Solid state Variable Frequency Drives to convert [genny 230 vac] to 3 phase 208 to run three phase well pumps, or other SHTF motors that may be hard to start with smaller generators.


The same tech that allows sine wave inverters to become low in cost has affected the price of VFD's -downward.

The latest VFD's are available on eBay for not much more than $100 for a 3 HP motor capacity unit. The units are small and weigh little.

The reason the surge load on a given 3 phase motor can be reduced is because the VFD's are programmable with regard to almost every parameter concerning the motor.


The frequency output [driven motor RPM] of the VFD can be varied and a true 'soft start' can be implemented, and the motor speed ramped up, avoiding the sudden high current load as the motor tries to instantaneously ramp up to the rated RPM.

A lot of appliances, A/C's, etc. use electronic motor controls to vary the speed of fans, blowers, compressor, etc.


If folks are interested, we can discuss this further.
 

I have an idea how an inexpensive 3 phase drive might be adapted to run most any capacitor start/run single phase motor, and if it works opens a whole new world of  low current motor starting.

I'll disclose it, it's pretty cool.

I think a 3 phase VFD can be connected to a motor with just two of its phases, without triggering a fault code.

Soooo, if you drive a single phase motor with the appropriate [reactor (inductor) instead of a capacitor?], to shift the VFD's phase channel to 90 degrees, vs 120 degrees, ordinary motors might be nicely controlled over a useful range of torque and speed.

Tho not as wide a range as a 3 phase motor, that operate almost perfectly over a wide speed and torque range when driven with a VFD.

That sounds very interesting, particularly for RV & window unit A/C.

Careful with trying to start AC units with a VFD as the sizing requirement are different to overcome startup loads.

As with any electronics, don't place your eggs all in one basket, they do fail. At least consider a spare.

I don't know that I would trust VFDs from Ebay. They maybe cheap for the simple reason they used cheap components and may even be of inferior design.

Even the name brand VFDs are made in China, but with substantial engineering design behind them with strict component specs.

I do have my personal brand choices, but they may not suit everyone's budget.

Becareful trying to run single phase device from a VFD. Most do some type of load imbalance sensing and will fault.

ETA:
VFDs are perfect for running pumps and fans. I don't want anyone to think they are going to being able to use one of these to run their HVAC or even an RV AC unit. Any electronics between the VFD and the motor you are trying to start will be toast!

Also not all motors will be happy running off a VFD. Problems arise in insulation overheating, acoustic noise ( as in a high frequency scream coming from the motor) which might be able to be tuned out by checking switching frequency..

Do not run a motor in continuous duty below 40Hz without some additional cooling. You can push to 90HZ but understand you will shorten mechanical life of the motor.

Here is a link to some that I would consider over anything on Ebay.
http://www.automationdirect.com/adc/Overview/Catalog/Motor_Controls/Drives

No they are not top of the line like Siemens or Allen Bradley but are of general use type.

I have a pile of 2hp-3ph motors, so yeah, tag.

Automation Direct VFD's on ebay...

Low as $39 start bid...


A-D VFD EBAY

So..... - could this be wired in between the fancy electronics and the compressor motor in a window unit?  

Sure.

Assuming the chosen VFD worked with the compressor motor.

I'm in.

Who runs residential 3 phase anything?

Very common in residential pool pumps now.

many (perhaps most) apartment complexes have 3 phase 208/120 service. 2 phases and a neutral are run to each unit.

not many single family residences have 3 phase service though.

I am not real convinced that a capacitor start single phase motor can be reworked to be used with a VFD. the centrifugal switch that takes the starting capacitor out of the circuit would need to be dealt with in some way. Most of these motors are not suitable for use on VFDs anyway as the insulation of the motor is not designed for it.

Most 3 phase motors these days have insulation designed to handle the high frequency and relatively high voltage  DC pulses that come out of a VFD.

I would not be trying to put any kind of a compressor that is not unloaded on a VFD though.

I do!

I have a 4-axis CNC milling machine in my dining room workshop.  

