Jump to content


Photo

DC chokes vs line reactors: Opinions?


  • Please log in to reply
9 replies to this topic

#1 jraef

jraef

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 683 posts
  • Gender:Male
  • Location:USA, California

Posted 01 August 2006 - 01:19 AM

I have been involved in a project to standardize a large aggregate industry customer on a VFD product line so that he has only one to learn and support throughout his facilities. In the process, I recommended (as I almost always do) that he incorporate line reactors in his drive packages because his environment is prone to some nasty power problems (i.e. digging up and shorting out cables as they did this morning) and it is my opinion and experience that VFDs with line reactors tend to survive major power events more readily than those without.

Vendors that are vying for this business have been attacking this issue from different angles, attempting to convince him that I am all wet in this regard and that they have different ways to solve these problems. One has said they have a "swinging choke" instead and that this precludes the need for line reactors (if you Google "swinging choke" you will see who it is, I don't want to give them free advertising). As I see it, DC bus chokes, "swinging" or not, help to mitigate current harmonic distortion issues. To my thinking however, they do NOT provide much in the way of component protection from line (or load) faults.

That said, I'll admit I have never been completely clear as to why line reactors help with survivability other than to assume that the reactive time constant that they introduce can allow the VFD's protective circuits to have time to react to an event before damage is done and also tend to be a low pass filter that clips the highest peaks off of the events. Am I all wet here? Would / could a DC choke do the same thing?

Just as a side note, several vendors have attempted to pass off their Level 1 EMC filtering as being the same (or even better) than a line reactor! They obviously were not listening to what I said was the problem, but that shouldn't surprise you.
"He's not dead, he's just pinin' for the fjords!"

#2 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,603 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 01 August 2006 - 02:01 AM

Hello jraef

There are a number of different issues here.

If we look at the basic drive, low cost approach, then the input rectifier is connected directly to the capacitors via some slow charge means. The three phase rectifier acts as a peak rectifier and only conducts on the paek of each sinewave. This creates major harmonic currents that extend to a very high order. Additionally, if there is any imbalance on the incoming phases, you quickly get the input current flowing on two phases rather than three. If there is a supply voltage transient or surge, this is directly applied via the rectifier to the DC Bus capacitors and will result in a high charging current. This current can destroy the capacitors, supply fuses, and input rectifier.
To reduce the harmonic currents, and to spread the input current across all three phases, a DC bus choke can be added. This has the added advantage of reducing the charging current when there is a surge voltage applied. If the DC Choke is large enough, the input current on each phase will flow for up to 120 degrees each half cycle. This will minimize but not eliminate the harmonic currents.

If the DC bus choke is in the positive leg of the DC supply and there is a ground fault on the output of the inverter, the positive current from the supply (not from the DC Capacitors), will be limited due to the impedance of the DC Choke. The neagive leg will not be limited so you will get full fault current flowing in the negative side of the rectifiers and inverter bridge. The standard DC bus choke does not help to protect the negative leg of the inverter or rectifier. A second DC choke can be added to the negative leg, or the DC choke can be split with one half in the positive leg and one half in the negative leg. This will have the same advantages in regard to harmonics but will offer some improvement in output short circuit fault current limiting. The balanced DC bus choke is preferable to the single DC bus choke because it then presents a balanced interface to the supply. If a single bus choke is used, there wil be some ripple on the centre pooint of the inverter because it is more closely coupled to the negative rectifiers than the positive rectifiers. A balanced choke arrangment will also offer improved noise suppresion between the output stage of the inverter and the AC input circuit.

The use of line rectors can be instead of or in addition to the DC Bus Choke. The line reactors provide a balanced impedance between the inverter output stage and the incoming supply. They will provide the best short circuit protection for the output stages of the inverter and the best noise filtering action, both into the inverter and out of the inverter due to the defined impedance at line fequency.

Most VSDs are either no reactor, or single leg DC Bus chokes. The addition of line reactors will definitely improve the reliability of the drive by reducing the susceptibility to output circuit problems and line bourne voltage problems.

