I used to have a "wild" leg 240v 3 phase system as well, so this is just what my beliefs and observations were. I'm not an electrician by trade so I can't say I have any concrete insights here, just what I think worked for me. I had a shop in an old industrial marina area of town, and it's the only time I've encountered such electricity. First, in my case the "wild" leg was about 25% higher than the other two legs when measured to ground (rather than leg to leg) rather than being lower. Leg to leg was a bit different, but you still got more voltage by combining the "wild" leg with either of the two regular legs, and got the expected 240 volts across the two regular legs. I can't recall the voltage from regular to "wild" leg, but by vague recollection is that it was something like 265 volts.
What I did was be sure to run the two regular legs on the inputs that eventually made their way to the controller. In most machines I've worked with or on, the controller and all single phase components run on the first two incoming leads, and the third lead is only used for the three phase loads. There isn't really a combination or balancing of these leads, meaning if R,S,T are the incoming lines, R and S do all of the single phase. T is never used on any of the single phase components. With that in mind, I made sure that R and S were fed by the two regular line values while T was reserved for the wild leg. If I needed to reverse a motor rotation, I reversed R and S when it needed to be machine wide, or any two lines leading to a load at the motor where a specific motor was backwards (like an external vacuum pump). I ran my machines on that kind of power supply for years without issue. Now, I have a smaller shop with single phase only, so I'm using a phase converter which causes the same general problems. Again, I power the single phase equipment with the two standard legs and run the "made" leg to the three phase loads only.
Not sure if that will help, but I'd recommend looking in the control cabinet and seeing if you are running your single phase off of the "wild" leg. If so, I'd recommend changing that.
If you Google wild leg or high leg Delta it will explain the setup, the neutral is not necessary on the feed to most CNC machines, just the 3 ph supply.Originally Posted by RCaffin
High leg delta is not used very much nowadays.
I agree, this is why I think that monitoring where the trouble ends up, and that is at the 100v 200v control is probably the easiest.
Al.
CNC, Mechatronics Integration and Custom Machine Design
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
I googled, and I was horrified!If you Google wild leg or high leg Delta it will explain the setup, the neutral is not necessary on the feed to most CNC machines, just the 3 ph supply.
High leg delta is not used very much nowadays.
That sort of connection is probably illegal in Australia, and it seems it is 'peculiar' to old American systems. It is not found outside America afaik.
I have a 3-phase supply, and I believe ALL 3 phase supplies in Australia bring genuine three phases plus neutral into the destination. To be sure, much heavy machinery runs delta, but very often the neutral is used for the control gear.
All I can suggest is that GITRDUN bites the bullet and gets his factory upgraded to what I (and the rest of the world) would call a proper three-phase supply. Yeah, it might cost, and yeah, it might disrupt for one day (no longer than that). But that would be cheaper in the long run than the mess currently blowing up his machines.
Cheers
Roger
Many machines to N.A. originate from elsewhere in the world and the majority require just 3ph and ground, either star or delta, which to the machine, appears no difference, the control circuitry, being a lower voltage usually originates from a 1 ph control transformer in the M/C enclosure fed off of two phases.
One side of the 120v secondary is referenced to earth ground and declared a neutral.
Machines such as Japanese origin usually use 100v or 200v 1ph control, of which is sourced the same way, a 1ph control transformer.
Al.
CNC, Mechatronics Integration and Custom Machine Design
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
This is exactly how my current machine is configured. You get a 100v or 200v and "neutral" coming off the transformer, which is running off of two legs of the three phase. That's where I think it's important to use the two standard voltage legs and keep the "wild" leg off the control transformer. Not sure it that's the advice that is normally give, just what I have assumed is best for keeping the voltage to the control circuits stable.
Well another melt down this morning. However the machine that blew a board was the only one in the shop that has not been left on for the last two weeks. It has been used every day but the operator has been turning it off every day (the machine is one of my personal machines i own so i instructed my operator to turn it off everyday). We had one other machine that had the control off when we came in but after turning the control back on its fine. So again something happened in the night. But again this doesnt explain to me why the machines are popping boards when they are switched on and not when they are left on. This makes no sense.
I agree about monitoring the supply voltage to the control boards. When the electric contractors return i will ask them if they can put their meter on one of the machines.
