[GITRDUN]

Our shop seems to have severe power problems. Our shop experiences weekly CNC breakdowns and thats no exageration. We have a total of about 20~25 CNC mills and lathes mostly built from 1980 up to around late 90's. At least one or two machines will toast a board almost every week. We have an employee with good electronics skills and he is constanly buried in boards that need repaired due to a blown chip or something.

We have some older Fanuc 6M controlled mills that seem to take the brunt of the breakdowns. Funny thing is the break downs almost always occur when the machine is turned on in the morning. Will be running fine the day before when shut down. But when the machine is powered up the next morning it comes up with an alarm and will have a board blown. My thinking is that the electronics are getting a small spike when the machines main breaker is switched on or when the control is turned on.

Our Okuma lathes experience the same type of break downs. but the lathes usually tear up something in either the spindle drive boards or the axis drive boards. Most of our break downs occur when a machine is powered up in the morning. Its almost rare for one to be running and break down on us.

We experince numerous voltage dips during a normal week, the lights will flash off and back on and a couple machines throughout the shop will shut down. Another business in our area packed up and moved because they believed the city electric was causing their problems.

Are the machines more succeptable to high voltage spikes or low voltage spikes? Will installing surge protectors at each machine help or do we have a more severe problem going on here? What are my options? We cant afford to spend $20k on this, but considering what we spend to keep the shop running it might be a fair trade.

[Geof]

Have you ever checked the incoming voltage at different times of the day, especially in the morning? In many locations the voltage will be high in the early morning and then fall as businesses get into operation and start using power. Your power may be way too high in the morning, above the normal upper limit.

I don't know what you can do to correct it if it does go from too high in the morning to too low later in the day.

I know I would look into getting an electrical consultant to put a data logger on the incoming lines to record voltage, current and phase shift. The idea being to see if I could get data showing the problem was on the supply side and amass the evidence for a potential lawsuit.

[GITRDUN]

The company which left town did exactly that. They hired "experts" to test the incoming voltage and suposedly found drastic problems. Dont think they actually filed a lawsuit as they realized it would be an uphill battle sueing the local authoritys. Im not interested in sueing anybody here, just want to see what we can do to protect our equipment.

So if power is above normal levels in the morning wouldnt it help to add surge protectors at each machine? Thats one option im looking at. But would like more insight before i spend any money. Also the surge protector companys want you to install a protector at each junction box as well as at each machine. I dont get why you need redundent protection, wouldnt simply putting a protector on each machine be enough?

[RICHARD ZASTROW]

You'll have to check with more knowledgeable people than I, but would isolation transformer(s) help with some of this?

Seems to me there was a tolerance of +/- 10% of line voltage allowed. Beyond that, the utility was responsible for damage. I'm NOT an authority on electric power.

Dick Z

[Geof]

Are they suggesting the same size protector for each location. Surge protectors have different energy dissipation capacities. It may be less expensive to put large ones at the junction box and smaller ones at each machine rather than large ones at each machine.

[Geof]

Dick Z's isolation transformer idea can help. I have used two identical transformers back to back as a surge suppressor but in these days of expensive copper that is an expensive solution.

[Al_The_Man]

Normally the power provider has a specialized team that deal with recurring power problems and apply the test equipment in order to trace the problem. I know here where I live in Canada they have a fairly large dept to deal with this.
It is strange that it only occurs at switch on, as usually when you turn the breaker on the controller is normally powered up after this takes place, so it is rare that the whole machine is powered up by the disconnect.
BTW, are the machine disconnects switched off? Or just the controller?
If the power Co. are reluctant to do anything and the majority of the faults occur to the CNC control side, I would measure the 200vac or control voltage supply side on the machines, it would be very simple and cheaper to reduce the 100vac or 200vac as the case maybe, with a small buck boost transformer, I have done this on these types of machines as Fanuc for e.g. allow ±10% on the supply.
Al.

[underthetire]

We have a remote shop in the midwest. Power there is terrible, we ended up putting filtering UPS systems on everything. Very expensive problem to fix. You can get a data logging multi meter that hooks to a laptop really cheap these days, might want to do some monitoring to see what your dealing with first.

