own transformer. None of them are cascaded and the noise permeates the entire system.

  The power company has been out several times and logged data, scratched their heads, brought out engineers and other white coat types and in the end told us that there is nothing really wrong and it just may have to run to failure if there is something wrong that they are not seeing. Fabulous!

  I was in the panel that feeds the offices to see if there was any breaker space to pull two or three 20A circuits to an isolatioin transformer to see if we can get a handle on this. I had done nothing other than removed the cover and was summing up what was available when sparks shot out of the box! Upon inspection the neutral connection and buss were loose and arcing when the machines brake so we have some panel maintanance to do this weekend.
 
  I have done a brief study on isolated transformers and it seem like the true double shielded units are the answer to the problem, but I sure don't want to buy one if the HF attenuation will not be adequate as they are quite expensive.

  I have a couple of oscope screen shots to show. I'm using a 1000X probe so the mV scale is directly scalable and the nasty spikes are about 22V in magnitude with at 360Hz.

  So, if anyone has experience with eradicating this type of noise, please help!

  Oh, one thing thats interesting, there is zero audible noise in my office stereo! I guess that's why I have lived with this for nine months.

  Thank you and all the best,
                                          Barry.
 

  I am having trouble with the pics, looks like I need to resize them. If they don't show up Ill try again.

[Josh Millward]

We have two 55 horse power direct drive spindle CNC lathes that are inverter controlled. On acceleration one can observe some bending of the AC waveform but on braking they generate the noise that sounds like arcing in all the transformers. It is severe enough that the UPS/AVR's jump off line for a second or two.

Hi Barry,

What does the manufacturer of those CNC lathes have to say about their machines spewing all that trash?

I would ask them first because if it is that bad, surely they have had issues with other customers.

I would try to fix the problem at the source rather than try to mitigate the symptoms of the problem. So, what could be done to keep those machines from creating all that noise? Perhaps a different type of PFC front end for those inverters, or maybe a different motor control system?

I don't really know, I'm not an expert at those things, but I think trying to stop it at the source rather than trying to protect everything else is a better plan of attack.

Good luck, have fun, be safe, and be sure to let us know what happens.

Thanks!
Josh

[TJ (Tom) Cornish]

Hi Barry,

What does the manufacturer of those CNC lathes have to say about their machines spewing all that trash?

+1 MFR should help.  Some VFDs have optional input filters to mitigate line noise.  There may be braking loads available too.

It does seem that a loose neutral wire in the panel could contribute significantly to the problem - the very first wireless transmitters used sparks to generate the RF - arcing is a great broadband noise source.

[Barry Singleton]

  Yeah I know about spark RFI, like fire, pretty cool when its your friend.

 It's pretty funny that The Mazak guys just shrug it off too. Mazak also sent a white coat electrical guy up from LA. He fixed an internal issue that was shutting the CPU down randomly but was no help on the noise issue. The machine has three internal 1:1 transformers for noise control but their attenuation seems inadequate if the real issue is not a connection issue between the service transformer and the machines.

  Form a dollar and cents point of view since I am pretty much on my own with the power company and the manufacturer having done all that they are willing to do and a 7.5KW isolation transformer for the office looks more affordable that a transformer sized to run +180HP of lathes.  They each have 20 HP for live tooling plus high pressure coolant and oil pumps if you wondering where the rest came from.

  It will be interesting to see what panel maintanance gets me but the buzzing in the monster box outside that feeds the whole building tells me it doesn't end there.

  Thanks again,
                     Barry.
 

[TJ (Tom) Cornish]

  Yeah I know about spark RFI, like fire, pretty cool when its your friend.

 It's pretty funny that The Mazak guys just shrug it off too. Mazak also sent a white coat electrical guy up from LA. He fixed an internal issue that was shutting the CPU down randomly but was no help on the noise issue. The machine has three internal 1:1 transformers for noise control but their attenuation seems inadequate if the real issue is not a connection issue between the service transformer and the machines.

