I have just completed this past weekend the intall of a TECO VFD / phase converter on a Bridgeport J Head mill. I had posed the question about the potential ill effects to the nearly 40 year old motor if I run it up past 60 Hz. I had posted this on the Bridgeport Forum and as it evolved into this question specificly one of the other members suggested this may be a better place to pose this question due to this being the local gathering area for the brain trust on VFDs
If you would like to catch up on the entire conversation you can do that here: Bridgeport Mill Spindle Bearing Max speed? The subject kinda shifted to the VFD specificly around post #9. I hope some of you guys that I know are way more educated than myself will chime in and shed a little more light on the subject at hand.
To say that I'm not one of the experts here is a gross understatement. However, I might be a few steps ahead of you on the learning curve as I went through this a few months ago with a J head mill, and more recently, with a South Bend lathe. The BP mill has a Hitachi 3hp unit and the SB lathe a TECO.
I conducted a fair amount of research into this matter and received a significant amount of feedback from the really smart guys here.
My conclusion is: I would not say it's a "non-issue". That is to say, it IS an issue. But apparently, not a very big one. Nonetheless, it is undeniable that the VFD's re-created wave form is choppy, and hence, hard on insulation. That's why modern electric motors are certified as being VFD ready although I don't remember exactly the term they use.
So, might it very well shorten the life of an older motor? Yes. But the guys here, to a man, have added "but don't worry about it". I suppose that if some kind of original antique motor is unreplaceable, and for some reason cannot be re-wound, it might rise to the level of a bonafide worry.
Again, I'm nobody to claim expertise on this. But that was the distinct conclusion I reached based on input from some people here whose opinions I value very highly.
The main problem with any motor is thermal. Running it significantly outside its normal speed can cause a couple of issues in that regard:
1) If you run it too slow the losses may increase and air circulation (cooling) decrease, both of which can cause temperature rise.
2) If you run it too fast you can overheat the motor windings and the bearings.
My suggestion would be to buy a remote-reading thermometer. Attach the sensor to the top of the motor housing and put the readout visible from the operator position.
Run the motor at normal speed for 1/2 hour and note the temperature. Use that as your limiting value.
If you observe a significant rise above that value, turn it off and let it cool down, then resume operation.
There's no guarantee that this will avoid all problems, but it will give you reasonable confidence in operating the machine.
Spoke with a Baldor Motor / VFD rep. today about these questions, I posed the question about doubling the motor speed per the motor plate he said that was absolutly true but the torque curve will usualy suffer once you surpass the rated motor RPM range via a VFD. We also spoke about the likely hood of a new VFD on a nearly 40 year old motor causing the motor harm. He is of the very strong opinion that it will indeed cause the motor to fail more than likely sooner than later. The reasoning he has stated was pretty simple, motors back then were not made to deal with the pulsing DC current of todays VFDs, the windings insulation and the quality of the copper back then was not as good. These 2 areas alone will make the motor fail eventualy. His suggestion was to either replace the motor with a VFD rated motor or have my current motor rewound with VFD rated windings.
Just kind of curious about this line of thinking? he was very adimate about his opinions and seemed very well versed in what he was saying. I am not saying he is right or wrong, Just curious what others thoughts are about this subject.
Technically, all true. Realistically, you ALREADY own the motor, what do you have to lose? Older motor winding insulation was not made to handle the voltages seen by the pulses coming from a VFD, that is true. But it's not the normal PWM pulses themselves, because those are no higher that the true peak voltage of the incoming supply anyway. What can happen is however, the pulses cause the motor leads to act like a form of capacitor and they have an electrical ringing effect that creates what are called "standing waves" that can reach very high voltages, and although for very short duration, they repeat themselves over and over and over, so the effect is cumulative over time. The pulses can be up to 2.5X the peak line voltage (not the RMS, which is what we normally refer to), sometimes maybe a little higher. So that means in a 480V system, the peak voltage is actually 678V so the standing wave pulses can be 1700VDC. In an old 480V motor, they would use insulation that was rated for 1200V peak because it was common to use 1.5X the peak voltage so it ends up inadequate for VFD use and the demise can be very quick. But on 230V motors, the motor mfrs often used the same magnet wire rated for 1200V and since the peak line voltage is around 340V, 2.5X that is still below the insulation level. It's not guaranteed that they used 1200V or even 1000V insulation, which would still be fine, but it is very common, as in a 90% chance (so I've been told). In other words, why not give it a try? You can always get a low cost motor lead filter and get rid of the spikes anyway, they are cheaper than a new motor.