You have a competitor just down the road who is sneaking in at night and doing the dastardly act!!
I don't remember if, (there have been so many posts now), if you are switching off right at the disconnect or just powering down and leaving the Disc. On?
Al.
CNC, Mechatronics Integration and Custom Machine Design
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
As of 11/4 we are simply pushing the E-stop in and leaving the machines on. This way the controls are powered up all the time. Before that we would turn the control off then switch off the main breaker disconnect built into the back of the machines. Which physicly disconnects the machine from the incoming power wires.
So as I a understand it, you have the problems apparently at power up but it is/was worse when you powered off the disconnect?
When you had the recent power recorder on, it most likely was on when the machines were up and running?
IMO you need that chart recorder or data logger hooked up, preferably to the control power, and powered on before the machines receive power, I am guessing that there are signs even when no damage is done, just that they are not sufficient to cause damage?
Al..
CNC, Mechatronics Integration and Custom Machine Design
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
That is correct. The machines were on while the recorder was recording. Ill see if i can get them back in here to connect the recorder to a machines power supply. And we will then turn the machine off at night and back on in the mornings while the recorder is monitoring it.
I suspect that is the only way to find it?
Al.
CNC, Mechatronics Integration and Custom Machine Design
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
I agree with Al. Ideally you would monitor both the supply going to the machine and the supply inside the machine which goes to the electronics, and that monitoring should run for several days continuous.
I will add that the moment of switch-on is KNOWN to be the time when the electronics are most susceptible. There is a big inrush of current into the transformer which can cause big surges on voltage. This especially happens when the switch-on is done at the peak of the 60 hz waveform. This is why savvy computer users never switch their machines off.
If machines have powered down overnight before, that does imply that the supply is serious bad. That wild leg will not be helping either.
I have had this sort of problem before. At one stage our computers were all blowing up at 9:00 am. Turns out that a couple of nearby factories were all powering up their machines at ... 9:00 am. Surges everywhere!
Cheers
Roger
You also need a record capable of capturing the transients. There are any number of recorders on the market right now Some of them optimized for certain jobs. You would want to make sure the recorder can capture those transients. Further you want to know what is happening line to line and line to ground on all phases.
This is true but better power supplies can help with power up behavior.I will add that the moment of switch-on is KNOWN to be the time when the electronics are most susceptible. There is a big inrush of current into the transformer which can cause big surges on voltage. This especially happens when the switch-on is done at the peak of the 60 hz waveform. This is why savvy computer users never switch their machines off.
As aside note sending the power supplies back to the factor asking for a diagnostic report of what they think might have happened can be useful. One machine I work on has a power supply with "crowbar" over voltage protection. Basically an SCR fires if an over voltage is detected on the DC bus. This protects the rest of the electronic in a spectacular manner. I didn't mention this before because it doesn't sound like that sort of failure. In any event; if they know how the supplies failed in the past the manufacture may have valuable clues for you.
Usually that would be an all or nothing thing. At least for the same controllers which should behave similarly. For example at work if RGE does something that flickers the power supply one series of mold machines all go down at once. I still wonder about grounding, poor ground has cause some odd failure in electrical systems. I'm leaning towards something happening internal to the plant.If machines have powered down overnight before, that does imply that the supply is serious bad. That wild leg will not be helping either.
You would think whatever recorder was put on the AC lines would have caught such surges though. However they could still have something internal to the factory causing these issues. I find it interesting that the problem is seasonal. If I knew a bit more about what was happening I might suggest strategic placement of Metal Oxide Varistors and other transient snubbing devices.I have had this sort of problem before. At one stage our computers were all blowing up at 9:00 am. Turns out that a couple of nearby factories were all powering up their machines at ... 9:00 am. Surges everywhere!
Cheers
Roger
I agree pretty much with everything Wizard was writing.
Story #1: A computer in the main office kept dying. The cause: the big solenoids in the big photocopy machine were creating big spikes on the mains. Filtering and mains conditioning was needed.
Story #2: Shortly after our house was built and we moved in, I noticed some of the lights flickering. I tracked the problem back to the main board. The supply wire coming in to the main neutral link was loose in its hole. The screw had never been tightened. The electrician was a bit subdued when I told him about this. Of course, the Council Inspector never checked it either.
