I built a rotary phase converter using a 25HP motor. I am using 400uf Cp and
20uF Cs with no Cpf. It just barley starts my brand new Ingersoll Rand 10HP
compressor with an empty tank but will pump it up to 175psi. It won't start
the compressor when it cycles on at 135psi. The compressor turns slowly
and after about three revolutions it slows down and the overload trips. I
checked the unloader and it is working. I tried starting the compressor
motor without the belt on and it snaps right on, but apparently with the
inertia of the flywheel and pump, the motor draws too much current. It
pulls down V1-3 and V2-3 to about 140 volts before it kicks out the
Do I need a different or bigger idler motor or is my compressor motor just
that hard to start?
I thought that a 25HP idler would be big enough but evidently not. The specs
on the idler are: Linclon A.C. Motor, Lincguard Dripproof, Frame 284T, 1765
RPM, 25 HP, Insulation B, 230/460V, 62/31A, 40º Max Amb, 60 Hz, S.F. 1.15,
Time Rating Cont., NEMA Code G, NEMA Design B. I have heard that some motors
do not make good phase converters, maybe this is one of them.
Any insight on this will be greatly appreciated. I bought the large
compressor specifically to do sandblasting and now I can't get it to work.
On a positive note, the phase converter seems to power up my 1HP Bridgeport
If I ever decide to do sandblasting, a gas-powered compressor on a trailer
is what I'm going to get. I had one once and even *it* could barely keep up.
I think you need to check current and voltages on your 3 legs. And your
10 hp motor *is* wired for 220, right? You may need to borrow a peak-detecting
Jeffrey M. Born> I built a rotary phase converter using a 25HP motor. I am using 400uf Cp and
Assuming adequate wire sizes for the starting load, you might try three
capacitors directly across the compressor leads. This could raise the power
factor and reduce current enough to permit starting.
Agreed! What is the function of the 400 ufd Cp? Is that the start
capacitor for the 25 HP (idler) motor? What about the "20 ufd Cs"? Are
they run capacitors? If so, they aren't nearly large enough for a 25 HP
idler and 10 HP load motor. Theory has it that run caps should be at least
25 X 30 ufd or 750 ufd. Insufficient amount of run cap will give the
symptoms you mentioned, esp. when trying to run a loaded air compressor.
IMO you should:
Install aprox. 750 ufd of run cap between 1 side of the input 240 V line and
the 3rd leg of the idler motor. Make sure the run cap is connected across
the same leg (phase) as the temporary start capacitor. Check voltage across
the 750 ufd cap with the idler running unloaded. If that voltage is more
than 15 % higher than the input line voltage, reduce the cap amount until it
comes down to aprox. line + 10 to 15%. Check the voltage with the load
motor running. It should then be aprox. the same as line in V or slightly
lower. This arrangement gives an "unbalanced" rotary phase converter.
Get it up an running satisfactorily in the unbalanced mode. Then, later you
can manipulate the amounts of capacitance in both legs to achieve the Fitch
style balanced rotary phase converter. This will probably require more
capacitance in a 60/40 arrangement between the two phases - that is from
both sides of the input line to the 3rd leg. As before, check voltges and
strive to achieve aprox. line voltage across all 3 "phases"; making sure
neither phase goes higher than 10 - 15% greater than line in with the load
I have THEORIES on why this isn't working. They are UNTESTED and
should be used as food for thought.
I have a 10 horse motor phase converter. I end up with 220 volts
across two windings of my milling machine, and with 209 volts on the
third leg. When an amp-probe is set on each leg of power to the mill,
the two legs going to the 220 V single phase have large amounts of
current in them, and the generated 3rd leg has essentially none. I
think that the generated third leg is more startup than run. Any
amount of 3rd phase is better than none, and gets most motors running.
Motors running off these home made rigs run mostly on single phase, in
I suspect that putting a step up transformer on the home-made third
leg, to boost the voltage where it matches the other legs would make a
large difference in the amount of power delivered, and hence increase
the starting torque. Never tried it, since my mill starts, I am happy.