Each axis is driven by stepper motors, there are driver boards to take care of that,

But the spindle motor is 3-phase 240VAC,
An Emerson VFD. drives it.

It is very reliable and pretty efficent.

Because it is a variable frequency I can run the spindle motor waaaaay faster than I otherwise could.
A 60HZ motor at 160Hz sounds like a dentists drill.

Additionally you can slow them to a crawl too.

You do NEED a VFD rated motor.

A VFD rated motor has an extra slip ring on the shaft which goes to ground

When a motor is run on a VFD there are stray magnetic  "eddy currents" generated in the armature and ground to the frame of the motor.

Those stray electrons will wipe out the bearings in the motor if not run to ground!
(Without the slip ring it grounds THROUGH THE BALL BEARINGS! -not good for them)

In a bind I would consider having a solid #8 copper wire rubbing on the output shaft and ground that

Very good explanation of what is going on inside the motor that wipes out the bearings.

Roughly 30 years ago when I first started working with VFDs ( back then we called them inverter drives) these drives were as big as one of those small dorm refrigerators for a 20HP motor. I have used them on about everything imaginable. Large inertia loads and compressors are the worse applications. I am not saying they can't be done, but they can be headaches get a good balance between startup and run protection. And yes you could probably insert one between any electronics and a compressor load. You would need back feed protection on the input side of the drive. Not something the average person could do without some electrical or electronics background.

Worse application I ever did was on a centerless grinder. The grinding wheel was roughly 4 foot wide and about 3 feet in diameter. So much inertia it was a bugger to get started without tripping the drive and if you compensated for the high inertia start you left yourself unprotected in a run condition. Stopping and slowing down was just as bad.

Last major application I did was two CNC machining centers. One had 15 stations the other had 13. Total drive count was over 35 drives!

Converting a capacitor start motor to run from a VFD would mean disabling the start capacitor switch. Then I don't know that you would have enough torque to start. I don't know I have never tried it. It would be like trying to start as if the capacitor was bad. You could spin the shaft everytime you wanted to run.

Would the drive provide enough torque boost to start the motor? If so then my question would be why? Would it be so that you could start a large load single phase motor from a generator that was too small to provide starting current? Just remember watts in = watts out plus loses.

I can't comment on the 208/120 thing as that is mostly commercial buildings and I always dealt with industrial apps.

Expy sorry but I don't understand exactly what you are wanting to achieve. Are you wanting to start 208 3~ loads with a 240/120 generator or is it high start current 240/120 1~  loads? The only reason I say shy away from Ebay is because of the proliferation of China clones appearing on Ebay. It may say GS1 or GS2 or Automation Direct type VFD. I would recommend that if you want to be assured it was genuine to buy direct from AD. By the way Yaskawa is one of the larger VFD manufacturers and might be the one that makes the drives for Automation Direct. Automation Direct absorbed Vishay a few years back so I can't say for sure, but if you compare the operator interface you will see they are similar. Also the programming is similar.
https://www.yaskawa.com/pycprd/products/product-directory

Why not try thinking outside the box?

Insulation is probably well over rated for a VFD.

Centrifugal switches can be easily bypassed.


Re the compressor, why not?

"I do!

"I have a 4-axis CNC milling machine in my dining room workshop.  



You -character!

Nah, you DON'T NECESSARILY need a VFD rated motor.




For example, last night I hooked a HY brand VFD to a Bridgeport motor and it works BEAUTIFULLY!

And prolly thousands of other folks have too in the past 10 years with nothing but gushing LOVE.

lots of things that are sub-optimal will work. For a while.

You do need it if you expect a reasonable service life out of the motor.

Will the motor spin? Of course,
But what you don't see is the moment you start it THE BEARINGS ARE ARCING INTERNALLY.

Bearing material transfers to the race, race material transfers to the bearing.

If you do not have a VFD rated motor you NEED to ground the shaft.  Somehow.

I could say putting vegetable oil in the crank case of my car works, because I have tried it before
(And I have)
But armed with the knowledge I am I can tell you you will not have an acceptable service life.

Likewise, my experience working with VFD's and motors in an industrial environment tells me when you lose that ground
You have a few hundred hours before the bearings are shot.

