[Phishaholic]

I have had a couple of random power outages in the last few days. I was wondering what happens when you are in the middle of a part and the power cuts off and then back on in a few minutes? Will it continue to run the part or will there be like an E-stop? I don't have my machine under power yet. I was going to put a battery backup for my computer, because I don't like it powering off with out me doing it through the computer.

I had an idea to wire a relay to the 110 outlet and the E-stop circuit. this way I can run a battery backup for the CPU and if the power fails the CPU will see the E-stop being pushed and stop the machine and code. As long as there is power the E-stop relay will be in the ready position. Any thoughts or am I over thinking things?

Larry O

[zephyr9900]

Larry, at the minimum you'll likely need to replace the broken cutter after a power outage. (Ask me how I know...) When the power goes out the Z will immediately start falling (backdriving its ballscrew) and bury the cutter into the part as the spindle is spooling down. The state the control PC comes up in after the outage is immaterial since the part is ruined one way or another--part gouged and/or cutter snapped and axes references lost.

In other words, having a power outage is a bad idea. Our area of town goes through periodic power glitches, and I had ordered a UPS before thinking the problem out fully myself. The UPS is going on my desktop/design PC when I get around to it.

Randy

[justgary]

Larry -

I think your relay idea has merit. If the computer has a UPS to keep it on and your power relay opens the E-Stop circuit, you just might have a chance at recovery. No guarantees, but I say go for it.

If you have a Series II mill your Z axis should not sink when the power goes out. The Z brake seems very effective, and I have now stopped parking my spindle on a block when I shut the mill down. I still put the block below the spindle, but I am convinced that it will not sink with the power off.

Regards,

- Just Gary

[zephyr9900]

If you have a Series II mill your Z axis should not sink when the power goes out.

Bah, you young'n's with your fancy-schmancy Z axis brakes and safety razors and indoor plumbing and all...

Of course, you're right about the axis not falling on "new" machines, Just Gary.

But with primary power (i.e. 208V) going out, I wouldn't count on a smooth shutdown of the machine. The axes and spindle will just stop. Estop is not a controlled deceleration either--you're pretty much guaranteed lost position. I wouldn't count on picking up the part program without re-referencing the machine (Mach doesn't like running unreferenced) and re-indicating the part.

Now, if you could put two UPS's in series across the two phases and get 30 seconds or so of operation of the spindle and axis drives, you might rig the line voltage sensor in a UPS to trigger a controlled stop in Mach and do an orderly shutdown.

I'll bring my safety glasses...

Randy

[300sniper]

Larry -

I think your relay idea has merit. If the computer has a UPS to keep it on and your power relay opens the E-Stop circuit, you just might have a chance at recovery. No guarantees, but I say go for it.

If you have a Series II mill your Z axis should not sink when the power goes out. The Z brake seems very effective, and I have now stopped parking my spindle on a block when I shut the mill down. I still put the block below the spindle, but I am convinced that it will not sink with the power off.

Regards,

- Just Gary

i have never had an issue with my z sinking down with power off. mine has sat for a few weeks at a time with the head up and still been where i left it when i returned.

this thread does bring up something that i never considered. i guess i will not run any critical parts during a major thunderstorm.

[zephyr9900]

i have never had an issue with my z sinking down with power off.

I don't know what it is, 300sniper. My head has always dropped. After I performed the latest Tormach-specified Z-axis gib adjustment, the head didn't drop for several days, then went back to its old ways. I guess I need to stop using the high-quality way oil...

Randy

[300sniper]

I guess I need to stop using the high-quality way oil...

Randy

i just don't use any oil

my machine is a series 2, i guess. i have the brake on top but the older stepper drivers.

[Phishaholic]

Guy thanks for the thoughts! My machine is a used First Gen, I plan on doing the spindle upgrade shortly. After that i'll make a thread when I get it under way. I like the Idea of two units to run 220, I'm going to see if they make one for 220V. The main cause of power outages here this time of year is a large load on the power grid.

Larry O

[justgary]

Bah, you young'n's with your fancy-schmancy Z axis brakes and safety razors and indoor plumbing and all...

Randy -

If it hadn't been for you pioneers blazing a trail in your chuckwagons, us noobs would not have Series II machines to buy.

I think the Series II machines have a bit better control with E-Stop also, but I can't really be sure about that. Maybe someone from Tormach can comment about that. Whatever the reason, I usually don't have much problem restarting after an E-Stop. It could be just plain clean living...

