Dual 20HP Idler RPC designed - Feedback requested. Build Photos to follow!

Hey guys! Just wrapped up the wiring diagram. Motors and all the parts are here. We have a GIANT cabinet and a nice rack made. About to kick this off and get it running. The info is in the diagram... Let me know if i'm missing anything.

Big thanks in advance to P. Haas for the diagram from 2007 that this is based on!




Main breaker box that feeds this is in the same building. All wires outside the cabinet will be in conduit. I have a big safety switch I can use outside the box if needed. Breaker feeding this is a 125 AMP 2 pole.

Isn't a 125 amp single phase breaker the recommended size for ONE 20 hp RPC?

Rob F. said:

Isn't a 125 amp single phase breaker the recommended size for ONE 20 hp RPC?

Click to expand...

I'll have appropriate sized wire per the NEC so the 125A breaker protects my wiring. Im aware I may not be able to absolutely 100% load these motors to their limit with this power feed. It may not be of much value to evaluate that breaker's size, as I can't get a single phase breaker over 125A anyway (within reasonable effort/cost) :/

RogueFab said:

I'll have appropriate sized wire per the NEC so the 125A breaker protects my wiring. Im aware I may not be able to absolutely 100% load these motors to their limit with this power feed. It may not be of much value to evaluate that breaker's size, as I can't get a single phase breaker over 125A anyway (within reasonable effort/cost) :/

Click to expand...

Yes, but why do you want dual 20 hp RPC's if you only have juice to feed one? Just to have a back up wired in?

Rob F. said:

Yes, but why do you want dual 20 hp RPC's if you only have juice to feed one? Just to have a back up wired in?

Click to expand...

My research showed that 125A will start and run both of these idlers. I can run one or both with this setup, and choose between those two options based on operating performance. I have an amp meter built into this system, and I can easily add another on the input side to monitor the draw on the 125A breaker.

This thread was intended to focus on my RPC design. Not my choice of feeder circuit overload protection. But we can go through that if desired...

If it were me I would split the run cap bank, one for each idler, to allow the system to stay in balance with either idler or both running, and not provide all the cap correction through one idler breaker. The individual idler cap banks could be placed locally at each idler motor to eliminate reactive current on the idler wiring and contactor. The cap wiring shown needs to be made, so that each run cap is connected to a different phase, not the same phase.

For the start cap bank,the resistors should be placed across each cap instead of across the bank. Much less current for the individual resistors to drain, longer life and smaller wattage, higher reliability.

I would want the start cap bank output to each idler contactor, instead of the panel bus. That way you could isolate one idler breaker and still start the other one. I would also want to provide a switch to select which idler is to be started, and connect the potential relay to that motor circuit.

Providing a separate 15A 3P breaker for the metering would provide better protection for it, versus a 60A one, and again provide ability for either idler breaker to be disconnected, and the other one to still be functional with metering.

On the control circuit I don't see a latching contact on the idler contactor. So that would meant that you would need maintained button operators to keep the contactors on. Not a good design choice. During a power outage the controls would be energized on resumption of power. Better to use contactor with 4 poles and use one pole as a holding contact.

Better yet would be to use motor starters with overload protection for each motor. NEMA starters are a better choice since IEC overloads normally have differential protection, and trip when there is a current imbalance between phases. And this is a normal occurrence on a RPC.

SAF

SAF said:

If it were me I would split the run cap bank, one for each idler, to allow the system to stay in balance with either idler or both running, and not provide all the cap correction through one idler breaker. The individual idler cap banks could be placed locally at each idler motor to eliminate reactive current on the idler wiring and contactor. The cap wiring shown needs to be made, so that each run cap is connected to a different phase, not the same phase.

For the start cap bank,the resistors should be placed across each cap instead of across the bank. Much less current for the individual resistors to drain, longer life and smaller wattage, higher reliability.

I would want the start cap bank output to each idler contactor, instead of the panel bus. That way you could isolate one idler breaker and still start the other one. I would also want to provide a switch to select which idler is to be started, and connect the potential relay to that motor circuit.

Providing a separate 15A 3P breaker for the metering would provide better protection for it, versus a 60A one, and again provide ability for either idler breaker to be disconnected, and the other one to still be functional with metering.

On the control circuit I don't see a latching contact on the idler contactor. So that would meant that you would need maintained button operators to keep the contactors on. Not a good design choice. During a power outage the controls would be energized on resumption of power. Better to use contactor with 4 poles and use one pole as a holding contact.

Better yet would be to use motor starters with overload protection for each motor. NEMA starters are a better choice since IEC overloads normally have differential protection, and trip when there is a current imbalance between phases. And this is a normal occurrence on a RPC.

SAF

Click to expand...

The run cap changes make sense. Thank you.