I also suspect that the only hope you have on your present setup is a
pressure relief valve which unloads your compressor until it gets up
to speed. I think these devices are available commercially.
Woops, I made a typo on previous post. I should have said a 200uF Cs
(Cp is between L1 & L3, Cs is between L2 & L3, Cpf is between L1 & L2.)
I have 1800uF worth of start caps between L1 & L3 which I momentarily switch
on to start.
The voltages I measured with different caps are as follows:
Cp Cs V1-2 V1-3 V2-3
0 0 238 211 204
100 0 238 222 210
200 0 239 237 217
300 0 239 246 220
400 0 239 256 230
400 100 240 256 244
400 200 240 257 255
400 300 240 258 265
400 400 241 258 276
500 200 239 267 261
500 300 240 268 271
600 100 241 276 259
600 200 240 276 268
700 100 241 287 267
I didn't go with all 800uF on Cp as the V1-3 was getting pretty high.
With Cp of 700 and Cs of 100, the compressor currents at 150psi are I1=21A,
I2=29A, I3=30A. This is with the compressor starting with no pressure in the
tank. It still won't start at 135psi.
I can email anyone an excel file of this if the columns don't come out
Are these (I haven't seen or read much but this ) the correct values ?
The caps are based on the Inductance of the winding, not the HP rating...
Naturally they are related in some ways.
THe concept is to have XC subtract from XL of the inductor so that the Copper
resistance Rs or Rcopper + |Xl-XC| is large enough not to blow fuses / breakers,
but allow enough in-rush current to turn the shaft. The shading winding gives
the first kick with several hundred amp-turns so the direction is set if shaded.
Once the motor starts, and is up to speed without twisting one end of the shaft
the switch is thrown and the capacitor is cut out and the run winding takes over.
So the point is this : capacitors are defined by the windings. For a given
the voltage determines the inductance - (considering things like number of
Bob Sw> Grant sez:
I have the same situation, 25 HP home made rotary converter, 15 HP IR
compressor. Here is what I do to get things rolling.
First, the converter. No caps. I use a 1 phase starter motor
directly coupled to the 25 hp motor shaft. Starting uses a time delay
relay and a contactor. When power to the converter is energized, the
starter motor comes on getting the main motor up to speed. Then the
starter times out with a time delay relay, and the main contactor
Second, the compressor. Make it easy to start by installing a time
delay relay to energize 2 solenoid valves. One relieves all
intermediate pressure in the interstage pipe. The other applies air
pressure to the unloaders. When the compressor starts the TDR goes
on, opening the valves, then times out in about 30 seconds, after
things are up to speed.
There may be 10 or 20 psi left in the tank. Not enough to open the
unloaders and enough to add a load. If that is the case, dump all air
Also get some heat tape and wrap it around the cylinder bases to warm
the pistons and cylinders. That helped mine to roll easier when cold.
After my rotary converter I tapped off to a step up transformer that
feeds another panel. That gives me "home made" 480 3 phase. The
compressor starts much better on 480 that 240. I switches all of my
big motors over to the 480 panel.
Finally, start another motor. I fire up my shaper, (15 hp 480 3 ph)
before starting the compressor. The added induction helps the
compressor to roll over.
This last step is only necessary on the first start of the day when
things are cold. From then on it is fine.
Good luck. This is real experience from someone that IS there and
DOES that. Contact me if you need more info.
I'm sorry to say I am pretty much drawing a blank, here. From your
description it appears you are working a balanced (Fitch type) rotary phase
converter with Cp between L1 and L3 and Cs between L2 and L3. L3 is the
3rd leg or "manufactured" leg. The 240-V "line in" is on L1 and L2. L1, L2
and L3 are parallel connected to the air compressor's motor.
