(AC power) Converting single-phase into 3 phase: What are my options?

[cvanc]

Hi all-

I need to run a machine that needs 3 phase power in a building that only has single phase.  So I need to manufacture roughly 10KVa of 208v, 3 phase, Wye power.  With a decently clean sine wave as well!  How do I do that?

Here's the details:

INCOMING BUILDING POWER
The building has the usual split 240v single-phase coming in, which becomes a pair of 120v rails in the breaker panel 180° out of phase with each other (in other words, the totally generic typical single-phase power provided here in the U.S.)

THE MACHINE NEEDS
208v Wye 3 phase power; the circuit breaker at the AC inlet inside the machine is 30 amps, but I am told max draw on any leg is never more than 20

THE QUESTION
How do I get 'from here to there'?  In the old days one option was a motor-generator set, which is nothing more than a single phase motor directly driving a 3 phase generator.  I've actually used these in decades past and while they absolutely work they are loud and smelly and hot and did I mention loud?  This approach just doesn't work here; it's an office setting and there will be people in the same room with the machine.  I'd have to mount it outdoors (roof? alley?) and put it in some kind of custom built soundproof, theftproof enclosure.  Ugh!

Surely there's a solid state answer this this kind of problem, but I have no experience in that arena.  I'd like to lean on those who've BTDT to see what might exist and get to a discussion of specific brands of gear that are respectable, or to be avoided.

"Why do I need a clean sine wave?" "Why do I need Wye?"
'Cuz this box is ANALOG and FUSSY, that's why.  And it has like 22 different power supplies inside, some of which are linear and some of which are switchers.  I need to baby it and give it exactly what it expects to receive.  This thing wants Wye with at least a half decent sine wave and I will not even TRY to rock that boat.

I'm all ears folks, I've got maybe 2 weeks to propose a solution so there's a bit of time to kick this around but not a lot.  Thanks so much for any insights you can offer.

[NiHaoMike]

Would it be possible to rewire it for single phase?

[sokoloff]

THE QUESTION
How do I get 'from here to there'?  In the old days one option was a motor-generator set, which is nothing more than a single phase motor directly driving a 3 phase generator.  I've actually used these in decades past and while they absolutely work they are loud and smelly and hot and did I mention loud?  This approach just doesn't work here; it's an office setting and there will be people in the same room with the machine.  I'd have to mount it outdoors (roof? alley?) and put it in some kind of custom built soundproof, theftproof enclosure.  Ugh!

Surely there's a solid state answer this this kind of problem, but I have no experience in that arena.  I'd like to lean on those who've BTDT to see what might exist and get to a discussion of specific brands of gear that are respectable, or to be avoided.

The old solution you describe above is called a rotary phase convertor. The newer, solid state, solution is a variable frequency drive (VFD). I've only used them to drive mill motors where I didn't care (and didn't check) on the cleanliness of the power.

So, look into that, but VFD/variable frequency drive is at least a keyword for you to search on...

[Floyo]

If its really fussy about clean power a VFD as is wont cut it, they just output a modulated square wave, and are really geared towards driving motors.
What you could look into is an AC source, they basically create a super clean grid on their outputs, fully programmable in voltage and frequency.
Since their outputs are isolated from the grid they can probably (not standard) be wired in parallel on the primary side, and their outputs in 3 phase.
One brand to look for is Chroma http://www.chromaate.com/product/list/ac_power_source.htm
Edit: here is another https://pacificpower.com/products/afx-series/

Having said that, AC sources are "call us for a quote" levels of pricey

If we had more info on the machine another option could well be possible since, aside from motors, not a lot needs native three phase.

[Jeroen3]

THE QUESTION
In the old days one option was a motor-generator set, which is nothing more than a single phase motor directly driving a 3 phase generator.  I've actually used these in decades past and while they absolutely work they are loud and smelly and hot and did I mention loud?  This approach just doesn't work here; it's an office setting and there will be people in the same room with the machine.  I'd have to mount it outdoors (roof? alley?) and put it in some kind of custom built soundproof, theftproof enclosure.  Ugh!

There won't be a silent solution for this. Either you use a solid state converter, which needs cooling fans. Or you use an rotary converter, which needs cooling fans.
Putting it somewhere else is not that hard, it just requires some work.

If you cannot rewire the machine, and it definitely needs a clean power source, then I think the rotary converter in some other room is the only way.

An office usually has a server room with AC. How about there?

Did you check if the building connection capacity large enough?

[cvanc]

Would it be possible to rewire it for single phase?

No.  The machine is old and was super expensive back in the day and it's failure-prone and hard to get parts for and fragile.  Plus, any problems it has going forward will be my problem, so no internal modifications of any kind will be done, period.

If its really fussy about clean power a VFD as is wont cut it, they just output a modulated square wave, and are really geared towards driving motors.
What you could look into is an AC source, they basically create a super clean grid on their outputs, fully programmable in voltage and frequency.

I don't need some perfect super-low distortion sine wave but I sure don't want a coarsely stepped square wave.  I was hoping to find products that live in the middle ground between those extremes but maybe 'slightly distorted sine wave generators' don't exist?

If we had more info on the machine another option could well be possible since, aside from motors, not a lot needs native three phase.

Please see above, no internal modifications allowed, we have to treat it like a black box and accept it as it is.

Having said that, AC sources are "call us for a quote" levels of pricey

I was afraid of that.

Thanks for your insights; much appreciated.

[cvanc]

If you cannot rewire the machine, and it definitely needs a clean power source, then I think the rotary converter in some other room is the only way.

You may well be right.  On the positive side, every now and then you can get a good used one ridiculously cheap on eBay (shipping can be $$$, though).

Did you check if the building connection capacity large enough?

Plenty, I've got two panels.  One has a 400a main and the other has a 200a main.  Both are currently fairly lightly loaded, especially the 400a one.

[langwadt]

Would it be possible to rewire it for single phase?

No.  The machine is old and was super expensive back in the day and it's failure-prone and hard to get parts for and fragile.  Plus, any problems it has going forward will be my problem, so no internal modifications of any kind will be done, period.

If its really fussy about clean power a VFD as is wont cut it, they just output a modulated square wave, and are really geared towards driving motors.
What you could look into is an AC source, they basically create a super clean grid on their outputs, fully programmable in voltage and frequency.

I don't need some perfect super-low distortion sine wave but I sure don't want a coarsely stepped square wave.  I was hoping to find products that live in the middle ground between those extremes but maybe 'slightly distorted sine wave generators' don't exist?

VFD outputs pwm but once it is "filtered" by the motor it is a pretty  good approximation of a sine wave

https://youtu.be/U71WrUihzO0

[drussell]

THE MACHINE NEEDS
208v Wye 3 phase power; the circuit breaker at the AC inlet inside the machine is 30 amps, but I am told max draw on any leg is never more than 20

What kind of machine is it?

If there is just one motor, a VFD is definitely the way to go.  They are readily available, inexpensive lego-type pieces these days.  It gets much trickier if the equipment has tons of different motors, etc. thus to make a realistic recommendation, we need to know exactly what it is you're needing to power... 

[Benta]

I'd go for a 15 kVA AC inverter with line filters. It doesn't need to be a VFD, a simple inverter will do the trick.

[capt bullshot]

A VFD rated for 10 kVA (kW) output usually has a 3-phase input. At least here in Europe.
One can add a sine filter to the VFD to get a decent sine wave output voltage.
There's one caveat: The VFDs output doesn't have a neutral, so if your unit requires a neutral, this can't work.

[Floyo]

I don't know of any "slightly distorted" options off the shelf, other than what's already been mentioned. A VFD with a filter could possibly work, but the filter also needs to account for the common mode voltage generated by the drive.
Another option could be (which is I think what Benta is suggesting) a solar inverter capable of islanding mode, but that'd need a DC input.

[drussell]

A VFD rated for 10 kVA (kW) output usually has a 3-phase input.

The vast majority of VFDs a just rectified inputs to get the main DC rail input.  Yes, larger ones have a three phase rectifier available but virtually every one will allow you to use just two of them from a regular single-phase 240 volt input.  You just leave the third one disconnected.  Sometimes there is a setting or a jumper to disable any phase-loss detection from being triggered by the (intentionally) missing input phase.

This application is very common.

[cvanc]

What kind of machine is it?

A motion picture film scanner.

If there is just one motor, a VFD is definitely the way to go.

Everybody, please lose all thought of motors.  There aren't any motors.  (well, there are, but they all run on DC)

What it has, internally, is about two dozen single phase (120v) power supplies of wildly varying types, power levels, and vendors, all poorly distributed across the three hot phases, and all returning to a single shared neutral.  This is why it has to be Wye power (5 wires: 3 hots, one PE, and a real copper neutral).  Any 3 phase topology that does not have a real copper neutral at the midpoint of the 3 legs will not work.  Plus they did a bad job distributing the load around the circle and even at idle the neutral current is several amps.  So Wye it is.

