Industrail surge protection

[kbsparky]

Today, one of my customers experienced a short power outage, probably for a second or less. It appears that this outage and the surge that followed fried a computer module in one of his machines. This is the second time one of these modules was cooked as an apparent result of a short-duration outage.

These modules ain't cheap: $3200 or so each. This particular machine has about 6 of `em in there, and he wants to do something to protect them better.

Since the high-tension lines out on the road are overhead, and pass through wooded areas, things like high winds can cause the trees to momentarily hit the lines, and make the power blink, flicker, and so forth. I figure when this happens, there is also a surge associated with the event, and this can cause the damages he is experiencing. Sometimes, when such blinking occurs, the power restores itself with one phase left dead. (I figure there may be a dead squirrel or buzzard out there as a result )

The service is a 3-phase 240 Volt open-bank delta connected, 400 Amp. The piece of equipment has its own 100 Amp feeder circuit, connected to its internal computer system, controlling various motors, servos, solenoids, etc.

At first, I thought about maybe trying to isolate the internal computer from the rest of the equipment, and using a UPS/APC unit might keep the computer up and running. But, it is integrated into the equipment and functions that isolating it may not be feasible. The servo/modules that got fried are connected to all 3 phases, so this is not your ordinary computer system.

So, I figured that installing a power-conditioning surge unit at the service panel should be able to help. I have outlined the service voltage and size to one of my suppliers looking for price/availability of an appropriate unit. These are probably capacitor bank/MOV units?

Do any of you have experience in dealing with this type of problem?

 

[wirenut1980]

Sounds like a CNC machine. I have investigated several of these and monitored afterwards only to find nothing out of place. I have seen monitoring data of enough momentary outages and restoration of power to say that a surge probably did not happen...at least not one large enough to damage anything.

The only theory I have heard is that many electronic devices have a current limiting device installed and with a quick off-on scenario, that device is bypassed. I have not been able to confirm this theory in any way. Your idea of putting the damaged part on UPS would stop this from happening, but usually is only cost effective if the damaged part is in the 120 V and below controls.

If the machine was single phased, that should be easily explainable why a motor would fail.

If anyone knows the problem and solution I would be interested to know as well.

 

[zbang]

Manufacturer have any ideas? I'd consult with them about adding a UPS for the computers. (I always question why machine makers don't deal with problems like this in their design. They've got to know that power blips happen.)

 

[__dan]

Suggestion

Suggestion

The question and the responses are heading in the same direction I am thinking.

At the service, for the incoming protection, you are looking at a TVSS type device, a relatively small box. If it has MOV's in it, which is likely, the MOV's are sacrificial elements over time.

At the machine, I also would be looking to separate the power feed for the control transformer, PLC power, and servo driver module power. Definitely with factory consultation and approval, if the sensitive electronic loads can be separated in the machine to receive a 'line conditioned' power source, I would be thinking of installing a ferro resonant type line conditioner as a second feed to the machine. The line conditioner does nothing for phase loss or outages over one cycle.

I have a preference for ferro resonant line conditioners in a certain range over a UPS. There are a variety of UPS types to choose from if you need the battery backed ride through capacity. There was a company, Best UPS, that made ferro resonant output UPS's. I buy these used, takeouts, when I see them and change the batteries. They are tanks compared to an small APC type UPS, very well built.

I would not be looking at a whole machine UPS or line conditioner, just a TVSS for the over 5 kVA load.

I'm thinking the controls, PLC, and servo driver modules are between 2 and 5 kVA total, and the bulk of the machine loads are large drive motors 15 hp+10hp+5hp+5hp. The large motors with starters are unlikely to get damaged or confused by utility switching transients. The motors with VFD's may prefer another TVSS at the machine.

 

[gar]

091215-1339 EST

We have 5 HAAS CNC machines. Several if not all were powered at the time of the northeastern blackout several years ago. I was not at the shop at the time, but I was in a basement with incandescent lighting. The lights went dim, then off, back on with over voltage, dim again, and off completely. Could have been one more cycle in there.

We lost several fuses in the CNC machines, but no other damage. An Ethernet switch had several channels destroyed, but no damage at the other end of the Ethernet connections. The primary line fluctuations were quite severe.

At another period we had arcing connections at the pole transformer on both a hot and the neutral. The neutral problem caused high and low voltages in the shop on the 120 circuits.

Our older CNCs have separate three phase to DC power supplies for the servos and main spindle. The newer ones use a single DC supply for the main spindle and servos. Since these are mostly 20 HP machines one does not want to use big Sola transformers at the input. Thus, one gambles on no major problems. Short voltage transients are not likely a problem because rectifiers feed capacitors and full wave rectification is used. Phase loss is detected by the control and shuts the system down. We probably should have some transient limiters, but don't.