Are you certain the wiring connections are all tight? Loose fitting
temporary clip leads could be a problem. Are you using at least #14 wire
between the rotary and the compressor motor? Use #10 here if running more
than a few feet. Are you really sure the compressor is properly "unloading"
for starting. 135 psi sounds like a high pressure to start against.
The next to last line in your chart shows Cp of 600 ufd and Cs of 200 ufd
with V1-3 at 276-V and V2-3 at 268 -V. The 276-V is about 15% higher than
line in V which is near the practical limit with V2-3 (268-V) a little
lower. You might try increasing Cs to 300 ufd to see if it raises V2-3 up a
little. Do not do this if V1-3 or V2-3 goes over 276-V (the upper 15%
limit). The obvious reason for this is to keep the unloaded idler motor
voltages from going high enough to damage insulation. Stay tuned - there
may be an alternate solution.
What follows is opinion and is unsupported by any practical experience on my
Loaded air compressors draw extremely high starting currents and your idler
motor to load motor ratio is about 2-1/2 : 1, which may not be enough for a
loaded compressor. Consider that more run capacitance may be necessary but
that you are already at the practical limit with the amounts of capacitance
you have tried. Consider addding permanent capacitance to the air
compressor motor, connected across L 1-3 and L 2-3. Make sure this
capacitance is connected directly to the compressor's motor, downstream from
the compressor's on/off switch. Begin with, say, 200 ufd in each leg.
Observe the compressor's starting "labor" and increase capacitance as
indicated. If enough capacitance is added to facilitate proper starting,
make sure that running voltages are no more than 15% higher than line in.
I removed all the wire nuts from the compressor motor, straightened the
strands and redid them. Didn't make a bit of difference that I could tell.
Are you using at least #14 wire
I have the compressor connected to the phase converter with 10 ft. of 10-4
AWG SO cord. I wanted to use 8 AWG, but I had the 10AWG on hand and I was
anxious to try out my new toy. Now I think I will buy some 6 AWG. The
converter is wired with left over 8AWG THHN except the caps, which are
paralleled with 10AWG to terminal strips. I am thinking of going to copper
bus bars instead of terminal strips.
Are you really sure the compressor is properly "unloading"
The compressor is unloading. At first I thought it wasn't, but the unloader
exhaust is plumbed into the air intake and is silenced by the air filter. I
removed the line to the intake and I could hear the "pssshh" when the motor
shut down at 175 psi. But...,this is a mechanical unloader which uses
weights that are activated by centrifugal force to close the unloader valve
(which is like a schraeder valve). After about 3 or 4 revolutions, the
weights are advanced enough to close the unloader valve. I removed one of
the two adjusting shims and it didn't seem to help. I will try removing the
other. I doubt if it will help, though, as the motor needs to overcome the
pump inertia and come up to full speed before pumping any pressure. A friend
at work has a big 3 phase compressor running from a home made phase
converter and he says his unloader takes a lot longer to close. I suspect he
has a oil pressure activated unloader. I was thinking (and another poster
has just suggested it), that I should use a time delay relay and a solenoid
valve to delay the unloader's closing until the motor is up to speed.
The other thought would be to add another phase converter or use a larger
idler motor. I read on the GWM phase converter site that ePact motors on
compressors may not work very well with rotary phase converters and may
require a 4X idler size. I even thought of using 3 separate identical phase
converters with the generated leg going to a different terminal on the
compressor. But this sound like a lot of complicated contraptions.
The last solution that I toyed with is to reduce the motor pulley diameter.
The pump on the compressor is used with 5, 7.5 and 10HP motors with
different sized pulleys. By dropping down to the 7.5 HP size pulley, my CFM
would go drop from 35 cfm to 27 cfm at 175 psi. Hopefully the motor would
have enough torque to spin against the load. This kind of negates getting
such a big compressor, but at a distress sale, I got a new deluxe 10HP with
intercooler and after cooler for the same price that I would have of had to
pay for a 7.5HP single phase compressor with smaller tank.