Phishalolic -

I think that you will find that a UPS for 220V is much more expensive than one for 110V. It would be interesting to see if two 110 units would work. I'm afraid that they don't synchronize to the line voltage before it goes out, so the output of two of them combined would be anything from 0 to 220 volts. It's simple enough to test with an oscilloscope, and we have UPSs stacked like cord wood at work, so maybe I'll test it sometime. If they do synchronize, it's only a matter of buying them large enough to handle the load.

A better solution might be to rig up a 220 genset to run from a large DC motor and always run the mill off of that, with a charger holding the batteries up while the power is on. Wasteful, yes, but a more certain solution.

Regards,

- Just Gary

[StephanWenger]

Hi all,

JustGary mentioned the issue of keeping the two 110V USPs in synch. This is serious. It's unlikely that two 110V USPs would keep the 60Hz up closely enough to maintain the 180 degree phase shift. Not for an extended period of time at least. "Extended" could be as short as a couple of seconds.

Also, most cheap USPs are not built with a large inductive load in mind. I don't know what the load characteristics of the Tormach are (and would expect a difference between series 1 and 2 machines) , but I consider it not unlikely that there is a large percentage of inductive load. Therefore, I would advise against a two 110V USP experiment.

What could work, though, is to import a cheap 240V 50Hz USP from Europe. I tried this many years ago to emergency-power the 220V supply of a DEC minicomputer, and it worked flawlessly then. However, the concern re inductive load remains.

Regards,
Stephan

[zephyr9900]

It's unlikely that two 110V USPs would keep the 60Hz up closely enough to maintain the 180 degree phase shift.

I was just blue-skying, Stephan, not thinking it was a realistic idea... But I thought that UPS's were phase-locked to the line frequency, which itself is tightly controlled, and have electronically-generated waveforms which themselves are very accurately timed? Or do they count on electronic equipment having large-enough smoothing caps on the input to their own power supplies so that a discontinuity in switching isn't important?

The spindle motor is powered through a VFD, so would the UPS be seeing an inductive load?

I think that Just Gary's idea of a motor-genset is good if the Tormach is doing mission-critical work and you absolutely, positively can't lose a workpiece.

Randy (not a EE, not even on television)

[StephanWenger]

Hi zephyr9900,

Yes, USPs are phase locked to the line frequency (while there is one) to allow for a smooth transition when the line power goes off. After that, you may have some clock drift as the PLL does not receive stimulus any more. The amount of the drift can depend on many factors such as the design criteria, temperature of the circuitry, age, and others). In most cases, you will not notice the drift ever, because a little bit of a drift would not record even on a line frequency powered clock. When the power comes back on, the PLL re-synchs to it, compensating for any phase drift. Needless to say, the power amplifier in the output of the USP is slaved to the PLL.

However, when you have two USPs, you would need to have the two PLLs drifting the same way so not to loose voltage.

With respect to the inductive load: my confidence level here is not very high. However, my understanding of the basic layout of a VFD is that it has, in its input stage, something roughly comparable to a switched DC converter. The DC is then remodulated to variable frequency AC. Switched DC converters can have incredibly high inductive and capacitive loads, and have virtually no resistive loads. After all, they have to stash away energy at a very high rate during the times of peak voltage. This led me to the assumption of a possible problem.

That said, and thinking a bit more, the main use for USPs nowadays are computer power supplies, and those are all switched. Therefore, I guess that the USP should be able to handle switched loads quite well, and that will probably hold for the VFD just as well.

Regards,
Stephan

[keen]

I did a stupid thing the other day - pcnc was making chips - and spraying a worrying amount of coolant around - I was checking for leaks around the plugs - fiddled with them too check they were fully engaged - and must have cut the power for an instant - It was all downhill from there............Too painful to remember and relay the sorry sequence of events that followed.

[justgary]

Keen -

Ouch. But I must pry (I must, I must!) into the details.

As for the UPSs, I don't think that the frequency will shift very fast at all. In fact, I would guess less than a few milliHertz per second with even the cheapest UPS around. If one were to use them for Larry's stated purpose, e.g. to hold the machine up for a few seconds while the E-Stop relay (that he wants to install on the power line) trips, then they should work fine.