Old copper? Piffle on that one. Not buying it on a small motor. Maybe an issue with really big motors with formed coils because impurities can cause localized heating, but on a small machine motor? Nah.
But the big deal he is right about is the torque issue. Once you go above base speed, you lose torque. This is because torque is dependent upon maintaining a constant ratio of Voltage and Frequency or the V/Hz ratio. As long as you keep it the same as what the motor was designed for, you get the same torque. That's essentially what a VFD allows you to do. So if your motor is designed as 230V 60Hz, that's a V/Hz ratio of 230/60 or 3.83:1. But of you take the frequency up to 120Hz, the voltage is still 230V because that's all you have. So now your ratio drops to 230/120 or 1.92:1. Torque will vary by the SQUARE of the difference, to at 1/2 of the ratio, you will have .5 x .5 or only 25% of rated torque at that 120Hz speed.
As I am fond of saying..... old insulation may not have been as good as modern materials, but it was usually placed more carefully....... So it can be pretty darn good if it has not been heated enough to begin to break down.
The wier insulation film.... that's a minor issue, the double and triple stuff is tougher against bending and wear, but not THAT much better against voltage....
2 old motors are usually cheaper (aside from union replacement labor) than one new..... and you likely will only need one of them anyhow....
The pulses can be up to 2.5X the peak line voltage (not the RMS, which is what we normally refer to), sometimes maybe a little higher. So that means in a 480V system, the peak voltage is actually 678V so the standing wave pulses can be 1700VDC. In an old 480V motor, they would use insulation that was rated for 1200V peak because it was common to use 1.5X the peak voltage so it ends up inadequate for VFD use and the demise can be very quick. But on 230V motors, the motor mfrs often used the same magnet wire rated for 1200V and since the peak line voltage is around 340V, 2.5X that is still below the insulation level. It's not guaranteed that they used 1200V or even 1000V insulation, which would still be fine, but it is very common, as in a 90% chance (so I've been told). In other words, why why not give it a try? You can always get a low cost motor lead filter and get rid of the spikes anyway, they are cheaper than a new motor.
This is an excelent explaination and easy to understand. Thank you for explaining it this way. With all of this said, My motor on the Bridgeport is a Quad voltage / Dual Hz rated motor, Corect me if I am understanding you incorectly. With my motor being rated for up to 460V, the windings would more than likely be the higher rated insulation. That being said I am running this machine on 230V power meaning the peak voltage in theory should be abouut 1/2 of what it is rated for?, and with that said if I understand you corectly this would mean that my VFD running on a 230V circuit on my 460V rated motor I should be pretty darn safe on sending DC electrical pulses to the windings seeings how it is only at 1/2 the rated voltage... in theory.
As far as a "motor lead filter", I am not familiar with that, can you elaborate on that a little?
Thanks again for the great responce and explaination,
Joe
must have been a post on other thread: Spoke with a Baldor Motor / VFD rep. today about these questions, I posed the question about doubling the motor speed per the motor plate he said that was absolutly true but the torque curve will usualy suffer once you surpass the rated motor RPM range via a VFD. We also spoke about the likely hood of a new VFD on a nearly 40 year old motor causing the motor harm. He is of the very strong opinion that it will indeed cause the motor to fail more than likely sooner than later. The reasoning he has stated was pretty simple, motors back then were not made to deal with the pulsing DC current of todays VFDs, the windings insulation and the quality of the copper back then was not as good. These 2 areas alone will make the motor fail eventualy. His suggestion was to either replace the motor with a VFD rated motor or have my current motor rewound with VFD rated windings.