This is why I really, really can't consider internal changes.  It's a black box for the purposes of this discussion.  Thank you all for your suggestions and ideas.

[Floyo]

If there aren't any native three phase loads, why not tie the phases together into a single one?
Otherwise so far only the AC source option remains, it has a real neutral.

[Benta]

Why do you guys keep talking about VFDs?
The OP needs a simple 3-phase fixed-frequency inverter. They're off-the shelf parts, the line filters as well. And yes, there are types with neutral for a wye connection.
The input rectifiers can be 3-phase, split-phase or even single-phase depending on model.
There's nothing magical about this.

[Ian.M]

If you are *100%* certain there are no PSUs inside it with 208v phase to phase input or three phase input, then it should run with all phases strapped in parallel externally and fed 120V, 60A.

How you get that 120V 60A is a different matter - I suspect a transformer off a 240V 30A split phase feed would be the best option.

[drussell]

Well, if it isn't a motor load, VFD is out.  You would need some kind of three phase inverter, not the nice simple, cheap VFD.

Personally, since all the loads are 120, I would open it up and re-wire the one third of the stuff that is on the blue phase to be half on the black, half on the red and be done with it.  Easy peasy, re-wired for 240.  (120-N-120, with neutral, that is, of course, not really 240.)

I could build you one of our APS-3720s, I'm in the process of building another one anyway (a variable frequency model, they're using it for 3-phase 400 Hz) for an aircraft company that builds specialized stuff for C-130s for the US military but they're about $5000, though a fixed frequency version at 60 Hz, maybe sans some of the metering, of one could potentially be a bit less expensive, I suppose. 

[drussell]

If you are *100%* certain there are no PSUs inside it with 208v phase to phase input or three phase input, then it should run with all phases strapped in parallel externally and fed 120V, 60A.

Except that he would have to beef up the neutral internally since everything would now go through the neutral instead of just the difference left over that didn't cancel between the three phases.

[schmitt trigger]

What Floyo and Ian are suggesting is THE best option.
I would go one step further and tie 2 inputs to one phase and the remaining input to the other phase. And the neutral where it belongs of course.

This will work because the line to neutral voltage in a 208v Wye is 120v. Exactly the same as in 240v split phase.

[SeanB]

You have a 120VAC split phase supply, which then gives you 2 hot wires and a neutral. Thus simply take the highest loaded phase and connect the one hot to that, and tie the other 2 to the other phase. Breaker at 40A, double pole, and the only neutral current will be the difference again, and as it is not a balanced load anyway it has a beefy enough neutral. Nothing in there is actually 3 phase, and even stuff that is 207VAC and across phases will generally be happy with 240VAC anyway.

No need for anything other than a look to be absolutely sure the thing has no 3 phase motors ( 3 phase motors run via an inverter or VFD will be happy here anyway, on 208V and 240V they get about the same DC  bus voltage anyway) DOL, just regular mystery 120VAC supplies, and the switchers will be happy at 120-240VAC in most cases, and linears anyway will be 120VAC or have a primary tap changer if they are critical in any way.

[tpowell1830]

Having had to deal with putting a machine shop with 3 phase motors in a garage that only had single phase power (the 240 VAC phase are 180° apart, not 120°) and 3 phase was not available in the area, we used the rotary converter method as Jeroen3 suggested. In your setting this will not work, since it is loud and you don't get the common neutral.

You are left with a very expensive option as was suggested, or you have to call the power company to install a 3 phase transformer at your shop, if you have 3 phase available in your area. In my area, getting 3 phase transformer installed is not very expensive if you use it for a large portion of your power requirements. You do not have that, since you only need it for this single device.

Tough predicament, but my first move would be to call the power company and tell them your needs and see what they can do to get your 3 phase power. If deemed too expensive or not available, then you only have the other 2 options; i.e. one very loud and the other very expensive. Granted, the power company option may be fairly expensive after having to get a separate 3 phase panel and the wiring required as well, but you would then have the 3 phase available from that point on in case 1. You added another of these machine or 2. you added other 3 phase equipment. The upside with this option is that the 3 phase transformer side would be true 3 phases at 120° with a common neutral and the transformer would be maintained by the power company plus no noise.

Hope this helps...

[drussell]

Installing 3-phase power would be silly overkill in this instance.  cvanc simply needs to double and triple check that everything really is 120 volt and then supply it from both sides of the available split 240.

As suggested multiple times above, this can be as simple as wiring two of the phases to one side, provided that the imbalance doesnt't cause an overloaded neutral, though that is unlikely as long as it is on both sides of the 240, not just all on a single 120 (which would be silly anyway.) 

If the imbalance will be huge or the OP wants to run from a smaller breaker, I really can't see why it would be too difficult to move even just a couple of the supplies from one phase to the other to balance things out, even if he doesn't want to move everything from the third phase.

I still think it would be pretty easy to move everything from the blue phase (or whatever one was easiest) over to the other two, but surely it can't be too difficult to just strap two of them together at the supply and move just a couple of the supplies from the now more heavily loaded "double side" over to the other phase.

I can't see why that would be very difficult.

Of course, I haven't seen any photos of the actual unit's internal horror show either, though. 

[Benta]

Well, if it isn't a motor load, VFD is out.  You would need some kind of three phase inverter, not the nice simple, cheap VFD.

Huh? An inverter is simpler than a VFD.   

[tpowell1830]

Installing 3-phase power would be silly overkill in this instance.  cvanc simply needs to double and triple check that everything really is 120 volt and then supply it from both sides of the available split 240.

As suggested multiple times above, this can be as simple as wiring two of the phases to one side, provided that the imbalance doesnt't cause an overloaded neutral, though that is unlikely as long as it is on both sides of the 240, not just all on a single 120 (which would be silly anyway.) 

If the imbalance will be huge or the OP wants to run from a smaller breaker, I really can't see why it would be too difficult to move even just a couple of the supplies from one phase to the other to balance things out, even if he doesn't want to move everything from the third phase.

I still think it would be pretty easy to move everything from the blue phase (or whatever one was easiest) over to the other two, but surely it can't be too difficult to just strap two of them together at the supply and move just a couple of the supplies from the now more heavily loaded "double side" over to the other phase.

I can't see why that would be very difficult.

Of course, I haven't seen any photos of the actual unit's internal horror show either, though. 

As far as installing 3 phase power being silly, I don't think you actually read his post. He is saying that the internals of his device is a black box and he is not going to modify it to balance any loads or re-arrange the internals in any way because it is very sensitive. In other words, he doesn't want to modify the stuff in the box.

I stand by my advice to get 3 phase power if within his cost structure, as I clearly implied. If getting 3 phase power is not within his cost structure then he may ignore my post entirely, up to him. It would still be appropriate to call the power company to see what they can do for him.

[drussell]

Huh? An inverter is simpler than a VFD.   

VFDs rely on the fact that they are driving a motor to make the circuitry simple and inexpensive.

A decent quality VFD in that power range is a couple hundred bucks.

Show me an inexpensive (somewhere even remotely cost competitive with a VFD) 3-phase-output inverter that takes wall power for an input or with a cost effective power supply to run it from whatever DC input voltage it requires.

[cvanc]

If you are *100%* certain there are no PSUs inside it with 208v phase to phase input or three phase input, then it should run with all phases strapped in parallel externally and fed 120V, 60A.

If there aren't any native three phase loads, why not tie the phases together into a single one?

I've supported these machines for 20 years.  Hell, there isn't enough therapy in the world to get them outta my head 

I'm 100% certain there are only 120v loads inside, and nothing 3 phase, and nothing 208v (leg-to-leg).

But here's the problem:  People have tried what you suggest before and the internal neutral burns up.  It's not sized for this and does not tolerate it.  What a mess THAT was, and here's why...  the power distribution panel in these things is literally the first assembly installed in the factory, and it's held in by really long bolts coming up from the underside.  If you need to pull it you have to lay the machine on its' back.  It weighs  sixteen hundred pounds.  And it's full of fragile, totally obsolete, delicately aligned optics.  Doing any kind of maintenance on the power distribution panel is the single most difficult task this design has.  It's just a non-starter.

It gets worse.  Before it burned up (it made it for 2 or 3 weeks) the output images had rolling humbars that nobody could figure out.  The power rewiring had done something to screw up the analog front end.  God only knows what.

Personally, since all the loads are 120, I would open it up and re-wire the one third of the stuff that is on the blue phase to be half on the black, half on the red and be done with it.  Easy peasy, re-wired for 240.  (120-N-120, with neutral, that is, of course, not really 240.)

