PDA

View Full Version : Battery Powered CNC

Geof
08-08-2005, 11:22 PM
The title started as "battery assisted" but it transformed into battery powered. This is an idea that has popped into my head a couple of times reading posts by people who have set up rotary phase converters for CNC machines. It is necessary to size the rpc for the largest expected load and
normally with a CNC this is during spindle acceleration. For instance a Haas VF2 with a 10000 rpm spindle will pull 50 amps at 208V 3 phase (18kva) for around 300 - 400 milliseconds when accelerating but this drops back to around 40 amps (14kva) when it is roughing at a spindle load of about
120%, is down to around 10 amps or less (4kva) for finishing cuts and down to around 5 amps (2kva) when the machine is being reloaded. This means the full capacity of the electric service is being used for only a small fraction of the total operating time; averaged over time the machine needs about 5kva. The hour meters on the machine support this conclusion; Feed Cutting Time is approximately half the Cycle Time which itself is less than half the Power On Time so this gives the rough calculation that the average power needed is about one quarter of the full load power or about 4.5kva.

By now anyone who has done any reading on hybrid automobiles may have an idea where I am headed: Many CNC machines rectify the incoming power onto a DC buss which is used to synthesize variable frequency 3 phase AC or supply servo amps for the DC drives. A small amount of single phase AC is used to run cooling fans, operate solenoid valves and relays and run coolant or hydraulic pumps but this rarely adds up to more than the equivalent of 2 hp. On the DC buss there is a very small amount of averaging of power draw because there is a small amount of capacitative smoothing of the ripple coming out of the rectifiers but this represents only a few milliseconds of maximum power draw. However a bank of lead acid batteries on the buss could give several minutes of maximum power draw. Using the Haas system as an example; the DC voltage can range from 320 to 340V so acceleration draws 55 amps off the buss for about 400 milliseconds, roughing 34 amps and finishing 8 amps all plus or minus a bit. There would have to be twenty eight 12 volt batteries in the bank each about the size of a small automobile battery. These batteries would experience the same type of use as automobile batteries; brief surges of moderate to high current draw followed by a recharge period. Even a small automobile battery can provide 55 amps for several minutes.

Just slapping all the batteries in series across the DC buss might not be the brightest idea; having 340 volts available with almost no limit on the short
circuit current, and DC at that, could make a spectacular bang if something went wrong and be pretty hazardous for maintenance. The batteries would
have to be separated by fuses and some form of power-off isolation would have to be engineered. Also the simple bridge on the incoming supply would have to be replaced by a proper charging circuit so there would be a limit to the recharge current going into severely discharged battery and protection against overcharging when the machine was idle for a prolonged time. All of this is standard stuff and could probably be assembled almost off the shelf and all operate from a 30 amp 240 volt single phase (7.2kva) supply which would also have enough capacity to run the single phase AC requirements mentioned above. But if you threw into the mix a 2kva inverter creating 120V and 240V AC from the DC buss you would have a CNC running off its own uninterruptible power supply totally immune to incoming voltage fluctuations and able to be shut down in an orderly manner in the case of power outages; or you would be able to run for a long enough period of time to start and warm-up a 7500 watt diesel genset which could be useful in some situations

So has it been done?

ViperTX
08-09-2005, 11:31 AM
Well sortof....most extremely large machining shops use some sort of "supply" to maintain the line voltage. Just think of 30 or 40 15HP spindles starting up simultaneously....

So, some use large banks of batteries...similar to what alot of semiconductor manufacturing companies use, other have turned to these "refrigerator sized power sources", which basically have a huge rotor (300 lbs or more) floating on air bearings that make up for any line sags....anyway there is alot of information available on the web....also some Telco's are turning to these "rotors" instead of the banks of batteries they use.

RonL
06-05-2009, 01:06 AM
The title started as "battery assisted" but it transformed into battery powered. This is an idea that has popped into my head a couple of times reading posts by people who have set up rotary phase converters for CNC machines. It is necessary to size the rpc for the largest expected load and
normally with a CNC this is during spindle acceleration. For instance a Haas VF2 with a 10000 rpm spindle will pull 50 amps at 208V 3 phase (18kva) for around 300 - 400 milliseconds when accelerating but this drops back to around 40 amps (14kva) when it is roughing at a spindle load of about
120%, is down to around 10 amps or less (4kva) for finishing cuts and down to around 5 amps (2kva) when the machine is being reloaded. This means the full capacity of the electric service is being used for only a small fraction of the total operating time; averaged over time the machine needs about 5kva. The hour meters on the machine support this conclusion; Feed Cutting Time is approximately half the Cycle Time which itself is less than half the Power On Time so this gives the rough calculation that the average power needed is about one quarter of the full load power or about 4.5kva.

