Machines constantly blowing boards on startup, cant figure it out

[GITRDUN]

Ok, ive posted a thread about this before and i did get a lot of good information but most of it was aimed towards dirty electric from the power company. Let me explain what is going on and i would like some knowledgable people to help me reolve this once and for all.

The problem :
We own around 20 CNC lathes and mills, most are built from mid '90s up to just a few years old and run on 3-phase 240V high leg delta. The oldest are early '80s model horizontals which are running on 3-phase 480V high leg delta. They all experience the exact same problem. We constantly have power supply and control boards blow when powered up in the mornings. No rhyme or reason to it, all machines we own have this problem. And it is extremely rare for us to have a board blow while running a machine during the day. It ALWAYS happens when the power is switched on in morning power up. The machine will be running fine the day before when it is powered off and the next morning when it is powerd up it will have alarms when the control boots up and a control board will be at fault. What is blowing on the boards the majority of the time is capacitors, volt regualtors and usually some micro chips that are in the same cuircut as the cap and regulator. My maintenance man is the busiest man in my shop trying to keep up with replacing the bad board components. In the last 15 working days we have had 7 control boards blown during power up. When the weather changes from warm to cold we are plagued by this phenomena for about a month or so and then it will settle down and we only have a few a month brake down. When the weather changes back to hot we will experience a high rate of brake downs again for about a month or so.

What we have tried to resolve this problem:
I have had surge protectors installed in every machine. No difference.
We installed ground rods at every machine. No difference. Rods have since been removed.
We tried using E-stop pushed in when turning machines off and not taking it off of E-stop until the machine is powered completely up in the morning. No difference.
Checked that all machines have a ground wire going from the control box to the ground wire coming in from the utility pole outside. There are a few machines that currently dont have a groudn wire hooked up but all machines grounded properly or not experience the same problem.

I have been told that placing a MAINS RATED capicitor at each machine might help our situation. I would like to try this but i am not experienced enough to know how to go about this. My maintenance man isnt sure what exactly a mains rated cap is or how it should be installed. Can someone please point me in the direction of how to do this?

In my opinion this is not a dirty electric problem, if it were i would expect to have problems during the day when the machines are running which is extremely rare for us.

Similar Threads:

• Open Source Boards Available, Stepper Drivers and Break Out Boards
• Help me figure out what machines to buy or if I should build a machine.
• Star VNC 32 Blowing fuses on startup.
• Light Machines Benchman MX CNC Control Boards
• Problem- VF-3 pcool constantly homing

[Al_The_Man]

I recall your previous posts.
The Caps referred to are AC rated as in Power Factor type.
The problem certainly appear to be (high) voltage at switch on.
I would think a sure way to obtaining an answer would be to use a chart recorder placed at a particular machine and start recording just prior to switch on to some time after.
Power companies, such as the the one where I am do this, including poor power factor monitoring.
Or the cost of renting a suitable one may be worth the cost in light of all the failures?
Al.

[GITRDUN]

I will look into buying a meter, from a quick search i see prices from $800 ~ and up. Certainly a justifiable expense.

But lets assume you are correct and the problem is high volt at switch on. Would installing mains rated caps at the machines help to buffer this problem? And what would cause high volt at switch on? We have checked voltages before switching machines on many times and have yet to see it higher than normal. But we dont check it every day and it could certainly be happening.

I am not interested in poking a stick at the local power company, i live in a small community where eveybody knows everybody and i would prefer to be on good terms with the city and officials. If i can possibly resolve our issues on my own thats what i would prefer.

[Al_The_Man]

Would installing mains rated caps at the machines help to buffer this problem?

At this point, the cause is unknown, so what effect the caps may have is also unknown?
Normally these are used for P.F. correction, which is mainly to improve and decrease power costs.
If you do fit a chart recorder at a machine, I would switch all other machines on first when monitoring, and then for a second test by switching the one at the meter on first.
Also a test at power off time.
There are different types of meter, but mainly you need to monitor the voltage and current if possible, dual trace recorder.
Al.

[GITRDUN]

Does this apear to be a suitable data logger for my aplication? Extech DL160 Dual Input AC Voltage/Current Datalogger Not sure if i need a meter which will monitor all three legs simultaneously or do we just need to monitor one leg.

[Al_The_Man]

It seems a pretty good unit.
You would need a multichannel to monitor voltage and current on all three.
I would detect which phase pair feeds the 1 phase control circuitry as this is probably the one that anything is going to appear on.
And place it on one of those phases.
If nothing seems to show up, you could also use it in the machine itself on the L.V. control to confirm anything there?
It would also be handy if the reading was triggered by a certain voltage level etc. I just skipped through the manual.
Al.