You have a 120VAC split phase supply, which then gives you 2 hot wires and a neutral. Thus simply take the highest loaded phase and connect the one hot to that, and tie the other 2 to the other phase.

Hmmm, this is something that I don't think anyone has tried before but given the history of the internal neutral failing with increased load I have to admit my doubts the box would be OK.  And there's always the hum-bars-in-the-picture thing to worry about...

Tough predicament, but my first move would be to call the power company and tell them your needs and see what they can do to get your 3 phase power.

Oh absolutely, this is very clearly the best and safest answer.  The utility was contacted weeks ago but a response has not come yet.  It remains my best hope but time moves on and we have to come up with an answer.

I've yet to visit the site and the folks there are not technical.  I tried to get them to tell me what's on the poles in the alley but no luck yet.  (I asked for pictures of the poles and they sent photos of the bottoms of the poles where they go into the ground - sigh)

For all I know 3 phase is 30 feet away from them and this is all no big deal to "do the right way".  But it's not an industrial area, so who knows.

Way back when I knew someone with this same issue and they found out the nearest pole with 3 phase on it was over a mile away.  The utility was happy to quote pulling the line to their address, however they wanted 35 grand to do it.  In the end it was decided to give the utility the money but boy did the people who chose to lease that building catch hell.

I really appreciate all of you for engaging in this conversation.  It's never too late for me to learn.  Thank you.

[Benta]

A decent quality VFD in that power range is a couple hundred bucks.

I must be living in parallel universe. A 10...15 kVA VFD for a couple hundred bucks? Do show.

[Brumby]

It's pretty clear the only acceptable solution is to provide clean 3 phase power that can handle poorly balanced loads across the phases - which is what the OP asked for in the first place.

To my mind, the motor-generator solution is the simplest.  The sine wave output should be pretty good.

Second up is 3 inverters at 120º phase difference, on a common neutral.

Last is a 3 phase generator set ... but that's a whole new level of noisy.

[drussell]

If you are *100%* certain there are no PSUs inside it with 208v phase to phase input or three phase input, then it should run with all phases strapped in parallel externally and fed 120V, 60A.

If there aren't any native three phase loads, why not tie the phases together into a single one?

But here's the problem:  People have tried what you suggest before and the internal neutral burns up.

NO!  You *absolutely* cannot do that!

That takes the neutral from carrying 0-30 amperes to carrying the full load of everything in the device (up to 90A in this case.)

That is a common mistake made by people installing a 120-volt inverter system.  You have to go back and re-wire anything that shared a neutral between phases to have their own separate wires.

Personally, since all the loads are 120, I would open it up and re-wire the one third of the stuff that is on the blue phase to be half on the black, half on the red and be done with it.  Easy peasy, re-wired for 240.  (120-N-120, with neutral, that is, of course, not really 240.)

You have a 120VAC split phase supply, which then gives you 2 hot wires and a neutral. Thus simply take the highest loaded phase and connect the one hot to that, and tie the other 2 to the other phase.

Hmmm, this is something that I don't think anyone has tried before but given the history of the internal neutral failing with increased load I have to admit my doubts the box would be OK.  And there's always the hum-bars-in-the-picture thing to worry about...

That would work from a power standpoint and as long as you keep the phases somewhat balanced, your neutral would be fine since it only carries the difference (unlike just jamming them all together, which absolutely will not work!  Egads!  That is guaranteed to burn up the neutral!  )

Unfortunately this does not help you if physically cannot get to any of the wiring.

The hum bars would very likely be due to the fact that the neutral was running at 3x its maximum capacity and probably many, many times what it would normally see when running on three phase where it only sees the difference.  This would cause all sorts of havoc and is exactly what I would have expected to see in that ill-fated scenario.    There would be severe voltage drop on something that was supposed to be essentially unloaded.

Tough predicament, but my first move would be to call the power company and tell them your needs and see what they can do to get your 3 phase power.

Oh absolutely, this is very clearly the best and safest answer.  The utility was contacted weeks ago but a response has not come yet.  It remains my best hope but time moves on and we have to come up with an answer.

If it is available and within budget then obviously that is certainly a valid method. 

I only say it is "silly" from a cost perspective compared to the minor modifications (if any) required to run it properly on 120-N-120...  i expect it would be cost prohibitive, though I do not know your budget, and like you say, maybe there is 3-phase available right there.  I suspect not, though, since they have a 400 amp panel that is only single phase.  Usually if you have a location that is getting a service of that kind of size, it would be 3-phase by default if it were available nearby.

Every area is different, though! 

[drussell]

I must be living in parallel universe. A 10...15 kVA VFD for a couple hundred bucks? Do show.

Ok, fair enough, probably more like 500-750 bucks for a new one in that power range.  The very common 10 HP-class ones are a couple hundred bucks, plus there is always the used market if cost were an issue.  VFDs are incredibly common units now.

I still doubt you'll find a three-phase inverter or a set of three that is stackable at 120 degrees for anywhere close to what a VFD costs.  180 degrees is a pretty commonly available stacking configuration but 120 is very rare, as are dedicated 3-phase units.  (It could solve the OP's problem, of course, though, if the cost is tolerable.)

[dmills]

Old Rank cinetel?

Here is a left field idea, grab three big audio amps (I have a certain fondness for old Crown MA5000VZs for this sort of stunt), add suitable signal generator and some wiring, job done.
You can have three phase of almost arbitary waveform and frequency, and could even lock it to an external SPG if chasing black and burst is a thing in your workflow.

Or do it the sensible way, motor inverter (or VFD, don't really matter) and a modest delta-star transformer (to get you the neutral), you can get filters for the outputs of the better sort of inverter drives to make decent sine waves, and the delta-star transformer will provide useful filtering anyway (as well as soaking up any triplen coming back from the scanner).

Regards, Dan.

[drussell]

Old Rank cinetel?

Here is a left field idea, grab three big audio amps (I have a certain fondness for old Crown MA5000VZs for this sort of stunt), add suitable signal generator and some wiring, job done.
You can have three phase of almost arbitary waveform and frequency, and could even lock it to an external SPG if chasing black and burst is a thing in your workflow.

Heh...  That "left field idea" is not nearly as crazy as you think. 

That's exactly what our APS-3720 is based on.  Three modified audio power amplifiers. 

The use of a commodity item as a building block allowed us to make them available for a fraction of the cost of similar units.  I originally designed it as a cost effective alternative to using an aircraft APU for testing their door systems in the hangar.  The customer was pleased with the cost-performance ratio and have ordered multiple units since I did the original design.  I think the reason they ordered the most recent one was they sent the one that was in Florida to some military base in The Philippines or something. 

This is the original prototype unit, from almost a decade ago:



The back of the rack:





Single phase in operation during testing (things like boiling water in a kettle using 400 Hz power sounds funky.  Like you're on a plane.  )

[dmills]

Yep, that would get it done.
My use case was a US touring band that had specified 120V, but forgot to tell us that they were driving a tonewheel organ and the 60Hz mattered (it usually doesn't much!).

Apparently we were the ONLY venue on that tour where the thing was in tune!

Regards, Dan.

[duak]

I would look in to a Phase Perfect converter: http://www.phasetechnologies.com/products/phase-perfect-240v/overview

These things take the center tapped single phase AC in and generate a third phase using a rectifier, bulk capacitor and PWM driver with output filters.  The FP websire has papers describing how it works.  I don't believe the input neutral can be used as the neutral for the output wye so you would also need an isolating delta-wye transformer to re-reference the load neutral.

These things are expensive though.  Another way is to get a big 3 phase motor, a few times the size of the load to act as a rotary phase converter.  Most 3 phase motors in North America are wound with 6 windings that can be connected in parallel for 208-230 V or series for 440-460 V.  As a balanced wye converter, three windings would be connected in a wye with the neutral brought out for the secondary or load.  The other three windings are usually connected as a wye already.  Two of these windings are driven directly from the single phase lines while the third is driven by a phase shift capacitor from one of the lines.  An additional capacitor may be needed to get the motor up to speed.  By taking advantage of the two sets of windings, the motor can serve double duty by also acting as an isolating transformer.  Most of the noise is from the cooling fan.  A good motor, especially 1200 or 900 RPM, is quite quiet.

Cheers,
 

[edpalmer42]

Just for comparison, I've heard of one brand of rotary converter, AmericanRotary.com .  They have prices on their web site.  It looks like their AD-15 model (21 amps) would do the job.  Price is $1649.

If you wanted to rewire the unit, but can't due to the overloaded neutral, is there any way that you could replace the existing neutral with a bigger one?  It doesn't have to be pretty - I'm talking drills, hole saws, hammer & chisel, dynamite - you get the idea.  Maybe run the new neutral external to the machine, but inside a conduit for protection.  You just want to get from the point where the power connects to the other end of the neutral cable.

