Performance of Matrix Harmonic Filters - for the eyes of Drive Experts 11

[edison123]

In another thread, jbartos referred an interesting article by MTE corp about the performance of the Matrix Harmonic Filter (with 6 pulse rectifier) claiming it to be the best bet as compared to 12/18 pulse rectifiers with line reactors for harmonic reduction.

The article can be downloaded/read at (thx to jb)

http://www.mtecorp.com/mitigation.pdf

I am interested to know from end users/drive experts whether all that is claimed in this article is true enough; or it is just a hard sell rubbishing competitors.

 

[Marke]

Hello edison123

Unfortunately, there is very little information about how this beast works. I have seen some information about filters for drives where they have a similar description and some very good claimed results. From what I have been able to glean so far, these are dsigned a little bit like a transmission line with a low frequency cutoff.

I have seen information coming and going for some years about this type of filter, but apart from the manufacturers data, I have not seen any other information.
I contacted one manufacturer for more information, even offered to buy one and install it in an ideal installation that was in desperate need of the quoted results. Things all of a sudden when quiet. I am not sure why, but it may have been that there was an inactive local agent. Often happens down here. The typical scenario, is that an Australian company collects the agency to add to their list, claim New Zealand as part of their territory, and then sit on the product etc effectively blocking access to us down under, and not even selling well in Aus!!

In theory, I can see merit in this type of approach, my concern is how this would perform on a high impedance inductive supply as is common out here in the rural areas. We find that supply resonance is a major issue and I wonder how this type of filter would behave under these conditions. Would it become part of a resonant supply??

I would be interested in any further information on these filters.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[jbartos]

Suggestion: Visit

http://www.mirusinternational.com

or

http://www.mirusinternational.com/pages/faq.html#uhf

for:
Lineator-Matrix Passive Harmonic Filter
and

http://www.mirusinternational.com/downloads/uhf_faq12.pdf

http://www.mirusinternational.com/downloads/uhf_faq19.pdf

for: Details

 

[edison123]

Thx Marke. While they have gone to town about minuses of 12/18 pulse rectifiers, they sure are pretty coy about their own product, which is why I posted it here. jb has given me another 35 pdf files to mull on i.e. when I finish downloading the damned things.

Thx jb. I don't know how you manage to catch those sites in the vast ocean of internet. Have you ever given a serious thought to own a dot.com start up to scare the pants off google?

 

[Skogsgurra]

Edison,

Claims like those that Mirus publish can be pure nonsense - or wonderful news. It is difficult to say which, and I would like to believe what they say. But I find it difficult for several reasons. One is that the technology ought to be known by most of us if it really were as good as Mirus says. Another reason is that the principles behind the UHF are kept away from the public. The third reason is that some of their material makes one wonder if they are careless or if they are somewhat bending the facts.

One example: [

http://www.ecsintl.com/pdf/Outperform K-rated.pdf

] This article compares performance of K-rated transformers vs the LINEATOR. Figure 3 shows that flat-topping reduces the DC link voltage in an SMPS - and we all agree on that one - but it also shows that the droop rate is more than twice as fast when the DC link voltage has been reduced (red curve) with ten percent. It is true that the inverter consumes more current when the DC input is reduced, but the droop rate should not be more than about 10 percent faster when starting from 90 percent of rated voltage (constant power to the load).

I wish that what they claim is true, but I do not find any explanation as to how they do it and that makes me cautious. Keep us posted if you find out how the LINEATOR works.

 

[edison123]

Thx skogs. I am pretty sceptic about some of the claims made in the article cited. Anyway we are also looking at other options to compare.

 

[jraef]

I have used both the Matrix Filter from MTE and the Mirus units at customer's requests. Both seem to work fine, at least when first installed. My concern on the MTE units is what happens when the system changes for some reason (as they always do). The Mirus unit is basically just a zig-zag transformer, a technology known for years as a harmonic mitigation technique (in response to skogsgurra's comment about it being known before). In most cases it will work fine if sized right, and it is fairly inexpensive compared to 18 pulse front ends.

The problem I see with all of them is that they all sell only one solution, often at the expense of another or any other. When you read through the sales literature you can't help but notice their slanted view and I for one am a little skeptical. The only reason why I ended up trying both was because I let my customers decide on their own and I stayed out of it. Were I to be put on the spot to decide which way to go I would opt for an active harmonic filter and damn the cost.