[Al_The_Man]

We have some older Fanuc 6M controlled mills that seem to take the brunt of the breakdowns.

One thing I was surprised to find a while back with the Fanuc 6's is that with only the disconnect on, before the machine control was powered, the spindle field circuit was on all the time.
So if the operator shuts the machine down at night by using the control off, the machine has the appearance of no power but the field supply board is still powered up.
Al.

[GITRDUN]

Are they suggesting the same size protector for each location. Surge protectors have different energy dissipation capacities. It may be less expensive to put large ones at the junction box and smaller ones at each machine rather than large ones at each machine.

That could be why they suggest one at the junction box and one for each machine. The ones for the machines are much cheaper than the large ones on the junction box.

[GITRDUN]

Normally the power provider has a specialized team that deal with recurring power problems and apply the test equipment in order to trace the problem. I know here where I live in Canada they have a fairly large dept to deal with this.
It is strange that it only occurs at switch on, as usually when you turn the breaker on the controller is normally powered up after this takes place, so it is rare that the whole machine is powered up by the disconnect.
BTW, are the machine disconnects switched off? Or just the controller?
If the power Co. are reluctant to do anything and the majority of the faults occur to the CNC control side, I would measure the 200vac or control voltage supply side on the machines, it would be very simple and cheaper to reduce the 100vac or 200vac as the case maybe, with a small buck boost transformer, I have done this on these types of machines as Fanuc for e.g. allow ±10% on the supply.
Al.

The powere dept is not interested in finding a problem, that would mean the city would have to spend some $ to fix the problem and thats not going to happen.
The machine controls are turned off and then the main breakers are turned off. The problem ocurring on initial power up is what is confusing. With the machine breaker off during the night how could a power spike get in to damage any electronics? So it must be happening when the main switch is thrown or when the controls are turned on.

[Al_The_Man]

If the machine disconnect on the main panel is off, there is no power to the cabinet at all.
And the CNC controller is typically powered up after with the control on P.B.
I would monitor the 100vac or 200vac 1 ph control voltage at first power up in the morning.
And see if it is within the ±10%.
Al.

[GITRDUN]

Ill have my maint man monitor the voltage when the switches are thrown. So if it is above or below the 10% what can we put in line to keep the voltage steady?

[Al_The_Man]

If it is a case of the supply fluctuating and being on the high side, and the majority of failures are prevalently in the area of electronics and controller cards and power supplies, you may get away with just taking care of the control supply as I mentioned earlier, if for instance on the 6M's the 200/100vac is over the 10% a buck/boost transformer on the output of the transformer supplying this voltage could be a cheap and easy fix.
If you have the schematics for the machines, it will help in the placement of the B-B Txfmr's.
You do not need to get the larger types advertised for motors etc, a smaller control transformer of the correct Va rating will work.
Al.

[GITRDUN]

The 200VAC is checking at 210V. Our incoming 480V power is checking at 495V. Both within 5%. However we went ahead and checked voltage on all the lathes and found that in one shop our incoming voltage is at 225V from the line and the other shop has 245V on the line. So some of our lathes were getting 225V fed into the machine after going through the machines step down taps. We adjusted the taps so they now are getting the required 200V. Dont know if this was enough to cause problems but that is a 12% higher voltage than what is suposed to be going in.

[GITRDUN]

Had another mill burn a trace on a board on initial power up this morning. Fixed the trace and replaced a blown fuse and it was back up and running. Its pretty much routine anymore.
We have started leaving our most problematic machines on by just pushing E-stop and not shutting the control or main breaker off at the end of the day. After a few weeks of this none of the machines that are left on have burned up any boards. That may be a fluke or not but thats the longest stretch weve had all summer without a control board getting burned on those machines. My only concern is alot of the machine components are left powered up at all times like solenoids and such but so far so good.

[Gundawg]

You should request the power company install recording volt meters to monitor your incoming service it is free. They will set these meter that will record time, voltage level and Amp draw.

If your power supply is the problem they have to fix it and a call to the PUC will get them moving. The problem could very easily be on your side of the service don't discount that. That is where I would start.

If your service is too small you could be getting a large voltage drop that could be causing you problems start with the power company they can be a good resource use them.