Ed

[Beamin]

Is this some sort of milling machine with CNC and motor controllers? Hence all the different power supplies? IF you have money to spend on this MrCarlsons lab of youtube fame  has a real business doing/making power supplies for CNC machines. Hes out of Canada. and has a website.

[IanB]

Is this some sort of milling machine with CNC and motor controllers? Hence all the different power supplies?

No it isn't. Why not read the thread instead of doing a drive-by posting? 

[cvanc]

Again, thank you everybody.

Best word of the thread so far?  "Dynamite" 

[nctnico]

At 10kVa getting 3 phase power to the machine is the best option. 10kVa translates to around 100A @110V single phase.

[6PTsocket]

That was my thought, too. The original was 3 phase to spread the load but the OP wants to run it on single phase that has to carry the whole load. I see no advantage to splitting the source into 3 phases that all have to come from his single phase source that must supply all the power.

If there aren't any native three phase loads, why not tie the phases together into a single one?
Otherwise so far only the AC source option remains, it has a real neutral.

Sent from my SM-G900V using Tapatalk

[IanB]

If there aren't any native three phase loads, why not tie the phases together into a single one?
Otherwise so far only the AC source option remains, it has a real neutral.

That was my thought, too. The original was 3 phase to spread the load but the OP wants to run it on single phase that has to carry the whole load. I see no advantage to splitting the source into 3 phases that all have to come from his single phase source that must supply all the power.

Another drive-by posting.

Because if you tie all the phases together the neutral will burn out. See higher up the thread where this was discussed.

Sent from my SM-G900V using Tapatalk

Why do we care about this? Please tidy up your litter and don't keep dropping it everywhere.

[BrianHG]

Build your own 1 phase to 3 phase rotary converter.  You will have your true real 3 phase sine wave, but, of course, you will hear an AC motor spinning all the time:  Note that other youtube videos include schematics and tests and what to avoid...

[richard.cs]

You have a 120VAC split phase supply, which then gives you 2 hot wires and a neutral. Thus simply take the highest loaded phase and connect the one hot to that, and tie the other 2 to the other phase.

Hmmm, this is something that I don't think anyone has tried before but given the history of the internal neutral failing with increased load I have to admit my doubts the box would be OK.  And there's always the hum-bars-in-the-picture thing to worry about...

If all the phases were equally loaded then the neutral current would equal the line current, so unless the neutral is smaller than the line conductors this would be fine. If they're unequal and you pick the highest loaded one to be on the opposite phase then the neutral current is lower. The only condition where the neutral current is higher is if you pick the wrong one, perhaps because the phase balance is time varying.

In terms of hum bars you would have to try it and see but at least it's not damaging to do so. Depends entirely what the pickup mechanism is as to if it cancels when phased like this or not. I suspect it'll be better than with all on one phase but perhaps worse than in three phase. There has to be some such effect on 3 phase because the balance is never perfect, just weak enough not to see.

[drussell]

Build your own 1 phase to 3 phase rotary converter.  You will have your true real 3 phase sine wave, but, of course, you will hear an AC motor spinning all the time:  Note that other youtube videos include schematics and tests and what to avoid...

A rotary converter would certainly work, however, for the power level that the OP needs it is going to have to be one seriously big honkin' motor, especially to handle the horrible power factor that is likely to be presented by a bunch of capacitor-input rectifiers in all those power supplies without the wave tops getting so smashed that it looks more like a square wave.   

Edit:  After more thought, I don't believe a rotary converter would ever be able to be made to work correctly in this case, especially since the loads are not static.  There would great imbalance between the phases to begin with that would be unlikely to be able to be compensated for with the capacitor arrangement, certainly not always stable with varying loads on each phase.  I do believe a motor-generator set would work, if appropriately sized, but I think the single-motor converter is out.

That will not fulfill his "low noise, somewhat sinusoidal" requirement. 

The motor will also make some interesting noises with power being drawn only at the peaks of the wave, especially when the phases are going to be randomly imbalanced and the imbalance is going to be changing during operation.

Remember, we're talking about needing about 20 amps per phase, all being drawn through DC supplies with (presumably non-power factor correcting) rectifier inputs.  I expect the peak currents to be very high, and only at the tops of the waves and that is probably why it has a 30 amp supply, even though the average power stays down below 20 amps.

<story mode on>
My first appreciation of just how much power factor actually meant in practice was a similar issue 15 or 20 years ago that I had just really never thought through, even though I had built dozens of different and all manner of power supplies and such before in my lifetime.  It took sticking my scope on there to see what was going on before the light bulb above my head finally came on.

We had installed a (for the time) fancy new inverter/charger system at a friend's off-grid cabin so there could be power without having a noisy generator running.  (A Trace Engineering DR series with 6 x GC2 batteries, IIRC).  The original generator that was at the cabin was the Honda (EX?)3500 workhorse that had been there for years.  The inverter supplier had indicated to my friend's dad that charging would be slow with such a generator, which it was, so he had already planned to buy a larger unit.

The next spring, they took up a new 6500 Honda but, of course, it didn't charge any faster either.  (Humorously enough, I calculated later that the old one was actually faster, and they are even in the example charge rates for various generators in the Trace manual, though Trace didn't do a very good job of explaining why, power factor, etc. and basically just said "you will need a much larger generator than you think")  The cabin owners were a bit dismayed, of course, but knew I would be coming out in the summer anyway.

At Trace's suggestion, I brought a nice big motor-run capacitor and wired it right across the generator output to help with the PF.  I also parallelled in a second 10ga cable to the inverter/charger from the generator shed to reduce the impedance of the feed line.  I had even brought a little PC-based scope with me (fortuitous choice) just in case and did various basic tests before I finally built a danger-divider using some resistors I happened to have in my laptop bag and put the scope right on the mains (scope and laptop floating, just in case I did something stupid) to see what was really going on.  When I saw just how bad the waveform really was, I knew exactly what the problem was.  Well, I already knew what the problem was but not the magnitude of the problem!  The charger was only drawing current from the part of the wave that was above the equivalent battery level since it was a smart, but relatively standard, non-PFC charger circuit.  The average voltage was being held correctly by the poor Honda but there is just no way it could supply all that power just on the peaks.  I knew it would be a bit clipped, that was expected, but I wasn't expecting the wave to me THAT mashed.

The next morning I was sitting on the steps of the generator shed, pondering the manual for the Honda when I came across the schematic and another lightbulb-over-the-head moment struck me.  Eureka!  Another friend (who is an electrical engineer) stumbled out that way, probably to re-tap the keg for a breakfast brew (we're at the fishing cabin remember, don't judge ) and I said, "Hmmm, Dös, take a look at this, what if I take this winding, flip it over and jam it on this one, that should work, right?"  He basically said, "Hmm, yeah, that should work, just don't get it backwards and burn out the windings.  LOL"

The 6500 generator has two separate windings which make up the two sides of the 240 volt output.  The 120 plugs are just distributed half on each side, so in effect our charger was only running on a 3250 watt generator winding.  The old 3500 Honda had a switch for 120/240 volt which used both halves when in 120 mode to actually supply more current than the 6500 could.

Anyway, I opened up the control panel, took off the winding wires, re-jiggered it so they were both in phase and connected together, buttoned it back up and fired her up.  The voltage gauge now only read 120, of course, but it only took a moment before Dös was out there sayin' "Dude, you gotta come check this out!"  I went into the cabin to the inverter and, sure enough, the inverter/charger was hummin' like it had never hummed before.  Dös said "Wow, that's impressive," when he felt the cables.  He was working for Nortel at the time and they have some giant honkin' 48 volt rectifiers and battery banks but he was impressed that our 00 or 000 cable running to the batteries was actually getting warm.   

Essentially, I had (more than) doubled our charge rate by doubling up the 120 at the expense of the ability of the generator to do 240 volts.  (Since we had only ever used 240 once before, to run my giant compressor during a cabin expansion the previous year, that wasn't an issue.)

Charge rate was now solved but I was still never able to properly equalize the batteries with that set-up.  You just couldn't get the battery voltage high enough to really do it, even after many hours of running at full charge.  PFC control chips were just starting to become a thing so I started drawing up plans to make a PFC'd battery charger but in the years since, they ended up added a solar panel setup which keeps the batteries fully topped up, even when nobody is there for 95% of the year so it has become less of an issue and I never built the PFC charger.
</story mode off>

My point is, you're likely to need a WAY bigger motor than you expect if you go this route.  That is where an electronic solution may well become more practical than the mechanical one.  Phase imbalance becomes irrelevant and any design with feedback will at least do its best to keep the waveshape correct within the limitations of the wiring impedance, etc.