Best regards,


#3 kens

kens

    Member

  • Full Member
  • PipPipPipPip
  • 148 posts

Posted 01 August 2006 - 02:33 AM

Hi jraef, I come from a background in HVAC so I have a strong preference for DC chokes. This is not really related to survivability though. We used a lot of IP rated drives so we could mount them in the plant room directly rather than in a cabinet. With the line reactors this usually required a seperate enclosure and therefore more work/expense. The chokes or reactors were required for harmonic purposes only so I must say I have very little experience in which works best in the limitation of damage.
Another problem is the voltage drop and losses across the reactors, this needs to be taken into account also.

By the way the "Swinging Choke" would appear to be about as effective as standard DC chokes when you really drill into their data regardless of how fabulous it is made to sound.

QUOTE(marke @ Aug 1 2006, 02:01 PM) View Post


The use of line rectors can be instead of or in addition to the DC Bus Choke. The line reactors provide a balanced impedance between the inverter output stage and the incoming supply. They will provide the best short circuit protection for the output stages of the inverter and the best noise filtering action, both into the inverter and out of the inverter due to the defined impedance at line fequency.


Best regards,


Marke, nice explaination, i assume that you are talking about Harmonic noise here? It is my understanding that line reactors have no effect on EMC noise.

I still like to see SCR fusing before large drives.

Kens


An expert is one who knows more and more about less and less until he knows absolutely everything about nothing

#4 frs

frs

    Junior Member

  • Full Member
  • PipPip
  • 8 posts

Posted 01 August 2006 - 12:36 PM

Hello jraef,

I think that "swinging choke" definition related to reactor characteristic curve, not to installation point.

If manufacturer have line reactor (or "line commutating choke") inside the VFD casing, that is good. But I never seen that.
Manufacturers like to offer line reactor as an option. Which is required almost always...

#5 mariomaggi

mariomaggi

    Senior Member

  • Full Member
  • PipPipPipPipPip
  • 230 posts
  • Gender:Male
  • Location:Milan, Italy
  • Interests:electric vehicles http://www.evlist.it , 3-phase AC sources to test PV inverters, high-power AC/DC power supplies up to 1500 VDC to test photovoltaic inverters, bidirectional AC/DC power supplies to test and simulate big batteries, infrared windows for thermography, renewable energies, special electric motors, special inverters, energy savings, power quality

Posted 01 August 2006 - 12:56 PM

Dear all,
my background is limited in this specific argument, but I would add some words.

In my opinion it is possible to have the maximum reliability also without line reactor, if there are good balanced DC chokes (on positive and negative legs). An RFI filter is always suggested, and also some fast overvoltage protection at least for spikes.
Line reactors are more effective for harmonic mitigation, but probably there are special solutions for DC chokes - not yet available on massive production - that could have a comparable level of efficiency. For example, DC chokes with permanent magnets in the core airgap, with non-linear windings (winding the coil, the linear translation speed of wire must increase).

This document could be useful to make some considerations.

I've few theoretical basis, don't ask me why I suggest this solution.

Regards
Mario

Mario Maggi - Italy - http://www.evlist.ithttps://www.axu.it


#6 frs

frs

    Junior Member

  • Full Member
  • PipPip
  • 8 posts

Posted 01 August 2006 - 01:48 PM

Here is copy-paste from catalogs of two major manufacturers:

Manufacturer 1:
QUOTE
The line chokes provide improved protection against mains overvoltages and reduce the current harmonic distortion produced by the speed drive.
The chokes recommended are used to limit the line current.
The use of line chokes is especially recommended in the following cases:

severe disturbance of mains supply by other receivers (interference, overvoltages),

mains supply with voltage imbalance between phases > 1.8 % of the nominal voltage,

speed drive supplied by a line with very low impedance (close to power transformers 10 times more powerful than
the speed drive rating),

installation of a large number of frequency inverters on the same line (reduction of line current),

reduction of overload of the cos phi power factor correction capacitors, if the installation has a bank of power factor correction capacitors.


Manufacturer 2:
QUOTE
Line commutating chokes are used to smooth voltage peaks or to bridge commutating
dips. In addition, line commutating chokes reduce the effects of harmonics on the inverter and the power supply. If the line impedance is < 1 %, a line commutating choke must be used in order to reduce the current peaks.