In a true emergency, yes, it will function.

OK, everybody using a VFD, ground the motor shaft.

Most any distrib of VFD's will have them for little cost.



My buddy who is going to put one on a lathe for home use probably won't bother.

Besides bearings like the 6204 and 6206 on the motor I hooked up last night cost a few $ and take a few minutes to change and I prolly won't live long enough to have to do it.



Oh- wait...


Those aforementioned bearings are going to be replaced today because one is squeaking --machine never had a VDF on it...

Until last night...

Ya know, as soon as I typed it, I was sure someone would tell me about machinery. lol

Ok, I have never seen a residential 3 phase panel here. Plenty of industrial and military, but not home stuff.

So how much current is generated by these 'eddy currents' ? At what frequency?



Are you saying a #8 wire for current carrying capacity, or for physical stability of a 'slip ring' substitute?





 

I like #8 solid for its physical characteristics, stranded would work, so would a spare motor brush.

The frequency and current is dependant on motor design, load, and the VFD.

You need an oscilloscope to measure it because a meter you own measures in RMS, and that won't work accurately at anything other than 60hz.

Put a 1 ohm resistor inline with the ground AND MEASURE THE DEVELOPED VOLTAGE ACROSS THAT.
whatever the o-scope reads in volts is what the milliamps/amps are going to ground.

 As I'm a bit stubborn, sometimes too smart for my own good, and a creative problem solver
I appreciate your candor sir.  

Pot -meet kettle!

small/cheap VFD's don't have the capacitors needed to provide surge current for cycling motors. They are designed to run a motor continuously.





just keep that in mind for your application

 

Sorry, don't follow, Gas...

Usually, VFD's are programmable with regard to the motor acceleration via ramp up time, that pretty much eliminates the instantaneous motor starting surge.

You can let the motor start, run up to programmed speed, dwell if desired, slow down and stop, reverse, rinse and repeat ad infinitum, without ever having any surges to be concerned with.

well if the VFD is only controlling a single motor load its fine. But if you have other items in your system like solenoids and switches that are cycling loads, it can really mess up a VFD.



example: I had to use a motor 2 to 3 phase motor converter instead of a VFD of a bottling line, due to the extra loads it would spike current which the VFD could not handle.





 

Yes, since devices like solenoids are inductive reactive, them switching on and off is "seen" by the VFD as a changing speed or load on the motor
and so it tries to (erroneously) correct the variable it mis-sees.  

Big VFD capacitors are needed for discharging into slow turning motors under a load.
If you use a VFD rated for the size motor you've got there shouldn't be a problem.
(I say that based on working with "real" industrial units like Mitsubishis.  The Ebay ones I don't know about.)

And I've spent waaaay too much time working on bottling lines myself!  

You guys have my interest. I've got a couple of low-pressure wobble-head compressors running air assist on CNC laser systems. I would love to be able to control the speed of these compressors. The motors are rated 120VAC 600W 1675RPM and use an externally mounted 25uF capacitor. The datasheet says the normal inrush current is 18A. We run these motors continuously 10-15 hrs/day 7 days/week. They never cycle on and off once they're started because they feed compressed air directly into the system with no pressure switch or tank involved.

Questions:
1. Can you guys recommend some light reading that will help me better understand how to make a motor like this play nicely with a VFD?
2. Is there a rule of thumb for the range of speeds possible with this sort of motor?
3. Is there any advantage or disadvantage to over-rating your VFD voltage or power ratings?

Thanks!

Capacitor-start motors are a problemwith VFD'S.

Without getting too technical, the frequency of the incoming power is matched
to the inductive reactance (impedance) of the start winding in the motor which is canceled by the capacative reactance
(Think of that as "negative impedance") once it is up to speed,

Esentially turning the start capacitor off once up to speed.

You would need to start the motor normally, then open the start capacitor circuit with a switch or relay.
(Essentially "manually ensuring it stays off")
THEN you can vary the frequency to control speed.

If you change the frequency with the start system still hooked up the start system partially (or maybe fully depending on the frequency)
Will come back on overheating the motor.



