[How do I?] My 2017 Rotary Phase Converter Build

7thKahuna

Registered
Registered
Joined
Dec 14, 2016
Messages
13
Hello all. I am new to your community and impressed by the range of experience and knowledge being shared here. Hopefully, in time, I will be able to contribute something useful. I am however quite new to metal working. My grandfather was a dedicated woodworker who built and raced wooden sailboats. He had friends who handled most of his metal work.

Last year, a friend of my dad's offered to sell him a lathe and a small mill. The machines were 3 phase and dad was under the impression that there was some way to adapt the wiring to get them to run on single phase. He still swears an old electrician at work had a way (isn't that the way it goes ;)). I knew in practice it was more involved than that, but that was about where my experience ended. In the apparent misconception that I knew something about what I was talking about, dad asked me to build him a phase converter. Ok, why not. My research lead me to rotary phase converters and ultimately to a YouTube video posted by a member of this site (marcaap). His video appears below.


The video caught my attention because the design appeared well thought out, nicely executed, and was working to run a lathe and a mill. Ha, just what I needed. Marcaap (Paul) had also shared a schematic of his design. I came up in 4-H and teen lead the electricity project. Give me a plan and I can build it and survive to tell the tale. I am not however an electrical engineer and by no means comfortable with 3 phase power. I didn't at first understand Paul's design. That was a couple months ago. I continued my research.

Last month I watched the video again and this time I got it, but simultaneously I discovered that the relay specified in the plan (Steveco Potential Relay 90-66) had been discontinued. I turned to eBay and was immediately rewarded with a NOS 90-66 still in the box. I decided to go ahead and source the other components and posted a request in Paul's thread asking if he could repost the parts list, just to be sure I had things right. I haven't heard from Paul but several other members chimed in.

http://www.hobby-machinist.com/threads/video-of-7-5-hp-rotary-phase-converter.14864/

Ken (4gsr) and Keith Foor (Keith Foor) have both added to my understanding of RPCs but Keith's suggestion that I replace the potential relay (the one part I had :D) with a timer relay has meant a return to the drawing board. Toward that end, I decided I better start my own thread rather than continuing to hijack Paul's.
 
Last edited:
Before I start asking about momentary switches and timer relays, I have another question. Last week, after several unsuccessful attempts, I finally secured the 3 phase idler motor. Actually I secured two, both Baldor, a 5 hp and a 7.5 hp, $75 each. At that price I was expecting to have to rebuild them, but they appear to be in great condition. I was looking for a 7.5 hp motor because they seem to be the most common among the threads I have read and will provide a good deal of flexibility in the shop. Paul actually notes that his design was intended for 10 hp but he ended up using a 7.5 because that was what he had. I assume that just meant adjusting the capacitors. As I understand it, the other components remained the same.

Ok, so here is what I think I know. If I am running a 7.5 hp idler motor, I should be able to easily run a 3 hp lathe. Following Keith's instruction I understand that I could actually use the two motors together (7.5 + 5 = 12.5), start one with the other, and I could comfortably rip oak planks on a 5 hp 3 phase table saw. Only thing is, right now we have neither the 3 hp lathe nor the 5 hp table saw. Currently our largest 3 phase motor is 1 hp. That will of course change in time. :cool: If we direct this conversation to amps for a moment, is there enough of a savings between the 5 hp 1750 rpm motor and the 7.5 hp 3450 rpm motor to consider building the RPC with the 5 horse motor? Per the tag we are talking 14.2 amps vs 17.2 amps, a difference of three amps. (I'm not running on batteries or trying to make it home from the moon.) My understanding is those are max values, occurring when the motor is under a load. This motor would not be, but then I am not using the motor as a motor so perhaps it's different. Is there any reason to go with the 5 hp over the 7.5 hp? My inclination is to go with the 7.5, and I wouldn't have given it any thought except for someone's comment about needing 125 amps to start a 40 hp idler motor. Paul's plan indicates a 50 amp double pole breaker supplying the RPC and a series of 30 amp internal fuses, but again, I believe that was designed to support a 10 hp RPC.

I understand there will be a spike in amps to start the idler motor (approximately twice the full load amps per Keith) then it should drop off. After that, is it just the amps drawn by the equipment plus a minimal draw for the idler and any heat loss in the system?
 