Consider addding permanent capacitance to the air
Adding more capacitance accross the compressor motor would be no different
than having them in the phase converter as far as the compressor is
concerned, right? But other connected loads on the converter wouldn't see
the high voltage, so I guess that may be an good option. My only other 3
phase load at this time is the 1HP Bridgeport mill, which spins up nicely -
even in high gear!
Do you have an oil pressue activated unloader valve? Mine is a mechanical
valve activated by centrfugal weights. I had the same thought as you suggest
after talking to a fiend that has a big 3 phase compressor running from a
home mase phase converter. He say his compressor gets up to full speed
before the unloader closes. I suspect he has a oil pressure activated
unloader which is taking longer to close for some reason. What kind of
solenoid valves and TDRs did you use? I kind of hate to hack up a new
compressor, but if it doesn't work, it doesn't do me any good. Adding the
TDR and solenoid valve may the best solution.
What kind of leg to leg voltages are you getting out of your converter?
Indeed. And if it doesn't work, you've wasted your time and money on
Can you hook a manual unloader, just to try it? Disconnect the unloader
line from it's valve and attach a hand/manual valve to the line. Open
the valve until the compressor is up to full speed then close it. Be
sure to open it again when the compressor stops. Bob
Jeffrey sez:> "Adding more capacitance accross the compressor motor would be
Adding capacitance directly across the compressor motor would be anagalous
to adding enough capacitance to the idler motor - except - that amount of
capacitance would not contribute to excessive voltage on the idler motor as
it would if it was placed there. Your symptoms, I believe, indicate there
is insufficient total capacitance to facilitate starting a heavily loaded
compressor. However, we cannot add enough capacitance to the idler motor to
handle the heavy load without excessive (over 15%) voltage across the idler
motor during standby. Placing the caps directly on the compressor motor
would (should) be a way of overcoming the excessive starting load, while not
causing excessive strain on the idler's insulation during standby.
My only other 3
Addendum to previous post:
It might be necessary to connect the capacitance to the compressor motor
with a separate switch or set of contactors, connecting extra capacitance
only during the difficult starting interval of the loaded compressor. This
in order to prevent possible excessive voltages after the compressor comes
up to speed. Those voltages will be felt by the idler as well as the
compressor. IMO, it would probably be well to limit this "running" voltage
to 15 % over line in.
You can try a larger idler very easily. Instead of using a larger one,
just add a second one in parallel. If you have another three phase
machine, all you have to do is turn it on (idling...) and see if this
helps your compressor start.
For wires short (less than 1/20th wavelength for practical purposes) with
respect to a wavelength (5,000 km for 60 Hz), the voltage will be the same
all along the wire (ignoring for the moment IR drop). So it doesn't matter
which end of the wires has the capacitor.
Were they thinking about having the C(extra) at the compressor motor,
simply because of the compressor contactor? That way when the
compressor wasn't running converter voltage would not tend up.
please reply to:
JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com
Were they thinking about having the C(extra) at the compressor motor, simply
because of the compressor contactor? That way when the compressor wasn't
running converter voltage would not tend up.
Adding a second potential relay, at the compressor, with the hysteresis changed
such that an additional run-type capacitor (A-B), or capacitors (A-B and C-B)
were placed on-line whenever the B voltage dropped on account of the compressor
starting would be a way to go.
Another way to go would be to add a state machine at the converter, one which
has a starting state for the converter alone, a running state where normal
loads would be supplied, and an overload sate which would handle the special
case where the compressor was placed on-line.
This overload state would place extra caps, hence extra VARS, on the converter,
thereby temporarily boosting the B-phase voltage.
(If the A- and C-phase voltages are sagging while the compressor is starting,
then the motor feeder is too small).
Another possibility is adding an autotransformer for this case.
(In fact, one of the first published rotary phase converters did in fact use an
autotransformer on the B phase).