There is no big need to over rate a VFD for higher voltage/power,
All the "real ones" I've worked with have that already designed in.

If you've got a 3hp motor, just get a 3hp VFD.

The possible speeds are limited by how much current the motor can take before popping like a fuse or overheating.
Set the current limit parametrs in the drive to the motors lable plate limits to start.
You can go higher if it is a good motor, a big motor, or a light load.

I won't say 150% current rating is usually a good max safe number for obvious liability reasons.  

Thanks for the detailed response-- that was very helpful.

I believe these motors are actually PSCs. When I put my multimeter on the capacitor and start the motor I get ~240VAC running at rated speed with minimal load, so I don't think there's a centrifugal switch disconnecting the cap from a start winding. I could be way off base here.  

My hope would be that I could run these motors at a significantly lower speed. We have to adjust air flow at the cutting nozzle to get different materials to cut and engrave cleanly. Right now we're running these pumps at full rated output and bypassing a lot of the output to reduce nozzle pressure for some materials. In the end, we may just have to use a different type of pump, but these have proven to be bombproof so far.

There will ALWAYS be voltage at the starter cap.
Your volt meter won't help you there.

There is a trick used (matched inductive and capacitive reactance) to STOP CURRENT FLOW
through the start circuit once it is up to speed.
It is not intuitive unless you've studied it, and the math that maked it work.

By opening the capacitor circuit once started you ensure CURRENT doesn't flow when the frequency changes.

Put an amp meter in-line with the capacitor.
THEN you will see it work!  

Thanks again! I measure current through the cap as about 3.1A with the compressor running freely with no real load. So, am I correct that this tells me that the capacitor is still connected to the run winding but that I cannot necessarily rule-out it also being disconnected from a start winding (i.e., a single value cap being used as a start-run cap)? In that case, I could have a conventional split-phase motor or a PSC. All of the diagrams I'm finding online show one of two conditions: 1) dual start-run capacitors with different ratings for the start and run sides, or 2) start coil with a start cap and run coil with no cap. Are there cases in which a motor will use the same single-rating cap for both start and run conditions? I'm thinking no, but I'm obviously not a motor guy.  

Yep, hang a clip on ammeter on one of the cap leads and you can see how much current is being drawn during operation.


The easiest way to solve the compressor issue is possibly to replace the single phase motors with 3 phase.

Plenty of nice motors on ebay and even new aren't too bad and might be worth doing to save wear and tear and electricity costs.

Your application is where VFD's and 3 phase motors shine.  


Once you hook up a VFD to a 3 phase motor and play around w/ it, you'll get a BIG smile on your face.

I guarantee it.  


Don't let folks scare you regarding mismatched parts, inferior construction, and that low cost VFD's are particularly inferior.

That's many times similar to the marketing BS ---like you hear abt India pharmaceuticals...   That's all you get and don't know it.  



For most aps, low cost VFD's may work just fine and for a long time.

The components used in the inexpensive ones are likely the same as used in the high dollar ones with fancy brand names.

In any case, these components are made in the zillions and have become very low in cost.

If the run winding has a cap this will never work for you.
(Because as frequency changes, what the cap does changes too)

The start cap should have very little if any current flowing during run.
3A is more than I'd expect (not impossible... but).

Start it and open that cap circuit.
If it keeps running that can't be the run side of things.

It is possible there is a run cap -but unlikely.
Testing will let you know.  

If there are two caps on your motor then you'd need a different motor
(Unless I can think of something nifty)
But I'd expect them to be different values if there are.

P-P, I may run some of these experiments this afternoon with a 208/240 vac single phase cap start/run blower motor.

If these wobble head compressors have an unloader on startup then sartup load is going to be typical startup currents. Yes the magical 150% overload limit comes into play here. Disclaimer: Some drives allow setting overload % higher and lower than 150%. Use at your OWN RISK. Sounds like since he is not running any receiver tank he has to  adjust the air bypass to control nozzle pressure. Where as a normal air system would be using a receiver tank and a typical air pressure regulator.