[drussell]

If all the phases were equally loaded then the neutral current would equal the line current, so unless the neutral is smaller than the line conductors this would be fine. If they're unequal and you pick the highest loaded one to be on the opposite phase then the neutral current is lower. The only condition where the neutral current is higher is if you pick the wrong one, perhaps because the phase balance is time varying.

Unfortunately it is time varying, depending on the operation of the machine at the time.

The OP also states that any re-wiring to re-balance is impossible since it is located in the bottom of the 1600 pound machine, so that's a non-started also, supposedly.

If the OP can get to at least enough wiring to move a couple of the right supplies to another phase to re-balance onto 2 phases instead of 3, everything would work great.  I have never seen the machine, obviously, but if there is any way that can be done, even having to add external wiring right to an external junction box or something, it would make everything so much simpler and less expensive than any other solution.

I can't fathom how it wouldn't be possible, even if the OP cannot get to the main distribution inside the unit.  If it is possible to get to a couple of the supplies that are on ANY one phase, the necessary minor re-balance should be possible... 

...but then again, I have not seen the actual machine....   

[BrianHG]

Though it would be big, looking at your current requirements, and this one is completely out of my area of expertise, what about a https://en.wikipedia.org/wiki/Scott-T_transformer, or, maybe use one of these in reverse http://transformer.sikes-elec.com/Transformer/Three-phase-to-single-phase-transformer .

[BrianHG]

Maybe use a phase shifter transformer to create your missing phases: http://transformer.sikes-elec.com/Transformer/ZYSFG-Phase-shifting-rectifier-transformer/Rectifier-transformer-Phase-shifting-ZYSFG-Series.html

[Floyo]

That scott T transformer is meant for four wire 2 phase networks, running at 90 degrees, not the 180degree split phase system.
This system 90 degree system is oldschool, and not (really?) used any more because you need four wires for two phases instead three wires for three phases with the 120 degree system, which is more efficient cabling wise.

The other option also wont work, you don't magically get the required phase shifts.

Interestingly the Phase Perfect converters mentioned by duak use a principle similar to the scott T transformer by electronically generating a phase that is 90 degrees out of phase with the split phase supply, so the line line voltages end up 120 degrees apart with the right voltage, but the line-neutral voltage of the generated phase is higher than 120V, and thus this wont work for the OP (unless using a delta-star transformer). The white paper has a decent summary of the issues presented in this thread http://www.phasetechnologies.com/downloads/products/phase-perfect/phasewhitepaper.pdf

[Fryguy]

Hi there ,

I'd go for the 3phase powerline installed to the building by the power company .

cvanc has to come up with a solution in less than 2 weeks and i don't think the proper rewiring of the machine for 2 or 1 phase operation can be done in this timeframe due to the complexity of the existing installation inside the machine and the problem of accessing it and moving this really big , heavy and fragile heap around without crashing it .

The rotary converter seems to be out of the question because of size / weight and noise issues .

Gluing the 3 phases together to 1 phase is a nice idea - i wonder how big the current impulse will be when you turn on 2 dozen big power supplies at the same time . . .    and i'm not sure this nice setup with the 3 audio power amps will be able to handel the startup current either ( that is where the rotary converter will work pretty good )

Well - these are my thoughts so far .

[Brumby]

It's been stated before that messing about with the internals of this "black box" is completely out of the question.

It should also be repeated that there is a known weakness with the neutral line inside the box - and since messing about inside is out of the question, then any solution has to be sympathetic.

You only need to think about the basics regarding current in the neutral to appreciate the need for a 3 phase solution....  Say you have 3 circuits that each draw 20A.  Use a single phase solution and the neutral current is 60A.  If your neutral is taxed at 40A, then running everything from a single phase is just not a long-term option.  Use a 3 phase solution and the neutral current is zero (for a balanced system) - but even when unbalanced, the neutral current can stay well within its limit for quite a range of unbalance.

And just to repeat - it is some part of the neutral line within the black box (that we cannot modify) which is the limiting factor.

The solution must not be a single phase one.

[NiHaoMike]

You only need to think about the basics regarding current in the neutral to appreciate the need for a 3 phase solution....  Say you have 3 circuits that each draw 20A.  Use a single phase solution and the neutral current is 60A.  If your neutral is taxed at 40A, then running everything from a single phase is just not a long-term option.  Use a 3 phase solution and the neutral current is zero (for a balanced system) - but even when unbalanced, the neutral current can stay well within its limit for quite a range of unbalance.

Run two phases from one hot of 120/240V and the third from the other hot. The neutral current would then be 20A with equal loads. As mentioned before, since the load is not balanced, put the phase with the highest current draw on one hot and the remaining two on the other hot to get something closer to balanced.

It's also worth noting that with non PFC rectifier loads, the neutral currents do not fully cancel with 3 phase, but do largely cancel with split phase. Therefore the split phase idea would be even less of an issue. I would say wire it up split phase and monitor the neutral current to determine if it's acceptable.

[Beamin]

Is this some sort of milling machine with CNC and motor controllers? Hence all the different power supplies?

No it isn't. Why not read the thread instead of doing a drive-by posting? 

I have a hard time seeing and often miss things or if a thread has more then one page.

[drussell]

I'd go for the 3phase powerline installed to the building by the power company .

cvanc has to come up with a solution in less than 2 weeks

I believe that when the OP said:

I've got maybe 2 weeks to propose a solution so there's a bit of time to kick this around but not a lot.

OP was trying to say that while he needs to propose a solution in the next week or two, that does not necessarily mean it can be implemented in that timeframe, though I am sure that it will become a "sooner, rather than later, please" affair shortly after that.

Since the power company hasn't even responded yet, there is a good chance that it could potentially be many months before that could be installed, if it is even within the budget.  If three phase is already in the building, it will be possible and cost-competitive with other proposed solutions, probably in the range of $5000 to add a minimal 3-phase service.

If 3-phase is not currently in the building, the price instantly becomes tens of thousands of dollars.  If it were in a remote location (unlikely), it could be $100,000.  Once you get into multi-tens-of-thousands of dollars, it may be more cost effective to change locations than to bring 3-phase power to that building.  Without knowing anything about the building or its location, I cannot speculate further, nor can the OP until better information is received from the client.

Any operation that is running an old telecine machine like this (OP, what is this, a MkIII or something?  ) is either doing something very specialized, or more likely trying to save money by using older equipment and that may well make the cost of bringing in 3-phase unattractive.  The OP knows more about the overall situation that I do, though. 

and i don't think the proper rewiring of the machine for 2 or 1 phase operation can be done in this timeframe due to the complexity of the existing installation inside the machine and the problem of accessing it and moving this really big , heavy and fragile heap around without crashing it .

For single phase, 120 volt operation, absolutely.  Not only would that require a complete re-wire of the main power distribution in the machine, it would be an impractical thing to do.  Why would you load down one side of your 120-N-120 with a single 60-90 amp load?

For split single phase, 120-N-120 operation, that remains to be seen.  While I certainly understand the OP's hesitation about attempting any modifications, and appreciate the fact that this position may well not change, it is still an option, and the main distribution inside the machine does not have to be accessed or touched in any way to achieve.

If we assume that there are 7 power supplies on each of the three incoming phases, the OP would likely only have to be able to access the power input wires to three or four of those supplies that are on any one of the three phases.  New wiring could be brought directly out of the unit and connected externally to whichever phase had the lower loading.  Cost?  Maybe a couple hundred dollars in supplies, absolute maximum, plus the OP's time to get to those three or four supplies within the unit.

The rotary converter seems to be out of the question because of size / weight and noise issues .

Size, weight and noise are actually relatively minor concerns with this approach.  I am more concerned with that kind of unit's ability to provide anywhere close to balanced voltages.  Not only do they typically NOT provide anywhere near a balanced voltage output, the mismatched and time-variant loading of this unit complicates matters further.  It will be virtually impossible to statically balance the capacitor arrangement to have anywhere near proper balanced power.  What may work fine for a motor in a machine tool is not necessarily appropriate for a fragile telecine machine that expects clean power.

I also think that harmonic currents are going to be a huge issue with a rotary converter.  Clipping of the wave is going to be severe unless the motor is huge and the cost of 25+ HP motors is not insignificant, nevermind the capacitor bank, etc.

A third option is a motor-generator set.  THIS has the potential to work well enough, provided the generator is sized appropriately to handle the horrible power factor without completely losing the tops of the waves.  Far superior from an electrical perspective than the rotary converter but, again, it is going to have to be vastly oversized from the nameplate rating for a resitive load to be able to handle capacitor-input rectifiers clean(ish)ly.

Gluing the 3 phases together to 1 phase is a nice idea

No, actually that is a terrible idea.