I already have bleed resistors sized for shared use like drawn. I'll have to order more smaller ones to do this, but it makes sense.

Changing the start cap bank makes sense if i will start idler 1 or idler 2 by themselves. My plan was to start idler 1 every time and use it to start idler 2. Do you not think that would work? This design wasn't intended to start idler 2 by itself (and it definitely won't!).

I like idea of the metering on its own small breaker. Thank you.

I will look into the motor starters vs magnetic contactors.

Thanks

RogueFab said:

The run cap changes make sense. Thank you.

I already have bleed resistors sized for shared use like drawn. I'll have to order more smaller ones to do this, but it makes sense.

Changing the start cap bank makes sense if i will start idler 1 or idler 2 by themselves. My plan was to start idler 1 every time and use it to start idler 2. Do you not think that would work? This design wasn't intended to start idler 2 by itself (and it definitely won't!).

I like idea of the metering on its own small breaker. Thank you.

I will look into the motor starters vs magnetic contactors.

Thanks

Click to expand...

SAF has done a useful "all one place" post. VERY useful!

It isn't all that different from the FOUR idler rig I've been re-doing my own one into. I do more tests and experiments than making of chips, so flexibility is key.

Here's a coupla kinda "borderline" possibilities to add to the mix that you might find of use:

NFPA 70 is a FIRE protection (and reduced risk of electrocution) Code. Insofar as possible, it stays TF out of what goes-on INSIDE the enclosures of serv-ED "systems". That's UL /CSA / TuV turf .. if any certifications are sought at all.

So.. "I could be (in the) wrong",, but the way I choose to interpret that for my use, is to carry multiple, individual runs of "lesser" breakers from the primary Square-D QO Load center to my idler's. Largest a 10 HP, then 2 X 7.5 HP, one X 3 HP.. which is basically a "trimmer".

Now.. RPC as they are, the two feeds from singe phase are common in copper to two of the 3-Phase legs.

Which means.. that I MUST have BOTH the contactors present to interrupt input side AND output side, and-then-also the reliable control logic to insure it actually OPERATES that way.

Then, too - given "shit happens" - alarm / monitoring to detect a fault in the anointed scheme.. and do "something safer" should any part of it fail to operate according to plan.

EVEN IF... that emergency response is the classical "crowbar" circuit .. that trips the breakers with such a vengeance it would be prudent to replace them after the fault is cleared.
Not a choice I expect to have to implement, but neither would it be "new ground" on a Telco pension!

Shorter take for your two 20 HP units is that I would be feeding them from TWO upstream breakers rather than one, making the investment in extra contactors ... plus not terribly complex logic. You don't need as much of that for two idlers as I do for FOUR ..

Why wuddja BOTHER?

Redundancy. Independence. Flexibility. Done and tested Day One to a plan.

Fewer potential surprises, later?

Done the way I am going, I don't end-up with a single, complex, 4-idler animal.

I have instead a collection of independently but "pre-wired as cooperative" modules I can not only put online in any combination as suits. I can re-purpose, re-locate, re-assign, dispose of, or swap independently.

Annnd my "worse case" starting load is only 10 HP for a rig that as presently enabled, tops out at 28 HP "worth" of idler-power. I'd have done your 2 X 20 as a 20 and two 10's, for example.

At which point, you have 20 or 40. but I would have 10, 20, or 40. Most often utilized? Sod's Law sez it's the 10 HP!

4 Idlers worth. Because I can.

CEO, C&W Mercury:

Ich: "Ian? You are running the whole show. How the Hell did YOU end up stuck with PABX extension: "666"?

Ian: "I got first choice!"

I made some revisions. I addressed the incorrect configuration of the run capacitors. I added thermal overload between the contactors and idler motors. I isolated the metering to be run off of it's own little breaker.

I left the bus feed from the capacitor generated phase because my only other choice is to have one motor feed the bus instead, and then still have the other motor feed the bus through a breaker. This is truly a "size of one, half dozen of another" call to me. I could also have a 2 pole breaker for the run capacitor bank and feed the bus through the other pole on it's breaker. Many options, none having much of an impact. Let me know if you have an opinion/argument otherwise.

Thermite - This has been a long process of acquiring the components when I could and for the right prices. In a perfect world, I would have a $5000 budget for this, and run brand new baldor motors in a 10/15/20 combo all driven off their own breakers, brand new motor starters, etc. I agree with you on the benefits of dissimilar idler sizes. I got the motors and contactors a while ago for a great price, so the build is based on the components already here. We already ran conduit for the wire to feed this RPC for a single breaker. We currently only have a 5 HP lathe to run (Takisawa TSL800D), so we will only be using one idler for a while. Down the road we can split the system into 2 feed breakers by adding another start cap bank and cutout relay/contactor if we end up tripping the 125A main feed breaker.