My suggestion is to contact a local motor shop or rebuilder and see if they can match a 3~ motor to replace the motor you have now. Then you could use a VFD. I do see a problem with slowing compressor as it will also reduce your available CFM.

Motor speed limit is constrained more by the cooling requirement and mechanical reliability more so then current limits. Current limits come into play more so on startup, acceleration, and deceleration. Too slow a motor speed is going to reduce cooling airflow around motor windings. Typical limits are 30 to 90 HZ on off the shelf motors. VFD rated motors can go higher. Blower cooled motors go slower.


I too cut my teeth in bottling plants and breweries. Project design and management, everything from inside the bottling room to packaging equipment.
Last 20 years with an OEM automotive drivetrain manufacturer. CNC machining, automation, and robotics. So glad I left the beverage industry,

Eliminate the caps completely. Let the drive provide the start currents. You already know the motor can handle the start current. Motor is not going to draw anymore than it needs. Drive just provides the available current boost. I would say an acceleration curve of 1 to 2 secs. Expy run a test before experimenting. Time the startup switching. You should be able to hear the centrifugal when it switches from start to run caps. Use that time as an accel time starting point. Set the motor current parameter to match the motor tag. Overload current to 150%. Motor speed to 60HZ.  Depending on the load you may have to extend the startup accel time. Drive will keep voltage constant and boost current.
Use a clamp on amp meter and monitor run amps. Unless your meter stores peak value you will not be able to see actual startup amps as they will spike very high at initial turn on. You don't need to be really concerned with that value unless it stays high for longer than accel time. Am talking milliseconds normally.

Thanks for all the input, guys. I like the suggestion to swap the motor for 3ph, but in this case it would be non-trivial to do that. The compressor has two heads, one on each end of the motor, coupled together in the middle, and crankshafts on each end, too. So, motor length, mounting pattern, shaft length, and shaft diameter and keying are all critical. If it was something easy like a belt-drive I'd swap it in a heartbeat.

There's no unloader on these compressors as they are not designed to restart under any significant pressure. I'm not worried about losing flow and/or pressure because that's really what I would hope to achieve by lowering motor speed. The point about motor cooling is well taken, though.  

Here's a link to the compressors we use: http://www.gd-thomas.com/workareashowcontent.aspx?id=24368

Expy, I look forward to your results. What you're proposing to test sounds similar to what I'd like to do.  

Good input on the "using the drives instead of the caps".
I wasn't 100% that would work ...because I'm NOT a motor expert
(But what I do know I try to share)

I'm an unemployed electronics/instrumentation/PLC/CAD/CAM/CNC guy.

Beer sales are down DRASTICALLY across the nation,
After years of dodging layoffs they finally picked my number,
MillerCoors let me go in Oct 13.

Trying to get the CNC/manufacturing gig going,
Its taken me a whole year to learn and get fluent with everything from Solidworks to feed and speed selection.

I think I've got it now!

I hope.  

This has been a fun thread!
Looking forward to hearing how it all works out!!!

It appears that they use what is called a dual shaft squirrel cage blower motor. You may actually be lucky. Inspect the motor name plate and see if it has a NEMA frame size listed. If it does then it should be a standard motor not some proprietary motor.

If you are serious about a vfd powered well pump, get a 3ph pump and vfd now and get it running right before you introduce the genset.
If you are feeding a vfd with single phase you need to have a drive that is rated at double the motor size. This is necessary to ensure the DC bus stays high enough to run the pump.
If it is a in well pump you will also need a load reactor for the drive to work properly with the long conductors.

Interesting abt the load reactor and the long conductors...

Would you go into more detail?

Think of the input or the load reactors in ham radio terms as a choke or filter.
Copy and pasted from here:
http://www.automationdirect.com/static/specs/aclr.pdf

Input line reactors protect the AC drive
from transient overvoltage conditions typically
caused by utility capacitor switching.
Input line reactors also reduce the
harmonics associated with AC drives, and
are recommended for all installations.

