Customer has a couple VFD's (Pump Smart PS75) One feeds a 7.5HP pump motor and the other a 15HP pump motor. I was told they are periodically tripping out on what they say are voltage spikes when it storms and they are having to reset them. They are fed from a 240V 3-phase panel with high leg. I'm thinking as a solution to just put TVSS on the panel or should I be looking into line reactors? Thanks in advance.
VFD's Tripping
- Thread starter DAWGS
- Start date
- Status
ptonsparky
Tom
- Occupation
- EC - retired
Line reactors.
Some drives don’t like that high leg.
More knowledgeable persons will chime in eventually.
Cow
Senior Member
- Location
- Eastern Oregon
- Occupation
- Electrician
For what you've described, a line reactor can only help.
I've started installing 3% TCI KDR style reactors on a good majority of VFD's now. I prefer TCI's lineup over MTE's, just for reason that TCI's enclosures are easier to get conduit/flex in and out of.
I've started installing 3% TCI KDR style reactors on a good majority of VFD's now. I prefer TCI's lineup over MTE's, just for reason that TCI's enclosures are easier to get conduit/flex in and out of.
- Location
- San Francisco Bay Area, CA, USA
- Occupation
- Electrical Engineer
Yes, line reactors. What a line reactor does mainly for a VFD is to slow down the rise time of any transients on the incoming line side, so that they don't cause the dc bus to spike up and either trip the drive, or damage the capacitors and diode bridge. I call them "cheap insurance" fro expensive drives.
If both drives are in the same place, you can use one line reactor sized as if it is a 25HP load, it will provide the exact same benefits. But that depends on how your circuit is run, i.e. you would need to have a single feeder going to both drives, then each drive has its own OCPD.
If both drives are in the same place, you can use one line reactor sized as if it is a 25HP load, it will provide the exact same benefits. But that depends on how your circuit is run, i.e. you would need to have a single feeder going to both drives, then each drive has its own OCPD.
Surge arresters eliminate transients, not just slow them down. The original surge arresters from the 1930s to 1990s were gapped SiC which has a delay as the SiC diodes took a while to react so parallel surge capacitors or series reactors delayed and flattened out fast rising transients. Today we use a MOV which is made of metal oxides, mostly ZnO. These are not diodes but nonlinear resistors. There is no delay. Voltage clamping is instant since they work on a different principle. So MOVs alone don’t need line reactors or surge capacitors. The VFD has its own small switching MOVs but these aren’t meant for protection from lightning. That is better implemented as part of an aggressive program.
Fast transients with SCRs cause self commutation as mentioned. This does not happen with free wheeling diode front ends or IGBTs. The big problem with these is the gates are physically thin and a transient can easily arc through and burn up the device.
So I would opt for the surge arresters solution,
Sent from my iPhone using Tapatalk
Fast transients with SCRs cause self commutation as mentioned. This does not happen with free wheeling diode front ends or IGBTs. The big problem with these is the gates are physically thin and a transient can easily arc through and burn up the device.
So I would opt for the surge arresters solution,
Sent from my iPhone using Tapatalk
synchro
Senior Member
- Location
- Chicago, IL
- Occupation
- EE
IGBTs are also subject to latch up because of their NPNP structure like in an SCR (which I like to think of as an NPN and a PNP connected with positive feedback):
https://www.google.com/url?sa=t&sou...QIARAB&sqi=2&usg=AOvVaw3mKJsBE04TJHUfb5-iNHNM
I agree that a surge arrester would eliminate most of the energy of a spike instead of spreading it out in time, if that is what is causing the tripping.
Last edited:
The TVSS solution would be cheaper.
ptonsparky
Tom
- Occupation
- EC - retired
That is not what I would have expected and I am not arguing that point at all.
Line reactors would reduce the di/dt. I personally would not use them for spike suppression. Make that have never used them for that purpose.
ptonsparky
Tom
- Occupation
- EC - retired
I do remember now, that I found that out a year or so ago. It was easier for me to fix the problem because I could identify and correct it with a low tech solution. Thanks to those here.
ptonsparky
Tom
- Occupation
- EC - retired
OP, it would be advantageous to actually know what the fault codes are.
Sounds like that wasn't possible. If a VFD takes a hit and it shorts or burns up the front end quite often power for the control board is gone too so no fault codes.
Sent from my SM-T350 using Tapatalk
ptonsparky
Tom
- Occupation
- EC - retired
‘Tripping out’ means a reset is possible, to me anyway.
They are supposed to let me know Monday what the codes are. Sounds like mixed responses on whether to go with panel surge protection or line reactors.
- Location
- San Francisco Bay Area, CA, USA
- Occupation
- Electrical Engineer
Surge protectors are for high energy rapid rising spikes on the line, well in excess of the line voltage, as in 10x the line voltage. In addition, most VFDs will have MOVs on the rectifier anyway. Spikes that make it past SPDs or MOVs tend to cause damage, not tripping. Lower level surges go right through SPDs and MOVs, that's where the line reactors help. They slow down the rise time of the surges and if the drive is running, the energy is dissipated into the motor. If the drive is off but connected and the surge is taking place for a relatively long time, as in seconds (vs mSec.), it may not help, but neither will SPDs.
They slow down di/dt. For spikes, the concern is dv.
synchro
Senior Member
- Location
- Chicago, IL
- Occupation
- EE
Yes, the current through a series inductor is the integral of the voltage across it (plus a constant) and so it will slow down the rate of rise (and the rate of fall) of the current. Of course if you wanted the highest protection level you would add a surge protector in front of a line reactor to clamp any large voltage spikes before the inductor can store much energy from a voltage spike.
It would be interesting to know what the stray capacitance is across a typical reactor inductance due to the interwinding capacitance. That would limit how much it could slow down fast rising voltage spikes
- Location
- San Francisco Bay Area, CA, USA
- Occupation
- Electrical Engineer
His concern is "tripping out", which unless he returns and discloses the meaning of that, means it is all speculation at this point. But as a general rule, I recommend line reactors on almost all VFD applications, they are cheap insurance.
They also limit fault current.
And there we get to the purpose of a line reactor. If you have a small drive on a big transformer it cannot handle the available current. It can also block harmonics (the “rabbit ears” current pattern).
- Status