The follow up question to that is, what happens if I walk out of the shop and forget to turn off the RPC? Do I risk damaging the converter if I leave it running without a load? Do I risk dad dropping dead when he sees the utility bill? I'm planning an indicator light near the door but the RPC will actually supply two different spaces so the light is of limited value. What I need is a design for a circuit that will shut the system down if it hasn't sensed a load for a set period of time, say 20 minutes.

I'll leave that idea for another day.

So what do you say, stick with the 7.5 or start with the 5.0?
 
Tonight I sat down to figure out the capacitors. My research has pretty consistently suggested 50 to 100uF per unit of hp for the start capacitor. Paul's design is about 63uF (475uF total) assuming his schematic was modified for the 7.5 hp motor he ended up using. Ken too suggested it looked about right. If I use five 100uF capacitors I'll be at 67uF per unit. I understand that too many uF isn't as much of an issue as too few.

Now the run capacitors aren't as clear. I've read that 12 to 16uF per unit of hp is appropriate but I've also read it should be 25uF to 30uF. Paul's schematic, if I am reading it correctly, shows 95uF between T1 and T2 and 160uF between T3 and T2. He also has a 40uF 'PF' Capacitor between T1 and T3. I figured out what PF stood for, I've forgotten. If I total all three of those together, I get 295uF but in Paul's comments, he says he actually used 340uF. Now 295uF is 40uF per unit of hp while 340uF is 45uF per unit of hp, well above any recommendation I have seen. Is there any rule to this? If, as I understand, it is custom tailored to each installation, where should I begin? Does the number of poles impact this calculation?

I understand that balancing out the system is somewhat a matter of trial and error and unique not only to the idler motor but perhaps also the equipment that is attached to it. So when it comes to ordering the run capacitors is it reasonable to order maybe one each 60uF and 100uF and several 40uF, or should I expect to order a handful of lower uF capacitors to make finer adjustments. 25uF? 10uF? 5uF? How small an adjustment should I be prepared to make?
 
As a young kid I remember hearing about the rocket graveyard and other industrial salvage places around Los Angeles. Are they all gone now? My google searches are coming up bare.
 
Sorry if I end up asking too many newbie questions. There is so much 'seemingly' or perhaps 'factually' conflicting information out there.

For anyone following along, do a Google search for "FitchWConverter.pdf". You should find a link to a helpful document that goes into the balancing process.

Of course caveats appear to apply. Rich Carlstedt, an experienced RPC builder makes the following comments with respect to the document:

I would say to forget your voltage readings.
You want "AMPS"
That's what does the work.
Voltage is only "Potential" , not HP
Balance the legs with amp readings, as Voltage numbers will drive you nuts
I use 15 MF per HP as a start
(for the run capacitors) and go from there.
I found motors to vary tremendously and you will be tuning to them specifically
Older motors with more mass are easier. Some of the "efficient' motors of late are NOT easy !
If you want to fine tune , Select the most prevalent load and use it.
Do not try to do both low and high speeds on your lathe for example, it will be different .

If this is a RPC , you can look at Volts, but as soon as you drop your slave motor in the circuit, it will change
Then you can add Caps to the slave motor to be rebalanced. This is especially needed if the RPC runs multiple motors either single or at the same time

In line with other material I've read, Rich suggests 100 MF per HP for the actual start capacitor and 100,000 ohms at 1 or 2 watts for the drain resister. 100,000 is more than I recall seeing elsewhere. Paul suggests 75,000 at 4 watts, Fitch employed 15,000 ohms at 2 watts.

At this point, I can't comment one way or the other. I guess maybe I need to keep my eyes open for a clamp on amp meter that is also capable of testing capacitors. My current meter does neither. Yeah, yeah, I caught that. :cautious:
 
Last edited:
OK, lots of questions.

TUNING
First thing with my designs are that everything I build is CNC rated. Meaning less than 5% voltage difference leg to leg.
Does it matter for older contactor/ switch controlled equipment with one or two basic motors and no electronics, it does NOT matter a lot as long as you are no over 10 % leg to leg different. For CNC gear, gear with electronic controls and the like it's very important.

