Flywheel on a rotary phase convertor

Mine is! Works fine.

Hanrahan seems to think it is the case. The presence of the capacitors notwithstanding, his motors are connected in parallel.

The part that disagrees with an acknowledged authority, Mr. Hanrahan:

I did. So did Hanrahan. His are in parallel.

He did.

Convoluted currents? Yeesh! RPC's aren't witchcraft and wizardry, mesh currents are mesh currents. The (parallel) caps provide some phase correction and resonant voltage boost to compensate for the reversal in I-Z drop because the third leg of the idler is a driving leg rather than a driven leg. They are not essential to the operation of an RPC, though they can improve balance.

My RPC has no capacitors, starts and plug-reverses my lathe and mill just fine.

The key point here is: "How my phase converter is wired when it runs a load motor". The flawed logic is that a phase converter does not "run" a load motor. Such a differentiation between phase converter and load motor is not possible. An idler motor and load motor, taken together as a network, are what constitutes a rotary phase converter. This all harks back to the misunderstood "generator" concept as applied to an idler motor. The two must work together to form a rotary phase converter. Remember a RPC (the whole RPC) acts to manipulate current flow in a network so that the load motor voltages, and currents, are the same as if the load motor was operating from a 3-phase source. Remembering, all the while, the whole thing is running on *single-phase* current.

Now as for parallel connectivity: The drawing is that of a classic RPC (idler and load motor) operating from Hot 1 and Hot 2, both sides of a single-phase source. Consider Hot 1. It connects to L1 of the idler motor and also to L1 of the load motor. The same can be said for Hot 2 and the 2 respective L2's. By definition, the points designated as L1 and L2 in both idler and load are operated in parallel across the line. That is to say the

2 main windings, L1 to L2 in your drawing of both idler and load are connected in parallel across the line.

Now look at the way current flows in the L3 lead. The idler's L3 wire has 1 end connected to the 3rd. winding on the idler, call that a source point for L3 current flow. The other end of wire L3 connects not to the analogous same start point on the load motor but to a point on the other end of the load motor's L3. This is not, can not, be considered a parallel connection. The only way the two 3-phase motors could be connected in parallel is if they are both fed from a 3-phase source.

In a manner of speaking, for a RPC (network) to do its thing, when taking in

*single-phase* current and delivering *3-phase current* to a load motor (also part of the network) there has to be current flow in 2 directions the 3rd leg. Of course, aggregate current flow is such that current will flow in the "right" direction in the load motor's 3rd leg. Parts of the RPC act as both generators and consumers, thus the heavier element will cause current to flow, seemingly backwards, into the other element. And so it is with the RPC - capacitor augmentation can enhance the convoluted current flow in such a way as to make emulated 3-phase current flow in the load motor.

Bob Swinney

Wait. Don and Jim.... separated at birth?

:^)

Jim

A phase converter could run a resistive load just as well. I use mine to run my welding machine. All such loads draw power from three legs L1, L2, and L3. That the "load" is a motor is of no great relevance to the fact that there is voltage produced on the third leg.

Well, duh.

Yes.

Well, connection of L3 to L2 effectively parallels windings L1-L3 and L2-L3 between the two motors. Points L1, L2 and L3 have the same potential, respectively, on both motors, because they are connected by wires (ignoring irrelevant issues like voltage drop on the wire).

Perhaps you and other participants mean different things when they say a parallel connection.

i

Ah, this *is* a semantic minefield. I think I see your point finally.

They're not, and cannot be in parallel, if one thinks that being in parallel means they each have that extra wire there. The third leg doesn't, it's missing the extra external connection that the line wire represents.

From a rough electrician's standpoint consider what the wiring looks like when I have my drum switch turned on at the lathe, but the converter is not energized at the knife switch on the wall.

L1 and L2 are missing. There's no external current source so the third leg is now identical to all the others. At that point then they truly are in parallel - from an electrican's view as well as a EE's view.

