Isolation Transformers for small VFD's

[mull982]

I notice that many VFD's have isolation transformers on the line side of the drive. I believe these are mainly for harmonics or surge mitigation, or maybe something else that I am missing?

My question is, if there is a cut-off point, or rule of thumb for which an isolation transformer is not required below a certain hp drive? Do smaller drives need isolation transformers for the reasons mentioned above? Should all drives have them, or only certain sizes?

I also notice that many drives now come with line reactors. Is the line reactor an acceptable substitute for an isolation transformer? Will it acomplish all the things mentioned above? Is one better than the other?

 

[PowerQualityDoctor]

VFD's have different harmonic filtration solutions. Many of them comes with standard filtration (inline reactor or transformer) and have optional filters. Minimal filters offered are for line side and motor side.

What to choose depends on the harmonic content of your site. The higher your harmonic content, the better filter you should use.

And of course, always ask yourself if you really need the VFD or just use it because everyone does. I saw many VFDs working on fundamental frequency. In this cases - either bypass them or use better motor control.

 

[Jraef]

VFDs don't really need line isolation transformers any more, but many people still think they do. In the "olden days" of DC drives and older AC Current Source Inverters, the method of achieving voltage control meant switching of SCRs on the line side, which created a lot of common mode electrical noise. To mitigate that, it was necessary to use what were/are called Drive Isolation Transformers. They were different only in that they were usually a 1:1 ratio (no step up or down), shielded with respect to ground and rated to handle the high harmonics without damage from heating. Because of this issue, a lot of companies have older design specs that require Drive Isolation Transformers and have never been updated, so people still put them in rather than buck the system.

But modern PWM drives no longer create common mode noise the way the older ones did, so the isolation transformer is unnecessary. It doesn't hurt anything but throughput efficiency so there is no problem with it, it just isn't needed. Sometimes they had them there for an older drive that has been replaced with a newer one and although they aren't required, they leave them in anyway.

One place you do still see them however is on larger drives where they have harmonic mitigation requirements and require 12 pulse or 18 pulse front ends. This means instead of a simple 6 diode bridge rectifier, they have two or 3 sets of diode bridges in parallel, each one fed by a transformer that is Delta-Y or Delta-Delta, which creates a phase shift feeding each set of diodes. This makes more of the harmonics created by the diode bridges cancel each other out, reducing the harmonics seen by the supply.

Typically you will not see this deemed necessary unless the VFD(s) represents a significant amount of the total load of a system. Harmonics mitigation is usually necessary when the voltage THD (Total Harmonic Distortion) exceeds 5% at the point of common coupling, so smaller drives don't often contribute enough to worry about. But there is no specific size at which it becomes an issue because it's always relative to the total load.

Reactors however offer added benefits besides just harmonic mitigation. They can also serve to help protect the VFD from line disturbances that could damage the diode bridge. So some people like to put them in every installation as a form of cheap insurance.

 

[ELA]

VFDs don't really need line isolation transformers any more, but many people still think they do. In the "olden days" of DC drives and older AC Current Source Inverters, the method of achieving voltage control meant switching of SCRs on the line side, which created a lot of common mode electrical noise. To mitigate that, it was necessary to use what were/are called Drive Isolation Transformers.

But modern PWM drives no longer create common mode noise the way the older ones did, so the isolation transformer is unnecessary.

Jraef,
It has been a long time since I worked with VFDs. The last ones I worked with were in the 1990s.
From your posts I can see that you have alot of experience with them.
I am curious how it is that modern PWM drives no longer create common mode noise?

In order to CE mark our equipment in the 90's we had to meet European conducted EMC requirements. In order to meet these requirements all of the drives had to have rather large external EMI filters connected to them. I can see where an isolation transformer could also be an option for reducing conducted EMC emissions.

Do these modern drives you speak of have internal EMI filters? I would expect that if they do not that they would require either an external EMI filter or isolation transformer if they were required to meet certain EMC standards.

 

[StephenSDH]

VFDs don't really need line isolation transformers any more.

I don't have any personal experience with having issues which required isolation transformer or line reactors to solve. Do you feel the same way about line reactors? What about adding the isolation transformer to add some impedance the the system.