Just kind of curious about this line of thinking? he was very adimate about his opinions and seemed very well versed in what he was saying. I am not saying he is right or wrong, Just curious what others thoughts are about this subject.
forgetting insulation a minute, he said 'torque would suffer." Jraef explained how this is but left out that it probably is immaterial upto 120 hz speed: although this BREAKDOWN torque goes down by square of the voltage, breakdown torque is still way higher than actual output torque avail on the 1/2 vhz curve; so until breakdown torque's V^2 curve intersects the v/hz curve torque, you get the v.hz torque. so at 120hz, the 'rated' or available torque will be 1/2 the nameplate, not 1/4: at 120hz most motors still have breakdown torque above this so it is not an issue until higher speed.
there is no heat issue as you go above base speed: the mechanical fan in the motor cools at a cube of the speed function, so plenty of cooling - dont worry about that.
most motors can go down to 1/6 or 1/10 base speed with shaft fan cooling without overheating; no ADDITONAL heating issues due to slower speed exist, other than the fan liekwise goes slower so cooling is down by same cube function; hence the 1/6 or 1/10 base speed min rule of thumb for most motors.
the baldor guys comment that the 'pulsing dc current' is problem is wrong: it is the pulsing voltage; if you ever scope the current on your vfd output you will find it pretty darn nice and having no spikes at all.
Not sure why he would say quality of cu back then was so worse than today..... seems silly.
since your motor is good for 230/460 then you are absolutely correct that your motor will NOT be adversely effected by the 800v spikes on voltage of the vfd output on 230v!
If you want to reduce those spike anyway to be a kinder gentler owner, stick a small "3% 3 phase line reactor" in series with the motor leads; this will cut those 700-800v spikes down to 1/2-3/4 that level.
did i say going above 60hz has no bearing on voltage spikes hense no bearing on motor life? spikes will be the same at any speed.
since your motor is dual voltage, I go with JST and say "run it like a rental!"
If you want to reduce those spike anyway to be a kinder gentler owner, stick a small "3% 3 phase line reactor" in series with the motor leads; this will cut those 700-800v spikes down to 1/2-3/4 that level.
You should not put a "line reactor" or anything else between the VFD output and the motor.
It would in fact have no effect, since the inductance of the motor greatly exceeds the inductance of the reactor itself.
I do put line reactors on the VFD input, between it and the AC line, to minimize current spikes at power-up. I had one VFD fail due to an inrush spike.
6.1 Description of the Motor Filter
The steep switching flanks of modern pulse converters, combined with
long motor lines, lead to transient excess voltages on the motor terminals.
When a motor filter is used on the converter output, these excess
voltages are reduced to harmless levels and the leakage currents of the
motor lines decrease.
This provides the following advantages:
• Limitation of power surges and the peak value for reducing strain on
the winding insulation.
• Operation of motors using long supply lines or many parallel motor
lines (e.g. for group drives).
• Reduction of interfering voltages on the motor line in the case of
increased EMC requirements.
see with and without simple inductor on output voltage waveform at motor. same exact inductor that can be used on input to reduce the pwm noise getting back to the power line
You should not put a "line reactor" or anything else between the VFD output and the motor.
It would in fact have no effect, since the inductance of the motor greatly exceeds the inductance of the reactor itself.
I do put line reactors on the VFD input, between it and the AC line, to minimize current spikes at power-up. I had one VFD fail due to an inrush spike.
- Leigh
Incorrect on two counts:
A reactor connected from the output of a VFD to a motor is not called a line reactor. It is called a load reactor.
A load reactor is helpful if long runs of motor cables are used, > 50-60 feet.
Another point is that a zero-phase reactor can be used to reduce noise interference
with audio equipment.
