VFD's and expectations of them

[Keith Foor]

I see a lot of questions about VFD's and some statements of personal expectations of them.
I want to put this out there to set reasonable expectations for VFD's. What not to expect and frankly their shortcomings. Hope you find it informational.

VFD's are actually a new spin on an old technology. Servo drives and DC servo motors have been around for years. And variable speed 3 phase motor controls are nothing new either, but the older ones used contactors and resistors to limit the current and voltage to the motors to slow them down. The excess power was just burnt off as heat. Now with newer technology the VFD or Variable Frequency Drive will control a 3 phase motor in a much more efficient manner.

WHAT THEY ARE USED FOR IN INDUSTRY
While VFD's are in some folks eyes are a God sent for the Hobby Machinist. They were not designed for the purposes that they have adapted them for. The two main applications for industry are air handling fans and material handling, IE conveyor drives. They can be found elsewhere, but this is the mainstay of them.
They provide a 50% or so reduction in the operating speed of the controlled motor and this of course is significant with these applications.
Fans are fans. I will not go into the specifics of them. They run slow, fast, or in between.
Conveyors are what shows the biggest strength and weakness in machining for a VFD.
Remember that it's a 50% reduction. Some can get a bit more but that's a reasonable expectation.
Now, with a conveyor the motor is NEVER coupled directly to conveyor, there is always a gear reduction that with a decrease in RPM increases torque dramatically. But these conveyors are setup to run a specific speed or range of speeds or be stopped. If a conveyor is designed to run at 100 foot per minute, it will never be expected to run at 1 foot per minute. This is where the VFD on your mill will let you down as well. You can't expect to set your mechanical drive gearing of your mill to 2000 RPM and then use the VFD to slow it down to 200 RPM. And even if you could, you still loose the mechanical torque advantage of the mills gearing so torque suffers greatly. At 2000 RPM with standard SFM cutting speeds and a .5 end mill you are either hogging out aluminum or plastic. At 200 RPM, you are going to be cutting really hard stuff or using a really big end mill on mild steel and not a half inch one. That requires torque and lots of it.
So whats reasonable? As mentioned before 50% or so.

A great application for a VFD is a cone drive (wide belt) lathe. If you have 4 settings that are 1500 1000 500 and 250 and you need a RPM of 750 what do you do? At 500 you are getting poor machined surfaces and at 1000 you are getting excessive wear on your tool due to heat buildup from the increased speed. So you set the belt for 1000 RPM and slow the motor until 750 RPM is reached. You have the correct speed and are not suffering for as much lost torque as if you were to set the belt at 2000 RPM and then slow the motor to achieve the needed 750 RPM. This applies to a number of vertical mills I have seen as well that used a step pully drive system to get a sort of close spindle RPM for their machines. Of course these are some of the machines that we are using in our home shops and suffer from the close but not quite spindle RPM issues on a regular basis.

Point is to not have an expectation that you can simply lock out your mechanical speed control system on your machine because you have a VFD and expect it to work at all speeds. It's not going to happen.

Reasonable expectation is key here. A VFD in unison with the mechanical controls of your machine and a optical tachometer will allow you to set your cutting speed correct and spot on every time. But you have to play within the rules to have a good experience.

 

[strantor]

The "50% and above" rule of thumb may be generally applicable on this forum, where the bulk of VFD installs are cheap Teco drives installed on non-inverter duty motors in open loop V/Hz mode. But the norm for this forum is not the norm anywhere else. A decent VFD combined with an appropriately rated vector motor and encoder feedback will perform without flaw at speeds below 5Hz; some go all the way to 0Hz and then have DC injection for holding torque.

There is no reason (other than price) that any hobbyist here couldn't achieve those results if they needed to. Just not by slapping a Teco drive on a tool and calling it done.