... - i wonder how big the current impulse will be when you turn on 2 dozen big power supplies at the same time . . .    and i'm not sure this nice setup with the 3 audio power amps will be able to handel the startup current either ( that is where the rotary converter will work pretty good )

A rotary converter would just have severe voltage drop for short time while the capacitors were charging.  That isn't really a good or bad thing.  A motor-genset would have the same issue, but that in and of itself wouldn't cause any problems.  There are far bigger problems with a rotary converter.

As for the electronic APS-3720 AC power supply / regenerator doodad shown above, the available output current depends on what the original amplifier capabilities are and how they have been modified.  The original specification for the design requirement given to me was for a 3-phase, 400 Hz, 115 volt, 20 ampere supply, later amended to be variable from approx 200-600 Hz for flexibility in testing once I offered them the option of it being variable.

My initial testing was done with a QSC RMX-2450, since I had one on hand anyway, and it worked so well that I have kept using that same circuit in subsequent units, even though I had initially thought it might need to be beefier.  The current limiting is adjusted to start to come in at about 25A and short circuit protection trip at a bit over 30A, IIRC.  I would have to check my notes.  There will be distortion of the sine wave by that point, but it will do it, and there is a distortion indicator on the front panel of each phase's LPA to indicate that condition.  I can start my 4 HP air compressor (wired for 120 instead of 240) off one phase.  It starts slightly slower than it does on MAINS power (takes a couple cycles to get to speed) just like it does on a generator (even wired for 240), but it works fine.

Even if it were an issue, this apparatus can easily ramp up the voltage to provide soft-start if required.  I don't recall if I ever tried a 2400W+ load that was that "nasty" like a huge rectifier-capacitor battery charger load simulation, perhaps I should characterize that, but all sorts of DC power supplies have been run from the APS-3720.

The one time that there actually was an issue with powering a DC power supply was shortly after delivery of the first unit.  It ended up being an interaction between the active power factor correction in the load (the DC power supply) and the feedback loop in the LPA setting up very high frequency oscillations in the LPA output stage and frying resistors in the snubber networks.  A design change was made on my end, the units updated and the resistors replaced.  There have been no subsequent issues powering any loads.  (I tried everything I could think of for load testing, from air compressors to microwave ovens, universal motors like power tools and vacuum cleaners, overloaded it with halogen lights and boiling water in kettles.)

I might just have to set one up for more torture testing, though, now that this thread has piqued my interest. 

In any case, even more robust amplifiers can be chosen if necessary.  The modular topology of the whole apparatus makes alterations like that possible.  The APS-3720 shown above is capable of pulling over 40A from each of the three phases on the AC input.  Some of that goes to heat, obviously, but there is a whole bunch of surge current available to run just about anything I could imagine to throw at it. 

[Urs42]

I did hear that someone i know is using a modified three phase online UPS to generate three phase power from a single phase, i'm not sure what exactly he did, i think that he connected all three phases on the input to the same phase, i guess that some more modifications are needed.

[Fryguy]

drussell - i know that gluing the 3 phases together is absolute BS and will never work without some modifications to the machine - this part was meant as a joke ! 

If your power-amp-rack can push that initial current without the overload-protection shutting it down - nice work ! 

How much would it cost to get a rack like that (working at cvanc's specs) ready to go and how much time would it take ?

[G7PSK]

If you really need a good 3 phase supply you need a motor generator as rotary converters don't produce even power on all phases and can lead to motors overheating, been in that situation myself and ended up with a diesel gen set as the requirement was for 40 KVA and I had somewhere to install the power plant.

[sourcecharge]

Again, thank you everybody.

Best word of the thread so far?  "Dynamite" 

Doing a quick search on startpage for 12kw 3 phase pure sine wave inverter:

https://www.electriccarpartscompany.com/12kW-3-Phase-Inverter-Charge-Controller-HY

12,000 Watts 12kW Energy Storage Package
Pure Sine Wave
Inverter and Charge Controller
3-Phase Off Grid
Works With Lithium or Lead Acid Batteries
Not UL Approved    
ECPC-3-Phase-12kW-Inverter-HY

Specifications of your 12,000 Watts 12kW Energy Storage Package Pure Sine Wave Solar Inverter:

·       Phase: Three Phase

·       Rated Power Output: 12,000 Watts

·       Pure sine wave output

·        Maximum >98% efficiency

·        Frequency: 50Hz/60Hz

·        CE, ISO9001, FCC

·        Operating Temperature: 5F to 122F (-15C to 50C)

·        Storage Temperature: 5F to 140F (-15C to 60C)

Price
Quantity    Price
1+    $2,517.00


Need more info? (801) 449.0681 info@electriccarpartscompany.com

The picture shows that it can use the AC mains, a solar array, or batteries...

I have no idea if this company is good or not and I would call them and verify that it would suit your needs before also doing the research into this company's reliablity...

hope that helps...

[Brumby]

You only need to think about the basics regarding current in the neutral to appreciate the need for a 3 phase solution....  Say you have 3 circuits that each draw 20A.  Use a single phase solution and the neutral current is 60A.  If your neutral is taxed at 40A, then running everything from a single phase is just not a long-term option.  Use a 3 phase solution and the neutral current is zero (for a balanced system) - but even when unbalanced, the neutral current can stay well within its limit for quite a range of unbalance.

Run two phases from one hot of 120/240V and the third from the other hot. The neutral current would then be 20A with equal loads. As mentioned before, since the load is not balanced, put the phase with the highest current draw on one hot and the remaining two on the other hot to get something closer to balanced.

That is something that had occurred to me as well - but it had already been mentioned and there did not seem to be any interest in that approach, so I didn't repeat it.  I still wonder if it might be workable, though.

[rx8pilot]

[I HAVE NOT READ THE WHOLE THREAD]

I use a Phase Perfect solid-state phase converter at my house to power my CNC machine. My particular model is 55A 3Ø output.

The good news is that it works very well, very reliable even when the CNC surges huge current. The bad news is that it is rather expensive. The CNC manufacturer told me a rotary phase converter would void the warranty so I went this route.

[Brumby]

I'd forgotten about that!

Definitely a candidate for this situation - and the OP won't need that high a capacity, so it would be cheaper than yours ... but probably not cheap.

[drussell]

[I HAVE NOT READ THE WHOLE THREAD]

Obviously.  You should go back and read the Phase whitepaper.

Those units are intended to generate a simulated third leg for motor loads while running the other two phases directly from the two input legs.  While it may be possible that this will regulate properly (you would have to talk to Phase) and give the OP a third, 240-volt phase, two of them will be 180 degrees apart and the third, generated one will be somewhere in between.

This means that the OP will still have the same problem with the neutral that he already has.  It won't burn up, just like it won't if he runs two of the telecine machine's phases off one input phase and the remaining one off the other, but will still have far more neutral current flowing all the time than in the stock configuration on 3-phase  Since one of his concerns is hum bars, etc. causing noise in the output picture, we are trying to find a solution that preserves the original configuration as much as possible (ie, minimal neutral current with imbalance only!)

Edit:  Plus, I forgot to mention, I don't believe you get a neutral out of those units that actually would produce balanced voltages.  The reason they can get it to look very close to a motor, even with a 180-degree input is by creating the third phase at would would be an "oddball" voltage to the neutral.  Remember that by adjusting the relative voltages of each phase, you are effectively shifting the phase leading or lagging (left or right of you're thinking about it on a graph.)

The good news is that it works very well, very reliable even when the CNC surges huge current. The bad news is that it is rather expensive. The CNC manufacturer told me a rotary phase converter would void the warranty so I went this route.

You're running a machine tool off of it, not a telecine machine with no AC motors, DC supplies with all sorts of imbalanced, randomly changing loads, etc.  The OP would need to talk directly to Phase about it but I do not believe this would work any better than just using the 120-N-120 he already has and has the potential for additional induced noise from the Phase unit itself.

[drussell]

I'd forgotten about that!

Definitely a candidate for this situation - and the OP won't need that high a capacity, so it would be cheaper than yours ... but probably not cheap.

Why would using a simulated third leg (at 90 degrees?) be any better than just using his existing 120-N-120??

Note that in the whitepaper, their "Case Study" is seems to be a bit of BS, since they were using two input phases that were 120 degrees apart, so their unit creates another one at 120, (pointlessly) re-creating 3-phase.

The OP has 180-degrees.  The Phase Perfect units only generate one phase, the other two outputs go directly to whatever power is on the input, so he wont' be getting 120-120-120 degree phases out.

Wouldn't a giant capacitor essentially theoretically work just as well to get a 90 degree shift?

