DC alternative to phase converter?

Garwood · Jan 27, 2021

Instead of feeding a CNC spindle drive 3 phase AC would it be possible to feed the drive 340V DC power from a battery bank?

Odd scenario, but lets say I could get a pretty fancy large, late model HMC for pocket money. 40HP gogetamus spindle. MTB says it needs 75KVA.

I will eventually have the power to run it, but things take time. I run off a 60RPC currently and only have a 50KVA single phase transformer on the pole.

If possible, what might it look like to bypass the rectification portion of a 2010ish fanuc spindle drive and pack the pixies straight into the bus?

DDoug · Jan 27, 2021

I don't think the rectifiers waste much power, so it might be a moot point.
Several have posted on here before of running "too much machine" on too little
RPC by limiting the ramp up speed on that big spindle.

Controlling that D.C. power bank might be a problem.

Garwood · Jan 27, 2021

digger doug said:
I don't think the rectifiers waste much power, so it might be a moot point.
Several have posted on here before of running "too much machine" on too little
RPC by limiting the ramp up speed on that big spindle.

Controlling that D.C. power bank might be a problem.

The +/- acceleration ramp settings for Fanuc stuff don't really give you much room. Back it all the way down and it's still getting there in a big hurry. I run a 30HP spindle CNC mill currently off my RPC and it does OK for all but a few 100 degree summertime days when everyone is running AC and line power droops, but I can't run any other CNC's simultaneously. When the big spindle ramps the other machines brown out. Iirc the ramp parameters are 1 to 300 which corresponds to milliseconds. 300 milliseconds is not a lot.

If I add a 1200MM HMC I can get rid of a couple VMC's, but I would have to run it and atleast one other machine simultaneously.

Maybe a better way to look at this would be is there a way to build a battery based buffer for spindle ramping?

I was thinking 14 24V lead acid batteries wouldn't be too bad. I have a lot of DC switchgear.

Pathogen · Jan 27, 2021

DC Voltages at what you are speaking of are dangerous, as in you die if you make a mistake.

Just a Sparky · Jan 27, 2021

Best case in feeding a three phase bridge rectifier DC, two of the six diodes are going to be dissipating all of the heat... that's a pretty tall order.

Honestly I would just go for a bigger service. At 208-240V 3 phase you're talking about a 180~208 amp branch circuit. That kind of power is not something you screw around with casually. It's asking for trouble. If you have the space, funding and genuine need for a 40 horse machine then you can justify a 120/208 or 240V, 400-800 amp 3 phase service.

I wouldn't dare try to bypass the rectification on a drive that size unless DC bus connections are provided by the manufacturer and identified for that purpose. There could be active rectifiers or other funky business going on which won't take kindly to being bypassed or fed DC. Even so, DC wiring at that kind of voltage and power level is wicked dangerous if you don't know what you're doing. Use the wrong type of fuses or put them in the wrong place and something goes wrong? It will sit there and burn. And burn. And burn.

If I'm mistaken and this is just a garage project, you could try messing with the drive parameters... longer ramp times and reduced full-speed voltage - but in a commercial or industrial setting you need what you need and cheating your way around it is just asking for trouble.

See here:
https://youtu.be/DpQeDcEpEn0?t=282

dkmc · Jan 27, 2021

Buy a nice used cheap Diesel Generator till you can upgrade grid service.

Just a Sparky · Jan 27, 2021

dkmc said:
Buy a nice used cheap Diesel Generator till you can upgrade grid service.

One thing to consider is the tendency for generator sets to not respond well to step loads. They generally don't tolerate wild changes in load characteristics (as are typical with CNC machines) as well as the utility can. Pushing a 75kVA genset from 10% to 100% in a fraction of a second may or may not cause enough of a voltage and frequency sag to fault the drive.

Examples:
https://youtu.be/tYmkKvohDSA?t=18
https://youtu.be/UEJ9ffPpvD8?t=17

Strostkovy · Jan 27, 2021

I can smell the arc flash from here

dkmc · Jan 27, 2021

Just a Sparky said:
Pushing a 75kVA genset from 10% to 100% in a fraction of a second may or may not cause enough of a voltage and frequency sag to fault the drive.

Well, wait. Who said a 75KVA set? He needs a 150KVA, Detroit 2 stroke powered for best torque response.

mmurray70 · Jan 27, 2021

Garwood said:
Maybe a better way to look at this would be is there a way to build a battery based buffer for spindle ramping?

I often wondered about this as well. How about doubling or tripling the amount of capacitors? Would this actually store any useful amount of power to help with spindle ramp or would it be negligible compared to whats needed?

Im in market for a new lathe, leaning towards a Doosan with Fanuc control and I fear I wont be able to slow the ramp enough for my needs as well.

Garwood · Jan 27, 2021

Any chance for a nuts and bolts response of "it would probably take this..." kind of thing.

I'm not the only guy running big machines on a single phase service.

We know the bus voltage is standard.

I fully appreciate the safety concerns, so what would it look like to do it safely?

There's a lot of things I do that would kill somebody if they did it unsafe.

thermite · Jan 27, 2021

Challenge here is that a big CNC critter actually NEEDS rapid dynamic response whilst "at work".

Ex; An SSD drive for a MANUAL lathe has an on-PCB ramp minimum set-point of two whole seconds. AND NOT 300 mS. Yes, terminals are "there" to over-ride. But we do not for start up. And never whilst "in the cut". Only for emergency braking. If-even then.