Even better would be to have your power relay issue a pause or stop command to the computer to help improve the odds that you save the part and can restart from where you stopped. After that, the machine is at idle, and the UPSs can certainly handle the load. I don't think that two to four seconds of loaded operation will blow the UPS, provided it is sized properly.

So the questions are:

1. Does the UPS synchronize to the line frequency? Stephan says yes, and it only makes sense for them to do so. It's easy to test with an oscilloscope by comparing the output of the UPS (after you unplug it) to the line it was plugged into. If you run two to get 220V, you must make sure that they are each on the proper line of the 220 breaker. In other words, you can't get 220 out of one 110 line and two UPSs.

2. Will a UPS handle the non-resistive load? I think so, but that also is easy to test with a fairly large motor (capacitive start, perhaps?) or a large linear power supply. We routinely run a medium-sized vacuum cleaner on a 1 KW cheap Chinese inverter (at a remote site), and it works just fine.

3. Do you need a true sine wave UPS? Probably not, since you are just running a switched DC supply anyway.

4. What size UPS would you actually need? 15A * 220V = 3.3KVA. I just checked, and Tripp Lite has a 5KVA (3750 Watt) unit for $2500. Maybe that's the answer after all, since you would pay nearly that much for two 2.5KVA 110 volt units anyway...

I wonder if the good folks at Tormach have considered this problem? Maybe they can recommend a solution.

Regards,

- Just Gary

[zephyr9900]

208V UPS--who'da thunk? Obviously not me! If it were a snake, blah blah blah... That's a brilliant idea. The 208V UPS Just Gary linked has an auxiliary 115V outlet that might suffice for the control PC and control circuitry (contactors etc.) that are supplied separately from the 208V in normal operation. Otherwise the scheme will need a separate 115V UPS for that stuff.

Randy

[keen]

Hi Just Gary - What happened? Actually it wasnt too bad as I have the braked Z model. Job survived. New cutter and complete re setup bla bla.

Many moons ago our ancestors evolved a brain 'trama memory protection circuit' (this happened to the survivers who explored too many cave bear caves) - Well that kicked in after the cutter broke so I forget most of the details after that - it really only troubles me when I'm asleep. I wake gasping and up trying to reach the stop button. Lesson well learned.

[MichaelHenry]

There was talk a while back, either here or on the Yahoo Tormach group, that you could finish a job after a power failure, by starting back up and resetting the axes back to 0 and then going back to your original work offset.

Haven't tried that myself yet, though.

Mike

[wildwestpat]

Hi Folks

I have a similar problem with the reliability of the public supply. I am using a 240 volt ups to protect the computer and have been thinking of extending this to cover the power to the mill. The mill uses a 40 volt 9 amp supply of the switch mode type from Motion Controls here in the UK. Hence I don't think the 700 watt UPS will mind and should achieve several minutes of running both the mill and the computer. My question is how long does Mach3 take to respond to the pause (manual tool change request) as this will halt the mill then the UPS has a reduced load to support for the duration of the powewr cut. My typical power outage is justunder the hour and my 700 watt UPS is not large enough to run my 2GHz P4 machine for that long.

Does any one know if March 3 can be instructed to go to sleep - after all it should not be doing anything whilst in pause - lost fingers during a tool change would not be a good idea?!

Has any one any experince of this as it appears to be necessary to get the following sequence of events:-

1. Public mains supply fails

2. UPS kicks in and powers both computer and cnc machine

3. UPS power detect used to start the pause Mach3 with the manual tool change comand (Ithink this is the correct name for the command but have not tried it as I am still fitting ballscrews.)

4. CNC machine becomes still and consumes little or no power from UPS

5 UPS continues to support computer.

6. In an ideal world Mach3 would then enter a sleep mode which would then permit the computer to enter its hibernagte state.

7. Public mains restored - operator intervention might be a good idea as it is possible that the power might only be on for a few minutes - time to boil up for a cuppa perhapse?

8. Computer backs out of hibernation
9. Mach3 back up and readt ro resume if the pause (manual tool change) completed command is given.


PLEASE comment as most of you will have a lot more practical experience.


Also suggest that the 240 volt issue could be inexpensively solved using an auto transformer - the old fashioned type with an iron core not the electronic matchbox type.

Hope the UPS can be made a viable solution as I hate wasting both tools and metal.

Good luck to all who are looking for a solution to the problem caused by mains outages however caused.


Stay safe and have fun with cnc.