Alot of the low cost drives today don't even have a DC Link Choke. From the manuals I have read, if you have a stiff system it is good to add line reactors or isolation transformer to reduce harmonics and reduce damage to the drives capacitors and circuitry from high inrush. In addition it helps protect from transients.

 

[StephenSDH]

What to choose depends on the harmonic content of your site. The higher your harmonic content, the better filter you should use.

The drive doesn't care about voltage harmonics at your site. It is just a simple 6-pulse rectifier circuit with a capacitor. You add filters to protect other devices from the harmonics the drive will produce.

 

[jdsmith]

I don't have any personal experience with having issues which required isolation transformer or line reactors to solve. Do you feel the same way about line reactors? What about adding the isolation transformer to add some impedance the the system.

Alot of the low cost drives today don't even have a DC Link Choke. From the manuals I have read, if you have a stiff system it is good to add line reactors or isolation transformer to reduce harmonics and reduce damage to the drives capacitors and circuitry from high inrush. In addition it helps protect from transients.

We had an instance a few years ago where some 5 HP 480V drives without line reactors or isolation transformers were tripping off every morning. After some investigation it was determined that the utility capacitors on the 138 kV station at our plant caused enough of a transient at the 480V level to trip the drives off. My boss and another engineer added 5% line reactors to insulate the drives from the transients and the trips stopped. Curiously enough, there were other "identical" drives fed from the same 480V bus that were in service for several years with no problems. I was not personally involved in this and those that were didn't take the time to figure out why two drives were tripping when 4 others were fine. I'm not sure if they were a slightly different series of the same manufacture, or if they were specified identically and the manufacturer changed the design somehow.

Keep in mind that harmonic issues are related to the ratio of load with electronic power supplies compared to the amount of inductive load on a particular bus. In my plant we have around 1000 480V motors, only about a dozen scattered around the plant have drives on them so our harmonics are hardly noticeable.

 

[Jraef]

I don't feel the same about reactors as I do about isolation transformers. I said in my earlier post that I think reactors are a great idea even if you are not worried about harmonics.

Yes, an isolation transformer could add impedance and/or help with harmonics, but so will a reactor and reactors are a lot cheaper and smaller. What you no longer need is the shielding to filter out the switching transients that were a result of GTOs used in older drive systems. Nobody uses them any longer, the rectifiers are now passive diode bridges, or they are transistor based active front-ends (in the case of regen drives) and the switching noise is just not there.

ELA,
CE is an EU/IEC thing, we don't have that requirement here. Most VFDs sold for IEC countries do need EMI filters, most of those use reactors instead of isolation transformers; again they are cheaper / smaller.

 

[StephenSDH]

I said in my earlier post

Thanks. Missed That.

 

[ELA]

ELA,
CE is an EU/IEC thing, we don't have that requirement here. Most VFDs sold for IEC countries do need EMI filters, most of those use reactors instead of isolation transformers; again they are cheaper / smaller.

US based industries that are savvy to the problems associated with EMC also require IEC testing and compliance with EMC standards. One example is the Semiconductor industry.

I did some quick searching on VFD's and you find that most of them either incorporate built-in EMI filters or have a disclaimer stating that an external filter is required in order to meet EMC standards. These drives, unprotected by an EMI filter, most certainly do emit common mode noise.

 

[mull982]

Thanks for the great information.

VFDs don't really need line isolation transformers any more, but many people still think they do. In the "olden days" of DC drives and older AC Current Source Inverters, the method of achieving voltage control meant switching of SCRs on the line side, which created a lot of common mode electrical noise. To mitigate that, it was necessary to use what were/are called Drive Isolation Transformers. They were different only in that they were usually a 1:1 ratio (no step up or down), shielded with respect to ground and rated to handle the high harmonics without damage from heating. Because of this issue, a lot of companies have older design specs that require Drive Isolation Transformers and have never been updated, so people still put them in rather than buck the system. .

But modern PWM drives no longer create common mode noise the way the older ones did, so the isolation transformer is unnecessary. It doesn't hurt anything but throughput efficiency so there is no problem with it, it just isn't needed. Sometimes they had them there for an older drive that has been replaced with a newer one and although they aren't required, they leave them in anyway.