[BrianHG]

These pure sine wave 3 phase online, true galvanic isolation, DC power supply/battery start support, UPSs should do the trick:
https://www.tesla-pv.com/three-phase-pure-sine-wave-inverter-ups/

Thanks to the true galvanic isolation, you can feed the inverter with a high current DC power supply from the grid without neutral/gnd fault.  I'd still check with the manufacturer.  I think it's 50hz though...

[BrianHG]

Just go with the 'Phase Perfect' unit, it's purpose built for what you want to do and need.
Here is a Canadian distributor: https://www.shopampro.com/SearchResults.asp?Cat=409

True website: http://www.phaseperfect.com/p/t/overview
http://www.3phasepower.ca/phase-perfect-480v-digital-phase-converter/
pdf http://www.phasetechnologies.com/downloads/products/phase-perfect/pt380specs.pdf

You got these, but, Im not sure they would be strong enough. https://adaptivepower.com/products/ac-ac-sources/cfs300-series/

Also these guys: http://www.3phasepower.ca/phase-converters/

Cheap units: http://www.phase-a-matic.com/StaticDescription.htm

I'll avoid all the Alibaba stuff...

[capt bullshot]

Just go with the 'Phase Perfect' unit, it's purpose built for what you want to do and need.

Better double check. The block diagram doesn't show a neutral at the output, so it won't work with OP's problem. It works by creating a third leg, but leaves the original phases as they come from the two-phase feed. That input neutral isn't identical with the required output neutral, the unit doesn't generate the output neutral.

Quote from the manual:

If the connected load has a neutral connection and requires wye  configured power, the output of the phase converter must be passed  through a delta-to- wye isolation transformer before connection to  the load.

[BrianHG]

Quote from the manual:

If the connected load has a neutral connection and requires wye  configured power, the output of the phase converter must be passed  through a delta-to- wye isolation transformer before connection to  the load.

Darn...  However, you can get huge-ass refurb/used 3 phase transformers for not all that much and they are pretty efficient.

Only this guy with a built in iso transformer may work, but, you need to supply it with DC power...
https://www.tesla-pv.com/three-phase-pure-sine-wave-inverter-ups/

[Brumby]

I'd forgotten about that!

Definitely a candidate for this situation - and the OP won't need that high a capacity, so it would be cheaper than yours ... but probably not cheap.

Why would using a simulated third leg (at 90 degrees?) be any better than just using his existing 120-N-120??

Note that in the whitepaper, their "Case Study" is seems to be a bit of BS, since they were using two input phases that were 120 degrees apart, so their unit creates another one at 120, (pointlessly) re-creating 3-phase.

The OP has 180-degrees.  The Phase Perfect units only generate one phase, the other two outputs go directly to whatever power is on the input, so he wont' be getting 120-120-120 degree phases out.

Wouldn't a giant capacitor essentially theoretically work just as well to get a 90 degree shift?

Well, I fell flat on my face by assuming the "Phase Perfect" unit produced true 3 phase (120º angle between phases) which it does not.

And from my read, that white paper makes no mention of phase angles other than for utility supplied power - and phase angles were the one parameter that I was looking for all through the paper.  As such, I can't see where you get the 120º input phase angle for their illustration.  A lot of that white paper is a basic education about 3 phase and VFD issues.  Not a lot is said about the product other than it will keep phase voltages in check and cut out if things go sideways.

As such, it's hard to comment on what the actual phase angle of the generated line will be.

[drussell]

And from my read, that white paper makes no mention of phase angles other than for utility supplied power - and phase angles were the one parameter that I was looking for all through the paper.  As such, I can't see where you get the 120º input phase angle for their illustration.

Page 11, first sentence:

A 10 HP model DPC-10 Phase Perfect® digital phase converter was connected to two legs of a 208V grounded wye three-phase service.  The converter generated the third leg voltage."

[richard.cs]

Well, I fell flat on my face by assuming the "Phase Perfect" unit produced true 3 phase (120º angle between phases) which it does not.

It does, the three outputs will be correctly phased relative to each other. What it is doing is generating a 3rd phase that when considered relative to the supply neutral and L1 is at 90 degrees and >120V (something like 120*(1+cos(30)) Volts), such that the 2 original phases and the new one are 120 degrees apart rotating about a virtual neutral that is 90 degrees and 70 V or so from the supply neutral. It looks correct from the viewpoint of a delta connected load which doesn't know there is a common mode voltage at 90 degrees and a delta-star transformer would allow you to connect unbalanced star loads.

It's quite elegant really, it's much easier to see when you draw out the vector diagrams. Note that it cannot produce simultaneously exactly 208V phase-phase and 3x120 degrees from a 240 V split phase supply, to get to 120 degrees it has to add enough 90 degree voltage to push the phase-phase voltage up a bit. If this matters the extra voltage can be taken out again in the delta-star transformer.

[tpowell1830]

I have been listening and trying to see another way to solve this problem, but every new solution that I could think of was either knocked down by a user or just not feasible. Although I think that a good re-wire, as drussell and others have laid out before, would effectively re-distribute the neutral load enough to do what is needed, it still leaves the original problem that the machine is old and fidegety and the OP is afraid that this might break the old machine, plus the fact that the internal wiring is very difficult to access only leaves the one choice. That choice is getting a true 3 phase feed into the building to run this machine that has the wye configuration with a common neutral. All of the options that have been mentioned so far does not directly address the unbalanced neutral except for the re-wire of the machine.

How many times have we seen a similar problem from someone with the 3 phrase needs and all of the commercially available products are either electrically noisy, audibly noisy, or both. drussell, it sounds like you have had some design experience in creating a power supply that can create a 3 phase system, and although I really liked the 3 phase audio amp idea that was discussed, this does not address large 3 phase loads. It looks to me that if someone could design and build a 25-30 KVA wye with neutral unit that is electrically quiet and reasonably priced that could be driven by a 240 VAC single phase power, this would be a win-win system that would be very popular among small shops. I have seen this problem over and over again where, either 3 phase was not available at the power poles at a location, or it was deemed too expensive to implement because the power company wanted an arm and a leg to place 3 phase service. This is always the problem for a smaller shop.

Exactly what would it take and what resources would it take to design this product and get it out to market? I have been looking for a product idea for a long time and it seems that this is a recurring theme here in the US, where a small company is stymied because of the high cost of 3 phase power. What would the price structure of such a device be after designing and building such a unit? In other words, this could be an opportunity to capitalise and create a product that is a win for a entrepeneur and a win for small businesses needing this device.

Any comments or negative critiques are welcome. I am going to seriously ponder this and consider creating such a device. I have great respect for the knowledge and intuitive thoughts that I have seen on this forum. I have had several small businesses in the past and have long been looking for a product to develop, and any thoughts as to why this is not a feasible product are welcomed. Anyone who has the technical know-how to develop such a product and the interest to do a project are welcome to PM me to further discuss the possibilities. I may go as far as doing a feasibility study.

AHA moments... are rare.

[coppercone2]

I saw a AC source before, their really nice bits of kit, but they defiantly do have a high end look to them .

I think you have one per section and you adjust the phase shift between them. The ones I saw were meant to change in frequency and voltage and phase, so you can do military testing like 40-400Hz/90-300V AC on all three phases.

yea, looks like a chroma unit is what I saw. they look good.

[drussell]

It looks correct from the viewpoint of a delta connected load which doesn't know there is a common mode voltage at 90 degrees and a delta-star transformer would allow you to connect unbalanced star loads.

It's quite elegant really, it's much easier to see when you draw out the vector diagrams. Note that it cannot produce simultaneously exactly 208V phase-phase and 3x120 degrees from a 240 V split phase supply, to get to 120 degrees it has to add enough 90 degree voltage to push the phase-phase voltage up a bit. If this matters the extra voltage can be taken out again in the delta-star transformer.

Indeed, that is a good explanation of what I was trying to say. 

It would work, and I expect it would regulate reasonably well even with varying loads on each phase but you need to add the transformer on the output and you will need to adjust the voltage to get close to 120 to N, so will probably need a transformer with multiple taps on one side or the other or a custom transformer.

[drussell]

How many times have we seen a similar problem from someone with the 3 phrase needs and all of the commercially available products are either electrically noisy, audibly noisy, or both. drussell, it sounds like you have had some design experience in creating a power supply that can create a 3 phase system, and although I really liked the 3 phase audio amp idea that was discussed, this does not address large 3 phase loads.

Why not?  Three phase loads run just fine from an APS-3720.  The amplifiers have been modified to be completely isolated isolated and the appropriate side of the output from each is tied together to form a neutral.  If I set the voltage controls on the front panel to be 120-120-120 you will get 208 volts across any two legs.  Three phase motors run fine on it.  I tested an unloaded 3-phase motor on them (had no good load available), I tested my 4HP compressor across two legs, with and without additional loads and some of the door units the customer originally ordered the design for have large 3-phase 400 Hz fans and things like that in them.  They all work fine.