 

[mksj]

I am a bit surprised on the limitations you put on VFDs, and also the narrow application you paint. Probably just about every CNC milling/lathe system uses some form of VFD or servo drive operating over a very wide speed range, and if you hold to the 50-100% operating range it is in stark contrast to every factory equipped VFD lathe and mill. The limitations you discuss may apply to a poor conversions using an under powered system, but it is not true of a properly designed system. So if you look at most smaller VFD mills and lathes (5Hp and under), they tend to have 2 mechanical speeds, and the motors cover a range of a 10 fold speed range. As an example my knee mill with a factory VFD has a frequency range of 20-200 Hz. It is also common to oversize VFD motors so that their Hp and Torque are somewhat compensate for frequencies below 60Hz.They also can provide an overload of up to 200% for up to 1 minute. Above the base frequency a inverter/vector motor Hp is constant, but a comparable gear/belt driven mill/lathe will deliver 1/2 the spindle Hp at 2X the base speed and torque will be similar when one factors in the drive ratios.

 

[strantor]

I am a bit surprised on the limitations you put on VFDs, and also the narrow application you paint.

I can only assume that his reasoning is rooted in motor cooling, specifically on non-inverter-duty rated motors. I can't cite a source, but I seem to remember reading in some VFD manufacturer's literature that VFDs should not be used to power standard motors, but if they are, the speed should be kept above 50% to ensure the motor has enough cooling, since the cooling fan is attached to the rotor.

Inverter duty motors typically have milled fins to dissipate more heat and/or have separately powered cooling fans to let the motor run at 3Hz all day without burning up.

 

[Keith Foor]

Yes, the 50% statement is based on having whatever old motor was on the machine to begin with. It also addresses the issues of trying to bring an OLD motor down to 10% of it's rated RPM and expect 20 times it's torque output which is a bigger issue.
Yes, I am aware that there are vector drives and matching motors that WILL do these things. There are motors that are designed for 200Hz input. But they are RATED for those RPM levels. I am pretty sure that if you ran 200 hz to the factory motor on a bridgeport mill it would come apart in short order.

And of course if you understand vector network drives and motors and have that type of gear on your machine then this isn't really directed at you either.
This is for a guy that drags a 1930's cone drive lathe or an old Bridgeport home to find it's 3 phase and he wants to run it on a cheap import VFD.

And I didn't get into it but over speed on a VFD is like nitrous oxide on a car engine. The bottom end af a car engine will take X amount of horsepower before it fails. Now you can make that horsepower with new heads and cams and parts, or just spray it. The issue with spray is the kit is a tenth of the cost of the parts to make the same power it's capable of creating. And you simply re-jet it to make more power which means you don't typically have the respect for it. You want more, you pop the hood and 15 minutes later you have re-jetted the motor to make 3 times what it can handle. An electric motor is no different. It's more speed that kills here and not overall power, but it still kills. You can pull your tired 1960's 3 phase motor off your machine and get a whiz bang brand new vector drive setup and go direct drive with a new 5000 RPM spindle on your machine or just run your Chinese VFD at 200 Hz and get the same RPM out of it, for as long as it takes to seize a bearing, have the winding explode in the motor or it otherwise eat it self. Or you can run the machine within it's design parameters and get a lot of enjoyment out of the hobby for a long time. So can you do these things, yes in some cases. You still are not going to take a 19560's 3 phase motor with a plate rated 3450 RPM. Run it at 345 RPM and get the same spindle torque as running it at 3450 with a mechanical speed reduction. That is 10% of the speed which would be 10 TIMES the rated torque give or take.

 

[strantor]

You still are not going to take a 19560's 3 phase motor with a plate rated 3450 RPM. Run it at 345 RPM and get the same spindle torque as running it at 3450 with a mechanical speed reduction. That is 10% of the speed which would be 10 TIMES the rated torque give or take.

I didn't realize people were making this assumption. Yeah I guess it's important to note that when using a VFD to reduce speed, you are reducing HP at the same time, since HP is a function of speed and torque. Torque stays the same (excepting some drive parameters that let you boost torque some percent for some finite amount of time) while speed goes down. The drive does not replace mechanical torque multiplication, if that is what someone really means when they say mechanical speed reduction.

 

[markba633csi]

It's good to point out that for a lot of conversions the main purpose of the VFD is to make 3 phase first, and give SOME speed reduction second.
Also the sharp waveforms can apparently be hard on older motor insulation?
Mark S.