Story #3: One of the power drivers on my CNC keeps dying right now. Just one of them. After closely inspecting a 'dead' board (this weekend) after replacement, I am now of the opinion that all that batch of PCBs have dry joints due to the politically-correct but technically stupid use of lead-free solder in production. You might like to note that the military (among others) have never accepted lead-free solder for this reason, and in fact the lead-free solder is fast disappearing from the shelves around here.
What do all these stories mean? They mean that to sort out a case like GITRDUN's, you will probably needed an experienced consultant electrical/electronics engineer for a few days. However, upgrading the incoming supply to a modern transformer (on the grid), setting the voltage taps properly, and installing a full earthing system around the factory will probably feature somewhere.
Yeah, dollars. But what's the cost of the continuing lost production and board repairs? Not to mention the health effects of the frustration?
Cheers
I am awaiting on delivery of a recording meter so we can start monitoring these electronics ourselves as i dont have time to sit and wait for the contractors to come back and poke around when they get to it.
In the mean time i have been reading allot about isolation transformers. It seems to me that an isolation transformer for each machine would completely eliminate the chance of transients in the main lines getting into the machines if that is part of the problem. A single isolation trans on the incoming main lines would eliminate outside transients from getting in but wouldnt stop transients that are genereated by the equipment in our shop from hopping around from one machine to another. In which case i would need an iso trans for every single machine in my shop would i not?
Does that sound like something that could be beneficial in this case?
Hi GITRDUN
Sadly, no, a basic isolation transformer would not solve allyour likely problems. An isolation tranny will block fast transients which may have come from lightning strikes or huge solenoid spikes, but it can't stop slow voltage surges or swings, and I suspect these are more likely to be your problems. Frankly, I would not bother.It seems to me that an isolation transformer for each machine would completely eliminate the chance of transients in the main lines getting into the machines if that is part of the problem.
However, what you could do is something similar. You could put ferro-resonant or constant voltage transformers (CVT) at the front end of the electronics parts of your machines (the bits which blow up), and they could work. I would suggest one per machine, to keep the machines separate, and they are only single phase. However, they are not nearly as cheap. But they can just drop in place, so installation costs are very small.
Hum ... they may not be cheap, but the capital expense might be insignificant spread over a couple of years of problem-free running? What does your down-time cost per hour? Many of these units are custom-built to your requirements. You would need to get quotes to see. The manufacturers could tailor the CVT to each machine.
I have an off-the-shelf one running our network. It handles lightning strikes and brown-outs and things that go bump in the night. It's been there for years. Mind you, it's just 1 kW and weighs >50 kg! Yeah, transformers are heavy.
Cheers
The CVT or CVS series that Sola sell in the N.A. including the Solatron for 3 phase applications might cure the problem, you could always try it on one machine to check the results.
They are not the most efficient way to regulate as they run in saturation all the time. but they claim to regulate to within 1%.
Al.
CNC, Mechatronics Integration and Custom Machine Design
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
Hi Al
Yes, it is a Sola CVT unit I have as well. Probably 25 years old now - or more. Outlasts many computers ... :-)
Cheers
Roger
Ok i finally recieved my voltage recorder. Set it up to record every 1 second and let it record from just after we shut down for the day yesterday until about an hour after we powered up machines this morning. I downloaded the data and it shows a serious drop in voltage on the high leg and a small drop on the 120V line that also powers all of our 110V in the shop and office. The drop hit about 45 minutes after the equipment was up and running at 6:47 this morning. The high leg which runs around 212V dropped to 145V and bounced up and down in that range for about 8 seconds, and the 120V line dropped from 123V to 110V at the same moment.
I plan to leave this recorder hooked up and recording 24hrs a day and will download and save the data every day. Next time a machine goes pow i hope to be able to see something on the recorder at that exact same time. Nothing solid yet but at least i know i can now catch something in the act. Ill post some results when i have more info.
It sounds like you are on the right track, I would inform your power service provider after all the test are done, especially about the drop, unless it is due to an on site dist. transformer undersized?
Al.
CNC, Mechatronics Integration and Custom Machine Design
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
Reply with Quote