The load doesn't know what is powering it and the amplifiers don't know what they're powering and don't care about the power being drawn "120 degrees off" between amplifiers.  They don't know the difference between that and a reactive loudspeaker load.

With appropriately sized amplifiers, that type of solution should work for the OP.  If I can rig up a test load with the same kind of harsh, wave-hacking effects as his telecine machine would be presenting, I could verify that and change that "should" to a "certainly would".  His load might be right on the edge, power wise, of what the pictured 7200 watt RMS rated model can do (like an inverter, it will do about double that for an instantaneous surge) but using something like the QSC 4050HD amplifiers as the outputs should allow 30A continuous.

In any case, I'm not sure what you mean by "large three phase loads" wouldn't work.

It looks to me that if someone could design and build a 25-30 KVA wye with neutral unit that is electrically quiet and reasonably priced that could be driven by a 240 VAC single phase power, this would be a win-win system that would be very popular among small shops. I have seen this problem over and over again where, either 3 phase was not available at the power poles at a location, or it was deemed too expensive to implement because the power company wanted an arm and a leg to place 3 phase service. This is always the problem for a smaller shop.

That's exactly what I designed the amplifier based unit for.  A cost-effective unit for a small aircraft engineering company here in Calgary at the little municipal Springbank airport hangars.  They happen to have three phase, of course, they just needed it to be 400 Hz. 

It has proven to be satisfactory in service.  At least satisfactory enough its use is written into test procedures for the US Military for those systems, hence the need for them to buy more units for use at various places around the world.

I have thought about building ones into a custom chassis of some sort instead of supplying in rack form but have never got around to that.  Airdyne loves the current rack mount design and have some mounted in military-grade flight cases with a complementary Lambda 0-60V 50A DC supply to be able to simulate aircraft power anywhere they want.[/quote]

[Brumby]

And from my read, that white paper makes no mention of phase angles other than for utility supplied power - and phase angles were the one parameter that I was looking for all through the paper.  As such, I can't see where you get the 120º input phase angle for their illustration.

Page 11, first sentence:

A 10 HP model DPC-10 Phase Perfect® digital phase converter was connected to two legs of a 208V grounded wye three-phase service.  The converter generated the third leg voltage."

My apologies.  I walked right into that one as well.  Not having a good day.

However, I find it difficult to reconcile that setup with the statement "Balanced three-phase power from a single-phase source."   Wait ... scratch that.  My cold clouded head has some maths bouncing around - and I think I need to sleep on it before embarrassing myself any further.

[IanB]

The load doesn't know what is powering it and the amplifiers don't know what they're powering and don't care about the power being drawn "120 degrees off" between amplifiers.  They don't know the difference between that and a reactive loudspeaker load.

I guess the corollary of this is that large audio power amplifiers can electrocute you. Maybe not something that immediately comes to mind when working with them. Do they have "Danger: High Voltage" stickers on the outputs?

[Brumby]

Well, I fell flat on my face by assuming the "Phase Perfect" unit produced true 3 phase (120º angle between phases) which it does not.

It does, the three outputs will be correctly phased relative to each other. What it is doing is generating a 3rd phase that when considered relative to the supply neutral and L1 is at 90 degrees and >120V (something like 120*(1+cos(30)) Volts), such that the 2 original phases and the new one are 120 degrees apart rotating about a virtual neutral that is 90 degrees and 70 V or so from the supply neutral.

That sounds like the direction my muddled thinking was heading.  The virtual neutral being offset from the supply neutral was a key fact.

It looks correct from the viewpoint of a delta connected load which doesn't know there is a common mode voltage at 90 degrees and a delta-star transformer would allow you to connect unbalanced star loads.

It's quite elegant really, it's much easier to see when you draw out the vector diagrams. Note that it cannot produce simultaneously exactly 208V phase-phase and 3x120 degrees from a 240 V split phase supply, to get to 120 degrees it has to add enough 90 degree voltage to push the phase-phase voltage up a bit. If this matters the extra voltage can be taken out again in the delta-star transformer.

That sounds about right too.

It is an elegant solution ... but it doesn't provide the real neutral our Op needs without that delta-star transformer.

[drussell]

I guess the corollary of this is that large audio power amplifiers can electrocute you. Maybe not something that immediately comes to mind when working with them. Do they have "Danger: High Voltage" stickers on the outputs?

Yes, they do, and newer ones are now required to have touch-proof connections like Speak-on or binding posts with a cover over them, etc when the voltage is above the (whatever it is, 48 volts, 60 volts?) safety limit, so any higher power PA amplifier now must have physical safety protection for just that reason.

An amplifier that can do 5000 watts RMS into a speaker load on its output terminals puts out some hefty voltages! 

Higher than 120 volt line voltage.  120 volts into 8 ohms is only 1800 watts (15A).

[rx8pilot]

An amplifier that can do 5000 watts RMS into a speaker load on its output terminals puts out some hefty voltages! 

Higher than 120 volt line voltage.  120 volts into 8 ohms is only 1875 watts (15A).

I never really thought about that. The modern super-high-power amplifiers really have a stunning challenge.

[SeanB]

Try electrostatic amplifiers, where even headphones have kilovolt power supplies in them, right next to your ears.

[BrianHG]

Ok, for those talking amplifiers...
Here, from my 60kw subwoofer thread, an amplifier with a supply of 227v-378v in DC, up to 220 amps out continuous, class D, 5hz to 1Khz....  (Designed to tie directly to EV car batteries, so, not step-up transformer)

http://www.stetsom.com.br/en/produto/force-one/

You will need 1 killer DC power supply for 3 of these guys...

[rx8pilot]

Ok, for those talking amplifiers...
Here, from my 60kw subwoofer thread, an amplifier with a supply of 227v-378v in DC, up to 220 amps out continuous, class D, 5hz to 1Khz....  (Designed to tie directly to EV car batteries, so, not step-up transformer)

http://www.stetsom.com.br/en/produto/force-one/

You will need 1 killer DC power supply for 3 of these guys...

That. Is. In. Sane.

I mean......seriously. Output power measured at 250mΩ

[BrianHG]

Ok, for those talking amplifiers...
Here, from my 60kw subwoofer thread, an amplifier with a supply of 227v-378v in DC, up to 220 amps out continuous, class D, 5hz to 1Khz....  (Designed to tie directly to EV car batteries, so, not step-up transformer)

http://www.stetsom.com.br/en/produto/force-one/

You will need 1 killer DC power supply for 3 of these guys...

That. Is. In. Sane.

I mean......seriously. Output power measured at 250mΩ

Well, the OP wouldn't have to worry about power sags on the peaks of the AC waveform due to load using these amps....

[capt bullshot]

Ok, for those talking amplifiers...
Here, from my 60kw subwoofer thread, an amplifier with a supply of 227v-378v in DC, up to 220 amps out continuous, class D, 5hz to 1Khz....  (Designed to tie directly to EV car batteries, so, not step-up transformer)

http://www.stetsom.com.br/en/produto/force-one/

You will need 1 killer DC power supply for 3 of these guys...

Not a big deal anyway. Usually such an unit would used as a high power VFD or servo drive / amplifier. One can buy them in the MW ranges, e.g. used for traction, or renewable energy, high frequency drives e.g. for turbine based air compression. The power electronics alone is suprisingly small, but you'll need some external "accessoires" like line filters and chokes, output chokes and water cooling. These usually exceed weight and size of the power electronics.

[taydin]

I think a motor/generator pair would be the most reliable solution. But if an asynchronous motor is used, the output frequency will be slightly less than 60 Hz, which should not be a problem if the 120V's are rectified and used as DC. But it might be a problem if the device uses the 60 Hz to do some type of timing/synchronization.

A motor/generator pair would also have the most graceful failure mode. If either the motor or the generator fails, the output voltage will gradually drop to zero. But with an inverter, you just can't know ... It could supply same phase 120V to all of the outputs, or some phases might disappear, leaving only one phase or two phases active. There could be other failure modes that cannot be predicted up front.

[taydin]

The motor/generator would also work reliably through load spikes, with the output voltage slightly dipping.

But an inverter could cut its output in response to overloading, and if overloads happen frequently, it could even get damaged with an unknown failure mode ...

If these people go to such great lengths to keep this device running for more than 20 years, it's gotta be very valuable, so the power supply must never have a failure mode that would kill the device.

[Towger]

In terms of hum bars you would have to try it and see but at least it's not damaging to do so.

The hum on the earlier attempt of single/two phase is a symptom of the overloaded neutral.  Another cause is running all the live wires together and separately the neutrals together.

What is using all the power? Does it have an arc lamp in it?