  Form a dollar and cents point of view since I am pretty much on my own with the power company and the manufacturer having done all that they are willing to do and a 7.5KW isolation transformer for the office looks more affordable that a transformer sized to run +180HP of lathes.  They each have 20 HP for live tooling plus high pressure coolant and oil pumps if you wondering where the rest came from.

  It will be interesting to see what panel maintanance gets me but the buzzing in the monster box outside that feeds the whole building tells me it doesn't end there.

  Thanks again,
                     Barry.
 

How about in-line inductors?

[Greg_Cameron]

Form a dollar and cents point of view since I am pretty much on my own with the power company and the manufacturer having done all that they are willing to do and a 7.5KW isolation transformer for the office looks more affordable that a transformer sized to run +180HP of lathes.

Either way, it does indeed sound like machines and the rest of the office needs to be on their own separate services. I vote for secondary transformer, both feed from the incoming utility service directly and not piggybacked.

[Mike Sokol]

Either way, it does indeed sound like machines and the rest of the office needs to be on their own separate services. I vote for secondary transformer, both feed from the incoming utility service directly and not piggybacked.

Considering that you had sparks from the neutral, realize that a high-resistance neutral can inject all sorts of noise into a system. And there's even more problems if the safety ground and neutrals are double-bonded in the sub-panels which allows neutral currents to flow in the grounds. Time to check exactly where and how the incoming Neutral and Grounds are bonded.

[Stephen Swaffer]

I agree with Mike check neutrals and bonding including the grounding electrode system at the service.  I have worked on a lot of these machines and in similar situations.  The machines will be 480 VAC delta -no neutral connection at all.  The office panel coming from a 480/240VAC transformer is most likely fed 480 VAC delta-no neutral, secondary is 3 phase 4 wire Y creating its own neutral which should be bonded to ground.  The ONLY common connection for a neutral (if it exists) on the 480 VAC system and a 240 or 208 system is the grounding bond.  For the noise to travel over the neutral it must travel over/through the grounding system-which is what bonding and a grounding electrode is supposed to try and prevent.  360 HZ isn't really a high enough frequency to start misbehaving too badly in seeking a ground-if it has a good solid path somewhere.

Otherwise, a good electrical engineer should be able to build a low pass filter to supply the machines-allowing 60 HZ to pass and attenuating noise above that.

[Mike Sokol]

The machines will be 480 VAC delta -no neutral connection at all.  The office panel coming from a 480/240VAC transformer is most likely fed 480 VAC delta-no neutral, secondary is 3 phase 4 wire Y creating its own neutral which should be bonded to ground. 

Another thing to check will be a high-leg delta transformer feeding the panel. That was commonly used in industrial buildings that need a lot of 240-volt 3-phase power for motors, and much less 120-volt single phase for office equipment. Of course a 480 high-leg delta would normally only split down to 240-volts single phase (if memory serves), but there were a lot of strange things done with 1970's power that we would never consider using modern code. Maybe there's a way they did high-leg delta 480v to single phase 120v. I'll think about this a bit, but a simple neutral to leg voltage measurement would confirm a high-leg delta.

[Barry Singleton]

  Hi All;
 
  Tomorrow when I shut down and do the rest of the panel maintenance I will attempt to acertain where and how the ground and neautral are bonded and where they are not. I will have a licensed jounyman electrician with me for the day.

  In answer to a couple of questions/suggestions:

  The buiding was built in 1996 and there is no high leg. 

  The huge transformer that feed this block of buildings is about 200 feet across the street from the distribution/meter panel on the building. The power that feeds my building comes in underground from the dist/meter panel to a 480V panel that in turn powers some equipment like heaters , some air handling equipment ,some smaller machines and three step down transformers for the rest of my building. There was inadequate power in the building for the two new machines so, there was another 480V line of service pulled in from the distribution/meter panel straight down the wall to a new 480V panel just for these two machines. This service has it's own meter and nothing else is on it at all.