QUOTE
A line reactor is needed for high system fault levels, partly to protect
the actual converter against excessive harmonic currents,
and thus against overload, and partly to limit the system perturbation
to the permitted values. The harmonic currents are limited
by the complete inductance comprising the line reactor and
mains power input inductance. Line reactors can be omitted if
the line infeed inductance is increased sufficiently, i.e. the value
of RSC must be sufficiently small.
RSC=Relative Short-Circuit power: Ratio of short-circuit
power Sk line at the supply connection point to fundamental
apparent output Sconv of the connected converters.

The following applies for xxxxxxx drive converter
cabinet units:

Output .............Line reactor can be omitted
kW ......................for RSC
< 200 ......................≤ 43
200 to 500 ...............≤ 33
> 500 ......................≤ 20

In practice, the line configuration on which individual converters
operate is often not known, i.e. the line short circuit power at the
connection point of the converter is not known, so it is recommended
that a line reactor always be connected in series with
the converter.

The line reactor can only be omitted if the values for RSC are
lower than those shown in the above table. This is the case
when, as shown in the following figure, the converter is connected
to the line through a transformer with the appropriate rating.

Attention: A line reactor is always needed, however, if a line filter
is used.


jraef, if you dig through catalogs/documentation provided by applicants you may find similar recommendations. Then it will be difficult for applicants to defend their position.


#7 jraef

jraef

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 683 posts
  • Gender:Male
  • Location:USA, California

Posted 02 August 2006 - 12:04 AM

Thanks everyone. This confirms my own bias, that reactors are important in terms of survivability of the VFD, but DC chokes may not be as useful in that regard.

Mark, do you know of anyone using balanced DC chokes on both legs? I have never seen this. You would think this would be a tangible benefit, although I'm sure that it would add cost, which is never palatable to drive mfrs. I gather though that even that version would not really protect the VFD from line side problems, only load side.
"He's not dead, he's just pinin' for the fjords!"

#8 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,603 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 02 August 2006 - 01:25 AM

Hi jraef

I believe that the level of output protection provided by the DC bus choke is somewhat limited compared to the line reactors and would suggest that the prime use is to improve the rectifier conduction angles and reduce the harmonics.

I have seen balanced bus chokes and I think that one of the earlier ranges of PDL may have used that technology, but I can not be sure. Certainly, it is not the normal approach. If you take the standard choke that is used an just split the winding in half with one half in the positive leg and one half in the negative leg, the additional cost is not too bad, certainly not as bad as two separate reactors.

The use of the DC bus choke and or line reactors does actually improve the EMC performance of the drive. It is much easier to filter a drive with a reactor or choke than one without. The reactor and/or choke will not provide the required filtering, but they will lift the impedance to allow the filter to work much better.

Best regards,

#9 Dinge

Dinge

    Junior Member

  • Full Member
  • PipPip
  • 5 posts

Posted 10 February 2011 - 06:31 AM

QUOTE (marke @ Aug 2 2006, 12:25 PM) <{POST_SNAPBACK}>
Hi jraef

I believe that the level of output protection provided by the DC bus choke is somewhat limited compared to the line reactors and would suggest that the prime use is to improve the rectifier conduction angles and reduce the harmonics.

I have seen balanced bus chokes and I think that one of the earlier ranges of PDL may have used that technology, but I can not be sure. Certainly, it is not the normal approach. If you take the standard choke that is used an just split the winding in half with one half in the positive leg and one half in the negative leg, the additional cost is not too bad, certainly not as bad as two separate reactors.

The use of the DC bus choke and or line reactors does actually improve the EMC performance of the drive. It is much easier to filter a drive with a reactor or choke than one without. The reactor and/or choke will not provide the required filtering, but they will lift the impedance to allow the filter to work much better.

Best regards,


#10 Dinge

Dinge

    Junior Member

  • Full Member
  • PipPip
  • 5 posts

Posted 10 February 2011 - 06:38 AM

Mark

You are quite correct, the PDL ADSDi series used a split choke in the DC Bus.
PDL were innovators and pioneers in many aspects of electronic motor control.
Sad ending for a world leading kiwi company.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users