One place you do still see them however is on larger drives where they have harmonic mitigation requirements and require 12 pulse or 18 pulse front ends. This means instead of a simple 6 diode bridge rectifier, they have two or 3 sets of diode bridges in parallel, each one fed by a transformer that is Delta-Y or Delta-Delta, which creates a phase shift feeding each set of diodes. This makes more of the harmonics created by the diode bridges cancel each other out, reducing the harmonics seen by the supply.

Typically you will not see this deemed necessary unless the VFD(s) represents a significant amount of the total load of a system. Harmonics mitigation is usually necessary when the voltage THD (Total Harmonic Distortion) exceeds 5% at the point of common coupling, so smaller drives don't often contribute enough to worry about. But there is no specific size at which it becomes an issue because it's always relative to the total load.

Reactors however offer added benefits besides just harmonic mitigation. They can also serve to help protect the VFD from line disturbances that could damage the diode bridge. So some people like to put them in every installation as a form of cheap insurance.[/QUOTE]

 

[mull982]

Thanks for the great information. Sorry I messed up on my last post.

VFDs don't really need line isolation transformers any more, but many people still think they do. In the "olden days" of DC drives and older AC Current Source Inverters, the method of achieving voltage control meant switching of SCRs on the line side, which created a lot of common mode electrical noise. To mitigate that, it was necessary to use what were/are called Drive Isolation Transformers. They were different only in that they were usually a 1:1 ratio (no step up or down), shielded with respect to ground and rated to handle the high harmonics without damage from heating. Because of this issue, a lot of companies have older design specs that require Drive Isolation Transformers and have never been updated, so people still put them in rather than buck the system.

Can you explain the theory or point me to a reference of how these drives used to create common mode noise? I've heard the term before but never really grasped the understanding. How were the transformers you mentioned shielded, and how did this help with this common mode noise?

But modern PWM drives no longer create common mode noise the way the older ones did, so the isolation transformer is unnecessary. It doesn't hurt anything but throughput efficiency so there is no problem with it, it just isn't needed. Sometimes they had them there for an older drive that has been replaced with a newer one and although they aren't required, they leave them in anyway.

Yes I have seen alot of companies that we purchase drives from push an isolation transformer. There is always an explanation of the Delta winding on the transformer somehow trapping harmonics.

One place you do still see them however is on larger drives where they have harmonic mitigation requirements and require 12 pulse or 18 pulse front ends. This means instead of a simple 6 diode bridge rectifier, they have two or 3 sets of diode bridges in parallel, each one fed by a transformer that is Delta-Y or Delta-Delta, which creates a phase shift feeding each set of diodes. This makes more of the harmonics created by the diode bridges cancel each other out, reducing the harmonics seen by the supply.

We do have larger drives in the 350hp and 500hp range that have isolation transformers. I believe these drives are still 6-pulse drives though. If these drives are 6-pulse drives, do we still need these large isolation transformers?

I always thought that the pulse rating of the drive had to do with the output transistors as opposed to the diodes on the front end.

Typically you will not see this deemed necessary unless the VFD(s) represents a significant amount of the total load of a system. Harmonics mitigation is usually necessary when the voltage THD (Total Harmonic Distortion) exceeds 5% at the point of common coupling, so smaller drives don't often contribute enough to worry about. But there is no specific size at which it becomes an issue because it's always relative to the total load.

So if the vfd only represents lets say 10% of the system load then is it not really necessary? At what system loading point would you say it might require consideration for an isolation transformer? Or is it more or less related to THD as you mentioned? Even so it sounds like a reactor can be used to help with this issue.

 

[Jraef]

This is a very good paper on the issue, covers almost everything I have said actually.

Yaskawa paper on drive harmonics

 

[RETRAINDAILY]

would you use a isolation transformer to boost the aic rating of a drive?

 

[weressl]

I notice that many VFD's have isolation transformers on the line side of the drive. I believe these are mainly for harmonics or surge mitigation, or maybe something else that I am missing?

My question is, if there is a cut-off point, or rule of thumb for which an isolation transformer is not required below a certain hp drive? Do smaller drives need isolation transformers for the reasons mentioned above? Should all drives have them, or only certain sizes?

I also notice that many drives now come with line reactors. Is the line reactor an acceptable substitute for an isolation transformer? Will it acomplish all the things mentioned above? Is one better than the other?

I let you in on a dirty little secret....

The ASD market is so competitive that distributors can not make money on them so they 'sell' the various add on's to make profit. Seldom do you need either line or load reactors.