My ideal concept for this type of application is to employ a common battery bank supplying the 360 V bus for all the machines. Then excess regenerative power would be stored in the battery instead of being dumped in a resistor. Also the batteries would keep the machines running for short power outages. Maybe as long as 60 seconds. This would eliminate the major time loss to restart some jobs, or damage to the part being made.

Ahead of all this would be isolation via a motor-alternator system with high mechanical inertia and a backup engine to drive the alternator on loss of primary power. This would eliminate the outside glitches.

The problem is cost.

For 1 KVA or less loads I favor a Sola transformer as a rock solid device. This is what I use for the input to our gaging systems. The video display and all electronics are supplied from this transformer. With this arrangement we have never had any problems in any plants with any sort of power problems. The power can fail, or lightning can cause transients, and on restoration of power our controls are back working again. No reloading of data, programs, or damaged equipment. Part of this results from basic system design philosophy. The Sola is instrumental in preventing component damage and reduces internal power dissipation problems with large changes in input supply voltage. However, the Sola itself is wasteful of power. The Sola is good from 95 to 130 V for the input.

.

 

[petersonra]

You should probably start with the most obvious thing - put a TVSS at the service entrance point.

You might also consider another TVSS close to the machine.

CNC electronics tends to be a lot more sensitive than you might think. They make them as cheap as possible.

You might also want to look at putting an isolation transformer at the machine. Check the manual for the machine carefully. Its possible it expects a WYE power feed.

 

[kbsparky]

The motors are all fine in this machine. IT is an automated saw station, which is part of a production line to cut sections of wood for building structural trusses. The main motor is a 7.5 hp saw assembly.

I'll try to stop back in there and get a high resolution photo of the control panel and post it here.

I had a look at the utility pole outside. On it is 2 transformers, an "open bank" delta configuration. Sort of a cheap, back-door 3 phase power supply. The larger transformer handles both the single phase and 3 phase loads, while the smaller one provides the high-leg completing the 3 phase requirements. All 3 phases are available on that pole, I don't know why the POCO only used 2 of `em, except that they might have tried to save a buck when originally installed. This factory has been in place for 30 years or so, and the open bank delta services were commonly installed in those days.

When they lose one phase, it is always the high leg, since the other 2 phases originate from the same transformer.

My supplier originally recommended a ferro-resonant type power line conditioning transformer for the circuit supplying this machine. I asked them again about the capacitor bank/MOV units that should be available. We may end up utilizing both of these solutions.

The main breaker in the machine is rated for 70 Amp, the breaker in the service panel is 100 Amp. Probably upsized to handle the starting loads.

Another factor here is the length of the circuit run: It looks like it may be 150 feet or more from the service panel.

Reiterating our main concern is prevention of damage to electronic components from surges/spikes in the main power line. We do not need to keep the production going in the event of an outage.

 

[kbsparky]

Manufacturer have any ideas? I'd consult with them about adding a UPS for the computers. (I always question why machine makers don't deal with problems like this in their design. They've got to know that power blips happen.)

The manufacturer agreed with me in that we should install a TVSS unit at the main service panel at the outset. We are inquiring about an appropriately sized unit for the 400 Amp service.

 

[gar]

091217-0711 EST

In our area open delta is very common and a good solution. Typically in a commercial application the single phase load is greater than the three phase. If the three phase load increases a third transformer is added. Two transformers is less capital equipment and installation cost.

Three phase 240 and a 70 A breaker in the equipment seems somewhat high. If the saw spindle is an induction motor directly off the line, then I might expect 20 A line current at full load as a ballpark figure. Guess inrush current at 100 A, but that is only for a short time. If the main spindle, saw blade, is on continuously, then inrush should not be a problem. If turned on and off once a minute, then it might overheat the breaker if a smaller breaker was used.

Try to determine how large the load is excluding the main spindle. If small enough go with a Sola for that part of the system.

You can take advantage of the long supply lines with a transient limiter at the machine because the impedance of the lines helps limit the current from a transient to the transient limiter.

Note: MOV type limiters as compared to Zener diode types have a soft clamping characteristic. For a 120 line an MOV may be designed to start conducting at maybe 150 V sine wave (212 sine wave peak) and might reach 800 V at 100 A. See the GE curve figure 3b at
http://www.eeel.nist.gov/817/pubs/spd-anthology/files/MOV announce BW.pdf
There are many Internet sites on MOVs here is another
http://www.protekdevices.com/pdfFiles/articles/ABD_Comparison.pdf

The major point is that the peak clipping voltage of an MOV is many times the peak of the AC supply. Thus, a high impedance before the MOV is useful and low pass filtering after the MOV is important.

.