 

[strantor]

It's good to point out that for a lot of conversions the main purpose of the VFD is to make 3 phase first, and give SOME speed reduction second.
Also the sharp waveforms can apparently be hard on older motor insulation?
Mark S.

Yes, and yes.

Inverter duty motors have upgraded electrical insulation to cope with the nasty voltage spikes caused by VFDs. Ordinary motors do not. That, coupled with the fact that on older machines the insulation is probably already degraded, and the old motor's life expectancy drops significantly when installing a VFD. But it's not the end of the world. When the old motor fails, you get to upgrade to an Inverter duty motor, which if you get one with an encoder, will allow you tight speed regulation and departure from the 50% thumb rule.

 

[mksj]

Most instances where I see a VFD used on older machines in our hobby is because one does not have 3 phase, and does not want to use an RPC. VFDs in many applications can be easily installed and operate a machine with minimal controls, while offering advantages of soft start, braking, overload protection, etc. Some speed adjustment may be used, but usually this is between 30-60Hz and individuals inquiring about this are told not to exceed these range parameters, and it is clear they are not going to run their motors at 200Hz and beyond.

Most 4P three phase motors running at their base speed can easily be run up to twice their base speed mechanically from an armature/ bearing stand point of view, there can be a cooling fan efficiency and the electrical design specifications. Mechanical motor cooling fans be an issue below about 30% of the motor base speed and usually not an issue up to 200% of their base speed.

Newer motors of say the last 20 years can easily be run up to 150% of their base speed and retain full Hp and Torque falls off in a somewhat linear fashion. Beyond this speed, they can perform poorly. Most 4P inverter/vector motors up to 5Hp can retain their rated Hp up to 3X their base speed and torque falls off in a somewhat linear fashion. Very old motors should not be run past their base speed and the VFD frequency should be kept as low as possible, preferably in the 2-6Khz range. Agree that the insulation integrity becomes an issue, and as mentioned they may have a shorter life span. But in reading reference material provided by the VFD manufactures, the issue of insulation failure is rarely an issue on 220-240VAC motors because the insulation most of these motors was also specified be operated at 440-460VAC (see attached document from ABB). There are other issues such as the VFD induced mechanical vibration on the motor windings and bearing eddy currents that can also damage motors, but this would assume you are putting a lot of hours on the motor.

On speed regulation, I have looked into this and discussed it with manufactures, and standard V/F control provides a 2-3% speed regulation, sensorless vector speed regulation is approximately 0.2% and 200% rated torque at close to 0 Hz, encoder feedback is ~0.03% speed regulation and is usually used where precise speed regulation is required and/or positional control of the shaft is required. http://machinedesign.com/motorsdrives/how-choose-right-control-method-vfds

As the application in the hobbyist environment specific to the machines we normally encounter, I think all the points are well taken. An older motor, the primary use of a VFD is 3 phase conversion and understanding the limitations. If you decide to upgrade the motor (and the mechanical aspects of your machine), then a VFD can be used to a much wider full potential.

 

[Keith Foor]

I am not trying to down VFD's here. They are great at what they do and if run within the constraints of an old machine, they will be fine and provide a user with a reasonably cost method of creating 3 phase power, giving the guy with the cone drive or step pulley machine the ability to set his RPM to exactly what he needs for proper machining of his material.

I am just trying to convey that they are not sliced bread with fresh peanut butter and jelly. They have some limits that while can be over some are still limits. And the issue is like I pointed out about nitrous oxide in a car motor, they have the ability right off the keypad to destroy an old motor if attempts are made to run them outside the design of the motor connected to not to mention the equipment it drives.

I personally took a 13 inch South Bend cone drive lathe to a spindle speed of 1500 RPM without knowing better. The oil cup on the top was empty and the bearing (saddle bearing) next to the spindle expanded from the heat and seized. I personally got lucky that once it cooled off the bearing was not hurt. But same thing applied here. That machine was designed for a max speed of about 700 RPM maybe 1000. So I was only 50% over speed with it. It didn't like it but it didn't bite me real hard either. A Bridgeport spindle is ball bearing supported I believe, but it's only designed for 2000 or so RPM. A VFD will let you get twice that, plus. Short story is just because you can doesn't mean you automatically should.