RPC Building Stratigies Articles

Here are you some articles for you to brush up on, for designing and building RPC's, a good blast from the past.

Look for the Fitch Williams Instruction PDF download link in one of them, He preceded Peter Hass in these discussions.

Another Explanation of RPCs, and Their Capabilities

Three-Phase vs. Single-Phase, Illustrated

Balancing vs. Power Considerations

Fitch Williams 10hp RPC questions

Two Stage Rotary Phase Converter

Is there any reason not to use the neutral on a machine supplied by a rotary ph.conv.

RPC Topologies Using Potential Relays

SAF Ω

RogueFab said:

Thermite - This has been a long process of acquiring the components when I could and for the right prices. In a perfect world, I would have a $5000 budget for this, and run brand new baldor motors in a 10/15/20 combo all driven off their own breakers, brand new motor starters, etc.

Click to expand...

In a "perfect ENOUGH world" the only Baldor motor under roof was an unfortunate accident bought "NOS" and still in the original factory crate

Baldor built it in Mexico... to be branded for a name that was still reputable long after they had trashed their own reputation: T.B. Woods.

I can't be bothered to even make connections to a Baldor motor!

The replacement for the tasking is a used Reliance RPM III that came out of a Hershey chocolate production facility - still has the Hershey property tag on it.

Back to the AC side, a Weg is ordinary "white bread" but better value for money than present-day disposable Baldors.

The brute-force and heavier than sin finned-case Reliance are still good. They go cheaply even new/NOS/New No Box (crate).. because no one wants to pay as much in freight as the motor's cost! Around $150 - $280 each, new, sometimes with minor cosmetic damage to a fan shroud or such. Easily fixed. Need not be "pretty", just functional.

I'm a retired finance guy. I just kid myself and book freight, rigging-in, or go-fetch trips on a different scrap of paper.

Then throw it away as an "unrecoverable" cost anyway!

BFD. I had to "be SOMEWHERE" regardless. Why sweat the petty details of it?



NRi = and"not only" have loads of decent switchgear and starters. Some of mine is out of Canada, Square-D, and "600 Volt class". It was cheap for lesser demand than 300 Volt class. Used goods pulled-out of shut-down factories and commercial buildings renovated or demo'ed are just NOT re-used by first-tier commercial installers.

So long as the guts are good - as all HAVE been - I don't mind removing labels & plates or prepping and spraying an enclosure.

Is there a Lovejoy coupling between the (2) motor shafts ?

digger doug said:

Is there a Lovejoy coupling between the (2) motor shafts ?

Click to expand...

Where did he say he had MECHANICALLY coupled the idlers at all?

Surely nothing of that sort on MY radar! They sync perfectly. Nature of the beast.

Have a look at major commercial makers.

Once above a certain size, the wiser ones "go modular" anyway.

Staggered start no matter how widely spaced in the time domain, they are GOING to be in-phase, first time, every time so long as one wires them in phase at the outset.

SAF said:

Here are you some articles for you to brush up on, for designing and building RPC's, a good blast from the past.

Look for the Fitch Williams Instruction PDF download link in one of them, He preceded Peter Hass in these discussions.

Another Explanation of RPCs, and Their Capabilities

Three-Phase vs. Single-Phase, Illustrated

Balancing vs. Power Considerations

Fitch Williams 10hp RPC questions

Two Stage Rotary Phase Converter

Is there any reason not to use the neutral on a machine supplied by a rotary ph.conv.

RPC Topologies Using Potential Relays

SAF Ω

Click to expand...

Thank you! I have the "new topology" one in my bookmarks already, that is the best article/write-up I have found so far. The one written by FRW (the PDF) is so neat and easy to follow and the balance process is VERY detailed. Thank you again.

I revised the layout again, I split the balance caps into 2 sections, one for each idler. There would have never been a decent no-load balance for the system while running in both conditions otherwise (both conditions meaning one or both idlers running).

Switches are IDEC (good quality), just FYI.

I will be balancing the system at the equipment if the variation is outside of desirable balance once we get it installed and running and load it up. For the reasons cited in many of the referenced write ups. In summary, you can't have the system balanced for every combination of loads. So it is best to balance the system in it's various idler operating configurations (just 2 in my case), and then balance the system further with the different combinations of loads possible. This latter balancing is done at the equipment (so I will add capacitors inside my lathe cabinet to balance that load, for example). With each combination I will verify that the voltage balance is within acceptable limits. This would be a lifetime of work to get perfect, but shouldn't be too bad to get within reasonable limits like described by FRW! At this time we have a 5HP lathe with a 3 PH coolant motor, a 2 HP mill with servo feed, an auto saw that has 2, 1, and 0.5 HP 3 phase motors, and we also have a 2 HP cold saw that will be coming online soon. In the future, we will get our 5HP auto feed cold saw (another 3 or 4 motor machine) up and running, and it may be ran off of a buck transformer if I Can't convert it down to 230V! It will be a LOT of equipment.