[RCaffin]

Hi

I think I did reply last time. Your problem is a bit wierd, but it WILL have a cause.

Two steps to take for a start.
* Have your maintenance guy check the name plates on the machines which give problems. Are they rated to 240 VAC, or to 220 VAC? I ask because a lot of America runs on 110 VAC, and a 2:1 step-up is common.
* Get (hire?) a 3-phase mains data logger. You may even be able to borrow one from your local utility. Monitor all 3 phases for several days continuously. It is not unknown for the supply to be a bit high in the mornings.

My thinking is as follows. It seems you are blowing the front-end capacitors and the regulators. The rest of the board dies when the regulator dies, so ignore that. This suggest the mains input is a bit too high for the design. If the gear was designed for 220 VAC and you are getting 240 VAC, going up to 260 VAC in the mornings, this is a possibility.

A test you could run yourself right now. Pick out a machine which has died several times. Leave it OFF until, say, 10 am. Then power it up. If it never blows, this suggests the morning supply may be a bit high. This happens: I have had it myself.

Option: ask your utility company to check the supply into your plant and to consider turning it down a bit. Often they can do this by tap-changing. You might find the local utility engineer is quite willing to help.

Cheers

[GITRDUN]

We went through all machines a couple years ago and adjusted their transformer taps to make sure the control voltage is correct. Most are built for 220V and our shop is on 240V.
I will be getting a data logger and monitoring the line voltage.
I dont have the option of not turning a machine on until 10AM, we have deadlines to meet and if the machine is not broken down usually they have to be running 8 hrs a day. All of our machines experience this problem so there is no way to know which will be the next to be stricken.

[acannell]

We went through all machines a couple years ago and adjusted their transformer taps to make sure the control voltage is correct. Most are built for 220V and our shop is on 240V.
I will be getting a data logger and monitoring the line voltage.
I dont have the option of not turning a machine on until 10AM, we have deadlines to meet and if the machine is not broken down usually they have to be running 8 hrs a day. All of our machines experience this problem so there is no way to know which will be the next to be stricken.

a 10% increase in mains voltage is probably not going to destroy capacitors, is it? (240V to 260V)

i would bet on a faster transient of some kind, something that rapidly charges and discharges the capacitors in the power supply to the point where they dissipate enough power to pop. that or a very high voltage transient which breaks down their dielectric.

it might be useful to buy an older digital scope, like a $800 tek tds 1000, and monitor the line voltage with that. most dataloggers are pretty slow and probably wont capture a transient with a period less than a second..or much less one faster than line frequency. i dont know how long of a transient and of what magnitude would kill your boards, so youd probably want to start with a more sensitive device like a scope.

you may need an experienced electronics technician to figure this one out..unless you do some detective work and see if you can figure out any patterns or relevant power activity (big machines being turned on/off, other people/companies having issues), that may coincide with the destruction.

can you take a picture of a destroyed board, and maybe put marker marks on the parts that are dead if its not obvious? might help us

[Al_The_Man]

I will be getting a data logger and monitoring the line voltage.
.

The one you posted the link on has a 1-800 toll free tech line for questions on the best way to capture your problem, it may be worth running it by them?
Monitoring the 120vac control side could be one best option, this is where all you low voltage control is fed from and what ever is appearing on the 3ph line is going to appear on the 120v 1ph side.
Al.

[RCaffin]

The one you posted the link on has a 1-800 toll free tech line for questions on the best way to capture your problem, it may be worth running it by them?

Too right! They might also be able to suggest a consultant who can come on site. He would be worth it.

Monitoring the 120vac control side could be one best option, this is where all you low voltage control is fed from and what ever is appearing on the 3ph line is going to appear on the 120v 1ph side.

True, but sometimes it helps to see all three phases, especially if the neutral is bouncing around.

Photos might help to confirm what we suspect.

Some specific 'mains data loggers' include provision for catching transients and spikes. After all, that is probably half their application. They can give you a readout on line voltage AND transients.

I dont have the option of not turning a machine on until 10AM, we have deadlines to meet and if the
machine is not broken down usually they have to be running 8 hrs a day. All of our machines experience
this problem so there is no way to know which will be the next to be stricken.

And if you don't solve the problem some of your machines will be broken down every day and you won't meet your deadlines anyhow. You may need to look on things like this as an investment for the future.

Cheers

[Al_The_Man]

True, but sometimes it helps to see all three phases, especially if the neutral is bouncing around.

I suspect the machines have a typical supply of just the 3 phases, no neutral.
Al.

[wizard]

Interesting discussion!