Line and load reactors are specific to an installation.
A reactor is a reactor. Whether you call it "line" or "load" depends only on where it's installed. The device is the same.
I've never heard of a 50'+ line run on a machine tool.
Such might be common in other applications, like fans or conveyors, but that's not what we're discussing here.
actually most of the machine tools we work with made by cincinnati milacron, G&L, K&T, have much longer than 50' motor leads - typically 100-250'......
actually, the first post I listed showing the reduced spikes on the motor output with inductor is from refu vfd drive company.
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.
Actually most a majority of vfd mfgrs use 50' as the max value here too.
Most very high performance vfd drives and motor systems use output reactors for the above mentioned reasons. for example, when you have a 75hp 12000rpm spindle motor in a 8" dia x 10" long package, it is required to squeeze so much power from such a small package....
yes, inductors ARE used on vfd outputs and for very good reasons.
Hello,
I used to work in the winding department at Marathon Electric. We made VFD rated motors as well as standard. For HIpot testing, we regularly went up to 1700 volts on low voltage motors (under 600 volts) and used the same wire for low voltage units. We also made medium voltage units, which were formed coils (square wire). THis was also varnished, then it was wrapped with one layer of mica and two layers of fiberglass. VFD rated low voltage motors had a different type of varnish, it was more like a plastic than varnish. From watching failed motors and generators, lot of them run on VFDS, heat is by far what kills a motor (electrically). Make sure the windings are clean, the vents and screens open, and fan blades in good condition, and I would not worry about it.
Joe
A reactor is a reactor. Whether you call it "line" or "load" depends only on where it's installed. The device is the same.
I've never heard of a 50'+ line run on a machine tool.
Such might be common in other applications, like fans or conveyors, but that's not what we're discussing here.
- Leigh
We cannot be concerned with what you have never heard of. Your idea of a reactor is some wire wound around
some pieces of galvanized pipe. That is pure guess work.
Yes, I said earlier that my comments were about short runs. Long runs with high line inductance are a different issue.
I would never remove a VFD output over long lines. Poor engineering practice to configure a system that way. It's much
better to put the VFD at the motor and remove the controls and supply lines. That's the configuration the VFDs expect.
For long lines I expect you put the reactor at the VFD output, to attenuate the transients down-stream.
Leigh, inductors on vfd outputs is not limited to long wire runs - that is just one example of place it can help. they can help an installation with only 2 foot long motor leads.
It is a fact that can been seen with your own eyes if you own a scope and want to see the vfd output side and then the motor side of a line reactor. Here is another excerpt from a vfd mfgr:
Output side of the drive
When installed on the output side of the drive, line reactors
protect the drive from short circuits at the load. Voltage and
current waveforms from the drive are enhanced, reducing
motor overheating and noise emissions
Leigh, please look at the following scope picture if you don't want to scope it yourself to see. No amount of "I don't believe its" will make the facts go away - output choke reduces voltages spikes at motor, reduces noise, reduces heat in motor, and makes it audibly quieter.
that being said, is it for everyone with a vfd? 'course not. if one has an inverter rated motor, or a 230/460v motor running on 230v it should be able to easily withstand the 600-700v spikes shown above. Do most people need the extra 2-5% cooler motor the inductor will give for a given load? no. is the audible noise an issue worth spending a few hundred dollars for most folks? no. does the avg guy need more motor short circuit protection? no. but all these reasons it is not required still does not change the fact that is CAN and IS put on the output side of many vfds, with the blessing of the vfd mfgr, and for good reasons. in fact, we all know capacitance on the output side of a vfd is a bad thing; but we still make LC filters where we take the reactor and ADD capacitors too! with this we can tune the LC to be filter and produce near perfect current AND voltage sine waves.
7Likes
LinkBack URL
About LinkBacks
Phase converter / VFD effects on old motor?
Reply With Quote
Originally Posted by Jraef
Bookmarks