Mike

[RCaffin]

Yeah, well-known problem. Been there, had that. Put the fires out.
Many sources:

1) Inadequate local grid.
That means the supply authority is running the local grid right at the top of the voltage limit, such that during the day at full load everyone gets just enough voltage to meet specs. But first thing in the morning, before the community draws a lot of power, you may be well over the reasonable voltage tolerances.
A technical explanation here is that the local grid has too high an impedance. It should be something like 0.1 ohm or less, but it is probably higher.
The solution here is to apply enough political pressure that the supply authority upgrades the local transformers. They will be running pretty hot during the day anyhow. Costs them $$, but you can bet the supply authority knows about the problem already, even if they won't admit it.
(When I told the supply authority I wanted 3-phase at 50 amps per phase they got all upset and wanted justification, but they did install a new transformer. Suddenly, all my neighbours were getting stable power - and they came and thanked me!)

2) Phase Imbalance
Let's imagine you suddenly pull 50 amps on two phases and 10 amps on the third phase - and the local grid is poor or has high impedance. What is going to happen? The neutral is going to be all over the place, rather than sitting at ground. I have seen the neutral at 100 volts in a 240 grid once. Extremely dangerous stuff.

3) Inadequate grounding
Many places around the world have really strict rules about grounding. Australia, where I live, is one of them. But America does not have strict rules, and poor grounding can add to the above problems.
The solution here can be as simple as 6' of water pipe driven into the ground near the switch board. But get a good electrician to do this. And then run good earth cables to every machine and earth them properly.

4) Switch-on spikes
You throw the breakers and suddenly lots of current flows. Of course this is gouing to generate lots of EMI. The inadequate local transformer or high source impedance mentioned above is part of the problem here, but it is also a fact of life.

5) Inadequate equipment design
Yes, the makers of the electronics in your machines have to take some of the blame. Ideally the equipment would have nicely regulated power supplies, so that the 5, 12 or 24 volts sent to the circuit boards were just that. And the heavy power side of things would be optically isolated to avoid any other problems. But I have seen many bits of American gear where the manufacturer has ASSUMED that the grid will be a nice stable 110 volts (or 220 V) without any noise. Ask them and they will probably tell you that supplying their gear with 245 V instead of 220 V is going outside their specifications, and the damage is your responsibility. Yeah, right! And yes, been there myself! (Melt-down.)

What to do?
Get the grid fixed. (Hard.)
Put heavy spike suppressors on the inside of EVERY machine - GE MOV varistors are very good. This is cheap!
Add filter capacitors inside EVERY machine - proper mains-rated ones! This is cheap!
Put ferro-resonant transformers on the more sensitive machines, or at least on the supply to their electronics. Not 'isolation' transformers, ferro-resonant ones. This is less cheap, but very, very good. I have one powering our network. It handles even lightning.

Cheers

[GITRDUN]

Thanks for the tips. I dont think its possible to get the local authority here to solve their end of the problems. Not at least without getting lawyers involved and i dont want to go there.

I have wondered if there is a device that can be installed to bring the power up nice and slow when the machines main breaker is switched on. Rather than going from no power to full power in a blink. Maybe it would cause more problems than it would solve, i dont know. Just seems like thats where the majority of our problems occur.

Ill ask my repair guy if he knows anything about installing the capacitors and such as you stated. (thats all over my head)

[RCaffin]

Thanks for the tips. I dont think its possible to get the local authority here to solve their end of the problems. Not at least without getting lawyers involved and i dont want to go there.

Which is, of course, exactly what they are hoping you will say.
Yes, I do understand your position, but I have to disagree. As long as you are not willing to call them to account, they will treat you like this. Politics, greed, laziness ... they have got it.

I have wondered if there is a device that can be installed to bring the power up nice and slow when the machines main breaker is switched on. Rather than going from no power to full power in a blink. Maybe it would cause more problems than it would solve, i dont know.

Yeah, you got the last bit right. Bringing the mains up slowly can often cause serious brown-out problems. Especially kills AC motors. I would strongly recommend against this idea.

Ill ask my repair guy if he knows anything about installing the capacitors and such as you stated.

Pretty basic stuff for an in-plant technician.

Cheers