Output line (load) reactors protect the
motor insulation against AC drive short
circuits and IGBT reflective wave damage,
and also allow the motor to run cooler by
“smoothing” the motor current waveform.
They are recommended for operating
“non-inverter-duty” motors, and for any
motors where the length of wiring between
the AC drive and motor exceeds 75 feet.

Motors run on VFDs do not always have the problem of induced current flow through the bearings and the shunt method of trying to solve it is not as successful as one would hope. It is a complicated issue that even people with a lot of experience in the field end up shrugging their shoulders over sometimes. The only known solution for this problem that actually works reliably is isolating the bearings so no current flow can occur. Fortunately, this problem is fairly uncommon.

A more common problem on low cost and especially older motors is that the insulation used in the motor windings is not really suitable for the voltage pulses being seen by the winding. VFDs don't produce a sine wave like normal AC current does. It produces a series of DC pulses that mimic a sine wave. The motor does not really care whether it gets a sine wave or a set of DC pulses timed correctly with the right phase shift. The pulses can have all kinds of weird effects on the waveform that actually gets to the motor. They can even bounce back and forth on the wires going to the motor and build up so they exceed the voltage rating of the motor winding insulation. A VFD rated motor will typically have much higher voltage rated insulation to accommodate this. Having said this, it takes a while for the motor to fail from this problem, and for motors that are not used much, it might not make a whole lot of difference. It is more common as the conductor length between the VFD and the motor increases, so it is a less than ideal decision to put a VFD on an older well pump motor that is a couple hundred feet down in the ground.

I don't understand your comment. VFDs, even cheap ones, work fine at cycling a motor as long as they are sized correctly.

it may have tripping out on high DC bus voltage. a braking resistor might have solved the problem.

High inertia loads will do that.



Trying to feed a pulsed DC voltage into the primary of a transformer will cause the same problem.

I decided to put the bottom line up front:
Yes, you can use VFD's as a soft starter, but I would advise against using them as a soft-starter to compensate for an undersized generator.

Detailed response:
Unless I'm mistaken, this isn't quite this simple (especially for well-pump size loads) when running off a genny. I currently have a Mitsubishi VFD running a 2 hp motor and it cautions that when running from a generator the recommended generator sizing is 3 times the FLA of the VFD so that the generator can absorb the harmonics fed back into the generator without harming the overload protection devices.

Sounds confusing but you have to consider that a VFD is rectifying AC power to DC, storing it in capacitors, then releasing it back as 3 phase AC. The way the inverter pulses the current out is in very high frequency steps. This can cause a high-frequency feedback in the lines (line filters are recommended). Also, because it is taking fixed frequency single-phase AC and turning it into variable frequency 3-phase it draws power from the lines very unevenly. On 1 cycle of the phase it may draw 10 amp, but the next cycle of the phase may draw 1 amp, and the next will draw 6 amps etc because the power demand of the 3 phase doesn't necessarily coincide with the peak power of the input phase. This is VERY hard on a generator, especially if it's sized somewhat close to the load.

its the other loads that are started downstream, solinoids



 

It might be hard on an INVERTER genny, however, on non-inverter genny's it likely makes little difference.

Please describe the failure modes some more, referencing high frequency pulses on ordinary gennys and why they would cause more significant damage than the high frequency pulses applied to ordinary motors.


"Also, because it is taking fixed frequency single-phase AC and turning it into variable frequency 3-phase it draws power from the lines very unevenly. On 1 cycle of the phase it may draw 10 amp, but the next cycle of the phase may draw 1 amp, and the next will draw 6 amps etc because the power demand of the 3 phase doesn't necessarily coincide with the peak power of the input phase. "


In practice, the large capacitors after the rectifying stage of a VFD, tend to smooth these high frequency pulses and prevent them from passing back through the rectifier bridge and entering the input power lines, altho, the capacitor filtering isn't close to perfect.

Personally, I wouldn't worry about it as far as small motor drives are concerned when used with common genny's.  A reactor might be installed on the VFD input to further mitigate any issues.


I probably wouldn't worry about it with inverter based gennies since the devices in them are most likely rated for far worse conditions and high frequency pulses, similar in specs to the switching devices in VFD's.


This said, I would like to see rational data supporting the contrary.