.
If you have too little run capacitance, you will typically have low voltage, which means that you will also have low current. Yes, current does the work but voltage is the potential to DO work. If voltage decreases then so does current. Second, TOO much capacitance. This will generate higher voltages on the generated leg and will typically cause even more problems than too little voltage. And you can get the voltage on the generated very high on the generated leg if you get way overboard. I have seen 350 volts on a 3rd leg a couple times. That WILL fry things.

An RPC is a tuned circuit of sorts. ANd do to that getting it tuned right is important.
First reason it's important is that it will draw the least amount of current while idling. This means lower electric bills for you or whoever is paying them
Second is that a properly tuned RPC will start a motor at least twice it's size when that motor is not under load. I have spun up 40 HP motors testing them with a 15HP RPC. Mind you the motor was bare and sitting on the shop floor but it started the same as it did when I got it and tested it on the sellers 3 phase power in their building.

Now, the actual tuning part, or HOW TO TUNE an RPC
First part is that you DO need to use run capacitors between L1 and L3 and L2 and L3. The 9 to 15uF per horsepower of te motor is about right.
But tuning can be tricky. First thing to know is to move up and down slowly. Don't make 10uF jumps. I typically will have the figured amount of capacitance in the circuit and then have 3 additional capacitors for tuning. two are 5 uF and one is a 10 uF. Here is an important thing on capacitors. Wiring them in parallel increases the total value of capacitance, while connecting them in series decreases the total value. Two 5uF caps in parallel is 10 uF while 2 in series is 2.5uF. So base tuning can be done in 2.5uF steps from 2.5 uF to 20uF with only 3 capacitors.
First voltage to look at is the L1 to L2 voltage. This one will typically not vary as it's line fed and would require excessive current draw to get it to decrease. That is the voltage you are working to obtain for both the L1 to L3 reading and the L2 to L3 reading. Start with the L1 to L3 and adjust the capacitor value until the voltage matches the L1 to L2 voltage. Turn the unit off every time before adjusting the capacitor values. Once L1 to L3 is adjusted, work on the L2 to L3 voltage. Again, power down before changing out the caps. Once it's right, go back and check the L1 to L3 voltage. Changes are it moved up or down. Again, attempt to adjust your values to get it back to where it was matching the L1 to L2 voltage. YOU WILL NOT TYPICALLY GET ALL THREE READINGS TO PERFECTLY MATCH. Not possible so don't beat yourself up over it. So you are working for close as possible. After you go back and forward twice with it that's typically gonna be it.

START caps
25uF to 50 uF per horsepower and sometimes as much as 100uF per horsepower.
Your motor should take about 2 seconds to start. Anything more than that and you are just killing the capacitors and the motor. Remember that START caps are not run caps and can't remain connected to the motor after it's started. They will overheat and blow up.
Something else on RPC design. Install a contactor disconnection the output while the start capacitors are in the circuit. Reason is simple.
Remember where I said that too much capacitance will increase the 3rd leg voltages drastically. Well the run capacitors are about 9 to 15 uf per horsepower but the start capacitors are about 25 to 50 uF per horsepower. Guess what happens to the 3rd leg voltage during startup? If your RPC is connected to any device that has a constant connection to the input voltage, like a power supply to run the control circuits, it's gonna get that 300 or better volts put across it. Guess what happens to a device designed for 240 volts when you apply 350 volts to it. The magic smoke leaks out and it stops working. So output contactor is important.
 
Keith, thank you. I very much trust your contributions, even more so knowing that you build for CNCs. We do not have any electronically controlled equipment now but barring major issues, I think I will do likewise. This RPC is going to be around for a while.

"Two 5uF caps in parallel is 10 uF while 2 in series is 2.5uF."

Thanks for sharing that. Somehow I missed it. I knew you could add in parallel, I didn't know about 'subtracting' in series. What is the math behind that? Obviously 5+5=10. Is it then 5/2=2.5 in series? Can't be that simple. certainly a 10 and a 5 in series aren't 5 uF. 10+5 in series equals smoke? Logic might suggest 7.5 or zero but that logic doesn't work in your example. More research. At least now I can plan my capacitor order.

Remember that START caps are not run caps and can't remain connected to the motor after it's started.