When the converter is operating of course there is one special lead that breaks the symmetry - it's missing the line connection. An electrican would say that the absence of that line connection does not change the fact the two sets of windings are in parallel. A EE looks at the entire network as a system, including the incoming power. He says parallel means all nodes have the same number of connections.

Jim

I'm glad I have VFD's and don't have to use so much brainpower figuring out how they work (:

Well, you're reading a lot more into the Hanrahan drawing (conv/fig 1) than I can see. Hanrahan didn't show the load motor, only assigned terminal numbers 1, 2 and 3 to it. I would have to assume he meant T1 and T2 on the idler went to windings 1 and 2 of the motor. And it would be a fair guess he intended for those same connections to go to windings 1 and 2 of the load motor; granted those would be in parallel. But T3 on the idler, let's call that an "outside" end of the winding for the sake of convention, goes to 3 on the load. Wouldn't it be a safe guess then that 3 on the load is an outside end of that coil also? I believe the confusion comes from the fact the 3rd leg is both source and load in a RPC instead of a single lead tied back to a common source as in a true parallel arrangement. Hey! You almost got it below, when you said, "because the third leg of the idler is a driving leg rather than a driven leg." The third leg is both a driver and a driven leg, plain and simple. The 3rd leg is a part of a complete network (mesh it if you like) and it cannot be separated into a generator only portion of the RPC. The RPC consists of an idler and a load with the 3rd leg serving as both generator and consumer - not server only as if it were a simple parallel connection.

See the explanation given to Iggy, copied here for your convenience:

""Now look at the way current flows in the L3 lead. The idler's L3 wire has

1 end connected to the 3rd. winding on the idler, call that a source point for L3 current flow. The other end of wire L3 connects not to the analogous same start point on the load motor but to a point on the other end of the load motor's L3. This is not, can not, be considered a parallel connection. The only way the two 3-phase motors could be connected in parallel is if they are both fed from a 3-phase source.

In a manner of speaking, for a RPC (network) to do its thing, when taking in

*single-phase* current and delivering *3-phase current* to a load motor (also part of the network) there has to be current flow in 2 directions in the 3rd leg. Of course, aggregate current flow is such that current will flow in the "right" direction in the load motor's 3rd leg. Parts of the RPC act as both generators and consumers, thus the heavier element will cause current to flow, seemingly backwards, into the other element. And so it is with the RPC - capacitor augmentation can enhance the convoluted current flow in such a way as to make emulated 3-phase current flow in the load motor.""

As for wizardry and witchcraft, I see none in the operation of a RPC, balanced or not. Try to embrace the idea of the 3rd leg as both generator and consumer, and you'll have less trouble when you try to mesh things out. I will leave the horrendous math up to you as you seem to have already figured it out.

Bob Swinney

Jim sez"

Jeeze, Jim! Thanks, I think. And just about when I pictured you with your ball cap turned around the right way and it said "Gary Coffman" on the bill.

Bob Swinney

I always thought that gary's insights to matters like this (EE related) were typically very valid. And he was the one who suggested that flywheels probably hurt rather than help.

I think that gary would see your "not really parallel" view of life and agree mostly.

But honestly, whenever I have to tell somebody how to "make" a phase converter, they're typically electrically-savvy folks who want to run a machine. The easiest way to explain it to them is to say "you put the two idler motor leads across the line, and then you put the load motor smack dab in parallel with the idler leads.

Basically there's a one-to-one correspondence between the leads of the idler and the load motor. No wires left over and all that sort of thing. Then they say "but what about the fact that the idler has the two incoming line leads hung on it?" and then I say "don't worry about those, the thing'll work just fine with them there."

Red to red, black to black, blue to blue. Electricians like color codes. This is why I drive them crazy when I wire my motorcycles with all white wire. Makes drawing the diagram easy. "What color is

*this* wire?"