One thing I haven't seen mention of is ground faults which could potentially lead to short circuits. Considering that you have wild leg delta systems you should have ground fault indicators some place in the system. If not they should be installed as it is considered good practice to have such in delta systems. If you don't have ground fault indicators you could have devices intermittently shorting to ground such as heater bands, coolant pumps and the like. It is interesting that this happens as the weather gets cooler, which makes me wonder how well the physical plant is maintained.

By the way a chart recorder or similarly a power analyzer, is a nice tool to have assuming you have a technician capable of leveraging it. What bothers me here is you have made no mention of actual line voltages which can be measured with a good quality DVM. If the line voltage is high in the morning (a very good possibility), this can be easily measured without an expensive power analyzer type device. It isn't uncommon for power companies to bump up the line voltage at the onset of winter, usually this isn't enough to cause problems though. I say usually because some electronics can be very voltage sensitive. If your maintenance guy hasn't already looked into voltage levels you might want to consider hiring a specialist technician that understands CNC and machine controls systems. It is one of the first things you would want to look at if you are having random failures.

[RCaffin]

If you don't have ground fault indicators you could have devices intermittently shorting to ground such as heater bands, coolant pumps and the like.

Delta wiring and intermittent faults with a 'soft' mains distribution - NASTY stuff! In the circumstances, VERY possible!

At the very least, ask the utility for help, or hire a specialist electrical techie.

Cheers

[Titaniumboy]

Is a neutral wire (in addition to the 3 hot wires and the ground wire) ran to any of your problem machines?

[Al_The_Man]

I am not sure if this was brought up at the last thread on this, but most machines of Japanese origin have 100vac & 200vac for the control voltage with a tolerance of ±10% for each.
Was this ever measured for tolerance?
Al.

[GITRDUN]

There is no neutral wire on our setup.

I spoke with our CNC repair company this morning and he told me it sounds more like a ground problem to him. After our conversation I called a company about 100 miles down the road to send in their "experts" to put some monitoring equipment on the lines and do an inspection of our electric system and ground. Hopefully they can find a problem and tell me how to resolve this.

In the mean time i had my maintenance man check the line voltage before any equipment was powered up this morning. Checked at 239V and the wild leg checked 199V. Once we powered up machines sure enough one of them blew the volt regulator out on its power supply board. If thats all thats wrong with it we will have it back up and running in a couple hours.

We have measured and adjusted control voltage on all machines in the past to make sure they are within their tolerence. All are currently.

So i should have some solid measurements and information in a few days.

[GITRDUN]

Ok so the electric contractors have checked the data from their recording unit and did not find anything unusual. I think it recorded for maybe a couple days. Doesnt necesarily mean there isnt a utility problem but could just mean that nothing happened while they were monitoring it. They are as puzzled by this problem as i am. Next step is to have them inspect all the grounds and maybe do some tidying up with the electric wiring in the building.

In the mean time we have been leaving all the machines powered up 24-7 and only pushing E-stop at the end of the day. Started doing that on 11/4 after the last machine popped a board on power up. Since then we have not had a single board blow. However one day a couple of the machines had the controls powered off during the night. One machine control was powered off and another right next to it was still on. And another next to it was off also. Powered them back up and no problems. So a quick surge or brown out definately hit that night. I know this isnt a solution to the problem but so far it seems to be helping. Either that or the problem has just gone away temporarily.

[Al_The_Man]

What I would consider an ideal test is for a pre/post trigger recording unit that only records when a voltage spikes over say, 10% - 12% and placed on the control voltage feed to the board systems, as obviously anything outside the machine is making its way through to the control portion of the machine.
Is it possible the record system posted earlier has this? It might be worth that toll free call?
Al.

[RCaffin]

There is no neutral wire on our setup.
I spoke with our CNC repair company this morning and he told me it sounds more like a ground problem to him.

I can understand the economics of skipping the neutral wire, but its absence is KNOWN to cause a lot of problems. For instance ...

In the mean time i had my maintenance man check the line voltage before any equipment was powered up this morning. Checked at 239V and the wild leg checked 199V.

I do not understand what is meant by 'the wild leg' here. If you mean two phases were at 239 V and one phase was at 199 V, I would be having kittens! That is a seriously UNbalanced supply and WILL cause significant problems. For that matter, it means the supply is way overloaded on one phase. NOT GOOD. If it means something else I am just confused.

Once we powered up machines sure enough one of them blew the volt regulator out on its power supply board.

I am NOT surprised.

So a quick surge or brown out definately hit that night.

Yup, fits in.

One thing is very clear to me already: your problems are coming in on the supply, imho.

Cheers
Roger