But as I understand it run caps can be used as start caps. Any issues other than maybe cost? Can start and run caps be mixed in the start circuit?

Something else on RPC design. Install a contactor disconnection the output while the start capacitors are in the circuit.

Ok, just to be clear, use a contactor to isolate the RPC from any of the equipment it powers until the start caps have been removed from the circuit. Yes?

Do you recommend, or have any thoughts on, using 24 volts versus 120 volts to operate your contactors?

Again, thank you for you help. I know I'll have more questions, some that I feel awkward asking given that you raise funds building these things. My apologies in advance if I overstep my bounds. Hopefully others will chime in as well.

When it comes to timer relays, if the motor should start up in 2 seconds, how big a window should the timer allow? I assume you would keep it to 2 or 3 seconds, yes? I guess that will be my next task, identify the relay.
 
OK, first the capacitor math.
When you connect capacitors in series you are adding their voltage handling ability because you are increasing the distance between the plates inside he capacitors basically. The actual math is reciprocals. a 5uF and a 10 uF with be figured as 1/5 + 1/10 = 1/X or 3.33 uF. To know you are doing the math right realize that your total amount of capacitance will never be higher than the smallest capacitor in the series.
Resistors acutally add up the same way only in parallel. Series resistors add up x+y=z.

OUTPUT CONTACTOR
Yes, isolate the output when the start caps are in circuit to avoid high 3rd leg voltages being applied to the connected equipment.

If you use 24 volt coils on your contactors, you will need to have a power supply there to create the 24 volts.
This can be a pain but there are also benefits to it. The biggest one is you can connect the 24 volt power supply to the lighting circuit in the shop. When you walk out of the shop for the evening and turn off the lights, it will shut off the power supply to the RPC and it of course will shut down.

TIMING
I do my timing in an odd way. I will use a stop watch and a momentary button to control the start caps and so I can see whats needed.
I start the timer and push the button. mind you the button will also need to be wired in a way that it will engage the input contactor as well.
I relaese the button and stop the stopwatch once the motor is spun up. Set the timer for that amount of time is a good starting point and you may need to tweak it a bit from there but it will be very close.
Use Omron timers or equivlant octal base timers for control. Easy to wire, easy to mount and easy to find replacements when needed. These timers only have a 3 amp contactor rating so do NOT attempt to use them to directly switch the capacitors. Use them to control a contactor instead.

On a side note. You can use the start timer to control the output contactor as well.
Wire the NO or normally open connection to the start cap contactor and wire the NC side to the output contactor. That way the timer will run and open releasing the start contactor and then close the output contactor. This will ensure isolation.
I would also advise you to connect the run capacitors to the output side of that contactor and not the motor side. Having the run capacitors in the circuit during startup can cause issues with starting if you are running L1 to L3 and L2 to L3 run caps.

LASTLY
As far as asking question and overstepping some boundary. The only way that will happen is if you ask me to build one for free and supply the parts myself as well. Anything else is fair game. I do build and repair these for funding the hobby, but it's just a hobby and I would rather see a guy enjoy it than fight it because I didn't share what I knew. I will litterally give you my number and talk you through it wire by wire if I need to and not be put out at all by doing so.
What I ask in return is pay it forward with knowledge. We are not here as union labor millrites. This is not a competition to see who is better. We are ALL on here to share what we know, what we have done and to share knowledge. And I enjoy being on here for that reason. That math we talked about before. I have not actually discussed that piece of knowledge in 20 years. I just knew it form college and high school electronics classes and have used it ever since.
In other words, ask whatever you don't know and I will do my level best to help. Don't worry about boundaries, there aren't any.
 
Following on from Keith's post, caps in series or resistors in parallel multiply the two values together then divide that result by the sum of adding the values - so for a pair of 3uF caps in series 3×3=9, divide that by 3+3=6, 9÷6 =1.5 uF. Similarly for 2 and 10, 20÷12 = (roughly) 1.7 uF.
Resistors, same procedure, 2k parallel with 1k: 2k x 1k = 2k k, 2k + 1k =3k, divide 2k k by 3k (note k cancels!) = 2k÷3 =667 Ohms.

Hope that helps, doesn't confuse!

Dave H. (the other one)
 
Back
Top