"White."

Jim

He does? In circuit analysis, a set of two-terminal networks are regarded as "in parallel" if they are each connected to the same pair of nodes so the voltage across them is identically the same. Similarly, a set of n-terminal networks are in parallel if they are connected to the same set of n nodes so the various inter-terminal voltages on each n-terminal network are identically the same for corresponding pairs of terminals. This is regardless of whatever else might be connected to, between or among those nodes and regardless of any external symmetry or lack thereof. Some nodes may well have more connections than other.

By this definition, if there are wires connecting each terminal of one device to a corresponding terminal of another device, they are in parallel -- regardless of what else might be connected to those terminals.

The roles of "generator" and "server" just have to do with direction of current flow, in a parallel connection or otherwise. For example, the elements of a parallel-resonant circuit alternately act as energy source and energy sink. Even though they have the same voltage impressed across them, they may (usually do) have currents flowing in opposite directions.

Why not?

How they are (are aren't) fed by an external system does not change the way they are connected! "Parallel connection" is a matter of physical topology.

Yet it isn't a generator. Hmmm......

It's threephase with or without the capacitors, though the phases may be unbalanced with different-than-ideal phase relationships and magnitudes. The caps just improve balance under a given set of conditions.

I think we each have things figured out, though clearly not in the same way. There are lots of ways to look at things. The concept of rotating fields is a fiction and an artifice, as is the practice of dealing with complex impedances using the (imaginary) square root of -1 as EE's are so fond of dong. There are lots of ways to think about what's going on. They aren't necessarily mutually exclusive.

Sorry, Don. The lead between the 3rd leg terminals does not place them in parallel with each other because it connects from one end of one winding to the *other* end of the other winding. Voltage symmetry and a parallel connection as commonly defined cannot exist under those conditions. By your definition below, the wire does not connect one terminal of one device to a corresponding terminal of the other device. Granted, the 2 line terminals do meet that criteria but the 3rd leg terminals do not. Thanx, Jim for taking Don by the hand and trying to help him out. I think he is getting there. At least he has come up with a plausible definition of "parallel". That's progress!

No. A RPC is not a three-phase device in the classical sense; It isn't fed with 3-phases and it does not "generate" 3 phases as would, say a 3-phase alternator. True, the currents circulating in the load motor may make you think they are 3-phase but that is because they are the products of a special network. That network is *not* comprised of motors having all 3 sets of leads connected in parallel. Because of the way it is connected, that network has the capability of taking single-phase current into two

3-phase motors and delivering currents that emulates true 3 phase current. Key to accomplishing this is the non-parallel connection between the 3rd legs.

Bob Swinney

Suppose I take a three phase motor and use some method to get it running on a single phase source, and then connect the three leads of the motor to a three phase rectifier bridge. Do I not have three phase power going to the rectifier bridge? Is my ripple current not pretty much what I would expect if I had the rectifier bridge connected to the output of a three phase generator? It seems to me that a RPC does not need a load motor to work.

Dan

Robert Sw>

You are right, it does not need a load motor. You could run a resistive load (a 3 ph heater), or, like I do, a 3 phase welder that starts with a transformer and rectifier.

i

It is. Same thing. Ampere's law. Field is proportional to current.

Ah, the *other* end of a three-terminal network. I have so much to learn!

There's only three freakin' terminals, there ain't no *other* end. Further, parallelism has no requirement of voltage symmetry. Parallelism, in conventional terms, is a matter of topology. Further, internodal voltages are what they are, however labelled, symmetrical or not, whether or not they match your expectations. Terminals connected together are nodes by definition -- in conventional terminology. If the labelling doesn't match then the labelling is wrong. The issue here may be "as commonly defined" as that doesn't match your theory. Progress here: it's good to clarify where your theory deviates from common definition and conventional engineering terminology.

Yes, thanks Jim! I particularly like the idea of using all white wires.