  I don't recall for sure how many lines come from the main transformer across the street to the meter panel here on the building but I think there were six. How that is split up in the dist/meter panel I don't know.

  What I think is most interesting is that I can also hear the sound of (sounds like arcing) in the main transformer across the street when the machines break and to me that's crazy!

  Thanks for taking the time to reply with ideas and help.

  All the best,
                  Barry.

[Stephen Swaffer]

Also, keep in mind that if you decide to filter on the machine end, there is no reason to buy a filter big enough for everything.  It should be fairly simple to install a filter just ahead of the spindle drives.

[Mike Sokol]

What I think is most interesting is that I can also hear the sound of (sounds like arcing) in the main transformer across the street when the machines break and to me that's crazy!

That arcing noise could be the hint. I've had a massive main transformer feeding a manufacturing plant lose a neutral connection internally. A lug had a loose bolt and created the problem. They said it was probably arcing and heating up for weeks, so we used an infrared camera to check all the other connections feeding the sub-panels. These was pretty big stuff with 600-amp 3-phase breakers feeding the sub-panels, and lots of damage happened from the open neutral.

It would be really interesting to have your journeyman electrician or power company pull the access panel from the big transformer and watch it at night while your machines break. However, be sure you use Arc Flash PPE gear (Personal Protective Equipment) since if a high current connection lets go while you're in blast zone, there will be an explosion of molten copper in a plasma hotter than the surface of the sun coming at you faster than the speed of sound. I've witness a few of those Arc Flash incidents, one up close and personal. And it's way more dangerous than simply being shocked.

[Jonathan Johnson]

I was in the panel that feeds the offices to see if there was any breaker space to pull two or three 20A circuits to an isolatioin transformer to see if we can get a handle on this. I had done nothing other than removed the cover and was summing up what was available when sparks shot out of the box! Upon inspection the neutral connection and buss were loose and arcing when the machines brake so we have some panel maintanance to do this weekend.

A good reminder to wear PPE (especially safety glasses) when opening electrical panels, even if you're not going to be working on them.

Even if the panel is dead, safety glasses are a good idea. You never know when something's going to let fly... a wire with spring pressure that releases unexpectedly... a bit of copper when you're snipping a wire... a bubble in the bottle of anti-corrosion goop that makes it splatter.

[Jerome Malsack]

It might be worth looking at the computer network connecting to the milling if there is one and use fiber to connect the computer back to the main frames. 
also if possible us a fiber cable to connect the computer to the machine to prevent any transfer of the problem to the office level of power.   

[Tom Bourke]

That arcing noise could be the hint. I've had a massive main transformer feeding a manufacturing plant lose a neutral connection internally. A lug had a loose bolt and created the problem. They said it was probably arcing and heating up for weeks, so we used an infrared camera to check all the other connections feeding the sub-panels. These was pretty big stuff with 600-amp 3-phase breakers feeding the sub-panels, and lots of damage happened from the open neutral.

It would be really interesting to have your journeyman electrician or power company pull the access panel from the big transformer and watch it at night while your machines break. However, be sure you use Arc Flash PPE gear (Personal Protective Equipment) since if a high current connection lets go while you're in blast zone, there will be an explosion of molten copper in a plasma hotter than the surface of the sun coming at you faster than the speed of sound. I've witness a few of those Arc Flash incidents, one up close and personal. And it's way more dangerous than simply being shocked.

I also agree that you should have every panel and transformer inspected and re torqued.  One loose connection indicates there may be more.  Also those outdoor transformers are subjected to massive temperature swings out here.  Also another forum that may be able to help you is http://www.cnczone.com/forums/mazak_mitsubishi_mazatrol/  As far as I can tell that site is the prosoundweb of the machine world.

[ProSoundWeb Community]