By now anyone who has done any reading on hybrid automobiles may have an idea where I am headed: Many CNC machines rectify the incoming power onto a DC buss which is used to synthesize variable frequency 3 phase AC or supply servo amps for the DC drives. A small amount of single phase AC is used to run cooling fans, operate solenoid valves and relays and run coolant or hydraulic pumps but this rarely adds up to more than the equivalent of 2 hp. On the DC buss there is a very small amount of averaging of power draw because there is a small amount of capacitative smoothing of the ripple coming out of the rectifiers but this represents only a few milliseconds of maximum power draw. However a bank of lead acid batteries on the buss could give several minutes of maximum power draw. Using the Haas system as an example; the DC voltage can range from 320 to 340V so acceleration draws 55 amps off the buss for about 400 milliseconds, roughing 34 amps and finishing 8 amps all plus or minus a bit. There would have to be twenty eight 12 volt batteries in the bank each about the size of a small automobile battery. These batteries would experience the same type of use as automobile batteries; brief surges of moderate to high current draw followed by a recharge period. Even a small automobile battery can provide 55 amps for several minutes.

Just slapping all the batteries in series across the DC buss might not be the brightest idea; having 340 volts available with almost no limit on the short
circuit current, and DC at that, could make a spectacular bang if something went wrong and be pretty hazardous for maintenance. The batteries would
have to be separated by fuses and some form of power-off isolation would have to be engineered. Also the simple bridge on the incoming supply would have to be replaced by a proper charging circuit so there would be a limit to the recharge current going into severely discharged battery and protection against overcharging when the machine was idle for a prolonged time. All of this is standard stuff and could probably be assembled almost off the shelf and all operate from a 30 amp 240 volt single phase (7.2kva) supply which would also have enough capacity to run the single phase AC requirements mentioned above. But if you threw into the mix a 2kva inverter creating 120V and 240V AC from the DC buss you would have a CNC running off its own uninterruptible power supply totally immune to incoming voltage fluctuations and able to be shut down in an orderly manner in the case of power outages; or you would be able to run for a long enough period of time to start and warm-up a 7500 watt diesel genset which could be useful in some situations

So has it been done?
Might be time to look at this some more, The system I put together might not be as large as needed for a full shop, but I was gearing for an electric drive system in a 1993 Ford Ranger. It finally got to heavy for the Ranger and the purchase of an F350 dual wheel came next. There was a number of evolutions between the two trucks, and now the parts are in other projects and still pretty much a dream.

What I did in relation to this thread came about as more of a project for learning. I took the 12 deep cycle marine batteries (100 amp hrs each) and wired them as a 1200 amp hr 12 volt battery, this group fed a 5,000 watt (10,000 peak) inverter 120 volts out @ 43 amps (85 peak). This voltage was passed through a 5,000 watt transformer to give 240 volts @ 21 amps which powered a Delta 2 hp planer motor, this motor turned a 10 KW generator head @ 3,500 RPM (for pure sine electricity).

The system worked well with just the flip of a switch (almost instant on, just like a power tool) and you had power up to 2 hp, I did a one time trial of powering a 225 Lincoln arc welder by adding a flywheel made from two 16" pipe flange caps centered on a 1-3/16" shaft. At about 2,400 RPM the flywheel carried the welder through almost 1 full 5/32 rod.

One duration test, (without the flywheel) the batteries were divided into two groups of 600 amp hrs. A sears 40 amp charger was plugged into the 10KW head and the system ran for just over 4 hours before the feed pack dropped to 11 volts and the inverter shut off.

There are other things in the works as we speak, I have decided to add compressed air to spin the flywheel(s) in an effort to relieve the inverter of the heavy amp load while using the welder.

Some of this might not make sense, but there is no particular goal other than learning and looking at what deficiencies might be filled by combining air and electric, and possibley the adding of a small gas engine (33cc) powered with natural gas. (long, slow charge of the batteries)

I have vane motors and piston compressors on hand, and the promise of a couple of scroll compressors (from a local AC tech) coming into the mix. I'm not far from starting the wiring and plumbing of this new project. I have quite a few 1.5 hp - 3.5 hp DC motors, a lot of playtime ahead, just hope the realities of life don't get in the way.

Maybe some pictures as I start progressing.

Later

RonL

amplexus
06-06-2009, 01:45 AM
very cool,
I have a power supply question; can you make a reasonably safe power supply using a ground fault interrupted breaker but no isolation transformer? just a big high current bridge rectifer and filter caps, feed direct from the mains. Use it to feed a synchronous H bridge pwm buck converter for variable voltage output. This would be cheap and flexible. plus the charge on the filter caps would kill you isolation transformer or not.

any thoughts
Amplexus Ender

Al_The_Man
06-06-2009, 10:18 AM
Basically you are describing the average VFD.
Athough in this case it is variable voltage-variable frequency.
Al.

Geof
06-06-2009, 10:32 AM
Basically you are describing the average VFD.
Athough in this case it is variable voltage-variable frequency.
Al.

Don't VFDs vary voltage with frequency to compensate for the change in inductance?

amplexus
06-06-2009, 10:33 AM
Frequency is fixed and synced to the stepper pwm frewuency only the duty cycle will change based on the minimum power needed at a given speed or load. I realize this has been done before I just can't find a good ap note. Also what about the gfi no isolation transformer issue? You have been doing this far longer then me (I'm 15) and you are ckearly one of the more knoledgable people on the forums (the smoking baby is a bit creepy though)
Amplexus Ender