So War ONE "coastal defense" submarine / U-boat tech is about right for the POWER asked for.

But the need of rapid response takes you out of forklift, marine/RV and golf cart battery into "starting service" battery flavours that can deliver massive current on-demand, even if not for very long.

Best-case? The rapid dynamic response technology you need is already built-in to a Tesla high performance electric vehicle.

Get the picture as to what it might cost to make any useful gain?
"Not cheap" if somehow you missed that.

More better to partially cripple the CNC rig with longer ramping.

Above my present pay-grade if that means the code that says "make THIS" has to be altered to EXPECT very much slower dynamic response.

But I'd expect that...

Should at least also be the CHEAPEST way forward as well as the safest. So long as the modest reduction in throughput is OK.

Usually it IS OK. In a way, you are just dialing BACK the overall performance level to that of OLDER generation CNC with lesser overall power and longer cycle-times according.

BFD. They made plenty-much chips.

Find a way to stretch the ramp. Run it that way.

Leave the battery bleeding edge to others.

2CW

Phil in Montana · Jan 27, 2021

There is a lot of control that's needed to be done to run the motor, down and dirty yes a veriac and a rec and away you go but .....Phil

9100 · Jan 27, 2021

............................

JST · Jan 27, 2021

Bill, I screwed up your post.....sorry..... if you can repost, go for it.

jz79 · Jan 27, 2021

you should first look at the type of the motor/drive combo there is on that machine, it is entirely possible your idea is completely incompatible with the design if it uses something like thyristor 3 phase rectifier to generate the dc to run a brushed motor for instance

and even if it is a VFD type of drive, as suggested before, if the drive doesn't support such a feature - it would simply not be worth trying to dig into it and try to macgyver something together there, too easy to blow something up... it would get expensive very quickly

a generator with a MASSIVE flywheel would make much more sense...

thermite · Jan 27, 2021

9100 said:
DC does not have some evil characteristic that makes it that much more dangerous. .........

Bill

Actually, it sorta does have.

With AC, people are just conductive enough that AC WILL tend to be conducted nearer the skin, and might not get a heart-stopping current down deep.

DC will flow completely according to resistance, and will more easily produce enough deep current to do it.

Also, DC has more tendency to cause you to "freeze onto" whatever you are touching (and get a more thorough shock) than AC does, although both can do it.

Yazzz Dinosaur Current is "stick and fry" - much above the neg-48 VDC -> plus charger "float" we used for a hundred years or so of Telco practice...

Massive power levels are still massive.

The 8800 VAC arc a 20 Ton GI rough-terrain crane drew EVER so briefly, alert operator triggering emergency drop of his boom.. need never have happened - his assistant hadn't felt the need to smoke a joint, opposite end, sitting on his ass, back against the deeply set-down push-blade. Instead of being where he was MEANT to be.. watching for overhead obstacles. Such as power lines.

Part of my job as the investigating Officer was to meet with the coroner and review the autopsy.

AC went MORE than "deep enough". Read like a menu in an Argentine restaurant as to what internal organs were baked, boiled, broiled, fried.. charcoaled.

They'd even managed to keep him alive for a whole hour and a half. Military docs do their best, regardless of how hopeless. A higher power calls game-end. T'was ever thus.

This s**t ain't LIKE rolling yer own "Pee Cee", stereo gear or robotics.

Massive power levels are still massive.

Hardplates · Jan 27, 2021

Garwood said:
The +/- acceleration ramp settings for Fanuc stuff don't really give you much room. Back it all the way down and it's still getting there in a big hurry. I run a 30HP spindle CNC mill currently off my RPC and it does OK for all but a few 100 degree summertime days when everyone is running AC and line power droops, but I can't run any other CNC's simultaneously. When the big spindle ramps the other machines brown out. Iirc the ramp parameters are 1 to 300 which corresponds to milliseconds. 300 milliseconds is not a lot.

If I add a 1200MM HMC I can get rid of a couple VMC's, but I would have to run it and atleast one other machine simultaneously.

Maybe a better way to look at this would be is there a way to build a battery based buffer for spindle ramping?

I was thinking 14 24V lead acid batteries wouldn't be too bad. I have a lot of DC switchgear.

What control? On a 16i/18i you can set it up to 4 seconds per thousand RPM. I also have 16/18A manuals as well as some others I can take a look in depending on what it has.

mmurray70 · Jan 27, 2021

Hardplates said:
What control? On a 16i/18i you can set it up to 4 seconds per thousand RPM. I also have 16/18A manuals as well as some others I can take a look in depending on what it has.

is this for lathes too? Any idea it this will work on a Doosan Lynx 300? No number listed. Just says Fanuc I control, or Fanuc I plus control.

Strostkovy · Jan 27, 2021

Has nobody mentioned how much harder it is to quench a DC arc? If a breaker or overload trips in that machine due to a fault, the current won't stop.

Also, even though most switch mode power supplies can run on DC, they won't start because they typically use capacitive droppers to charge the control circuitry.

DC alternative to phase converter?

Diamond

Diamond

Diamond

Hot Rolled

Hot Rolled

Diamond

Hot Rolled

Titanium

Diamond

Stainless

Diamond

Diamond

Stainless

Diamond

Diamond

Stainless

Diamond

Stainless

Stainless

Titanium

Similar threads