New layout below.

RogueFab said:

I will be balancing the system at the equipment if the variation is outside of desirable balance once we get it installed and running and load it up. For the reasons cited in many of the referenced write ups. In summary, you can't have the system balanced for every combination of loads.

Click to expand...

Yer over-thinking it. Greatly.

You have a minimum 20 HP idler, and - at present - only a 5 HP primary load motor for the lathe.

Your DRAW will be all over the lot as it goes from idle to in the cut. Light cuts and finish cuts and tedious stuff, but low HP draw, such as single-pont threading DOMINATE over heavy cuts anyway. Coolant pumps and such actually help stabilize, not worsen.

Bottom line "One" is that a 20 HP idler is largely going to ride right through the small variations. Any heavier loads you add later are for future tuning.

Bottom line "Two" is that so long as you have the "general goodness" of an RPC, not static converter, even rather large imbalance, generated leg, isn't all that big of a problem for a motor run well under its rated max load. Heating is out of balance. BFD so long as run at only partial max.

Short Answer: Don't waste a lot of time on trying for perfection in balancing unless you have a steady-state load such as an HVAC air-handler or an irrigation pump.

With about a 5 HP + aux load, VARIABLE load?

You'd be FAR better-off to put the effort into including something less than a 20 HP idler. A 7.5 HP or a 10 HP, for example.

thermite said:

Where did he say he had MECHANICALLY coupled the idlers at all?

Surely nothing of that sort on MY radar! They sync perfectly. Nature of the beast.

Have a look at major commercial makers.

Once above a certain size, the wiser ones "go modular" anyway.

Staggered start no matter how widely spaced in the time domain, they are GOING to be in-phase, first time, every time so long as one wires them in phase at the outset.

Click to expand...

Why not ?.....don't follow....LEAD the pack.....

digger doug said:

Why not ?.....don't follow....LEAD the pack.....

Click to expand...

Because an already wasteful rig gets stupifyingly inefficient if it has to haul an unpowered idler around.

RPC are OLD technology. Any "magic" wudda been standardized, long since.

If you already have the components, take some plywood and "breadboard" it out and try it.

"Down the road we can split the system into 2 feed breakers by adding another start cap bank and cutout relay/contactor if we end up tripping the 125A main feed breaker". I do not follow you there. With the original potential relay in the system when you connect the second 20 hp motor, one of three thing will happen. #1- it starts right up and you are good to go (unlikley) #2 the breaker trips from the inrush current of added the the preexisting load. #3 the inrush current drops the voltage and triggers the original potential relay and the original bank of start caps kicks in. At that point it either starts or you go to #2. The point is, you should not need a second band of start capacitors.

Here is some food for thought to avoid the high inrush current with the second 20 hp motor starting. Use the running rpc to feed a VFD to slowly bring the second motor up to just under the speed of the RPC. Have a voltage sensing system to detect when the voltage between the RPC leg 1 and the VFD leg 1 is minimal and then have it turn off the VFD first and then connect the mains to the now synced second idler. No- I have not done this- but I have used an undersized single phase VFD to bring an idler up to speed, then manually turned off the VFD and connected the idler. It worked as an "I wonder if" experiment.

I see something in that design I don't like. It appears that single phase is passed to two legs of the three phase machinery when the RPC is not on.

I designed my RPC around a common three phase bus. The bus had single phase power connected directly to it. Each idler motor was connected to the bus through a contactor (interlocked with the pony motor contactor, if applicable), and the output was also connected through a contactor.

The output contactor could only be initialized if a third phase was detected.

The system was very simple to construct, easy to balance for different configurations, reliable, and easy to add to or modify. I really ought to draw out the schematic.

Strostkovy said:

I see something in that design I don't like. It appears that single phase is passed to two legs of the three phase machinery when the RPC is not on.

I designed my RPC around a common three phase bus. The bus had single phase power connected directly to it. Each idler motor was connected to the bus through a contactor (interlocked with the pony motor contactor, if applicable), and the output was also connected through a contactor.

The output contactor could only be initialized if a third phase was detected.

The system was very simple to construct, easy to balance for different configurations, reliable, and easy to add to or modify. I really ought to draw out the schematic.

Click to expand...

Please do so!

I think I've done the same, near-as-dammit exactly, using the SQ D QO panel's busses as common meeting point.. except that I've presumed 3-P WILL happen when I ask it to do, rather than probing to confirm it actually HAS done.

Given my "drawing" skills are still in the India ink on vellum age, if your one is done "on screen" it would be easier to publish over the wire!

And safer, yet than my assumptions?

TIA