"Venditori de oleum-vipera non vigere excordis populi"

 

[Skogsgurra]

Thanks jraef! Snake oil again?

It would be nice if they had told us. The way the UHF is presented made me think that there is a new unknown way of doing it. We have been using Z transformers for improving short-circuit power on small generating units for many years and they also improve the V THD, mainly because of the lower impedance.

Nothing more to it than a Z transformer and some capacitors?

 

[jraef]

They seem to imply that it is more than a zig-zag but when I saw the internal diagrams I recognized it. On the one I used I didn't even see caps!

BTW, rumor is that some of their (Mirus') engineers split off and formed a new competitor. I have their name somewhere, I'll post it when I find it.

"Venditori de oleum-vipera non vigere excordis populi"

 

[edison123]

Thx jraef. That was the 'end user' view I was looking for. btw, did your clients find the claims made by MTE/Mirus to be true ?

 

[jbartos]

Comment on skogsgurra (Electrical) Apr 26, 2004 marked ///\\
One example: [

http://www.ecsintl.com/pdf/Outperform K-rated.pdf

] This article compares performance of K-rated transformers vs the LINEATOR.
///I was trying to find LINEATOR cited in the link and I could not find it. However, the presented comparison of K-rated transformers with Harmonic Mitigating Transformer (HMT) is carefully documented therein. Obviously, the HMT is a different harmonic mitigating device from LINEATOR. The HMT is essentially modified transformer with special winding connections to trap harmonics. The LINEATOR is a passive harmonic filter with high input impedance to harmonics and high output impedance to harmonics. This means that it is the low pass passive filter fitted in line conductors of the nonlinear load.
Visit

http://www.mirusinternational.com/downloads/L-S001-C5 (Lineator Tech Sheet 60Hz).pdf

http://www.mirusinternational.com/downloads/UHF-PS01-C (Lineator Brochure).pdf

(see the advantages of LINEATOR)\\

 

[HarryDampers]

I have had some practical experience with the Mirus product and in practice have achieved THDI figures of 5,3 to 6% in an average Industrial installation. This is similar to that achieved by an 18pulse drive .

No, it is not just a Zig Zag transformer and for those who want an explanation of how it works plus some field test results , I will be happy to supply them or ask Mirus direct .

The main problem that these devices have had in the past is that to avoid excessive volt drop at the drive terminals , large amounts of Capacitance have been used.This then leads to high DC bus voltages and leading Power Factors at reduced loadings. The Mirus product (and possibly others) have solved this problem.

 

[edison123]

Hi HarryDampers,

Thx for your feedback. Could you possibly mail mail me (to the id in my profile) your article on how Mirus products work and also the e-mail id of the contact person at Mirus. I sent a mail to them last week thru' their website and recd no reply till date.

Thx again.

 

[ICEMAN]

OK. I have been sitting reading all the comments from so called experts re MTE Matrix and Mirus Lineator. The vast majority of the comment is absolute crap ! It's reminds me of "hausfraus" gossiping over the garden fence. Let's get real ! Engineering is about facts not gossip.

The MTE Matrix is a rough copy of the Mirus unit. It's about 3-4 times the size and inferior to a significant degree (in my opinion) to the Mirus unit. Also, read the small print re "cannot guarantee performance if the Vthd is more than 0%", etc or "if voltage imbalance is more than 1%"....

The Mirus units is not a zig zag transformer, which as some of the so called experts should know, is for mitigation of zero sequence harmonics. I don't see any of those in 5, 7, 11, 13...etc. {+ve and -ve, yes !} Triplens (3, 9, 15...) but not 6n +/-1 characteristic harmonics.

The Lineator is a patented multi-limbed reactor (which has a large number of mutual inductances)with a small capacitor bank. When connected to standard 6 pulse VFDs with diode or SCR pre-charge front bridge DOES reduce the Ithd from around 37-40% (based on either 3% AC line or DC bus reactance {5% source impedance} to 5-8 % GUARANTEED. The performance I have seen on various sites is often below 5% Ithd and on a par with active filters. Unlike the Matric filter this is irrespective of voltahe imbalance (which seriously effects 12 and 18 pulse drives) AND any pre-existing Vthd. On many sites (ask Toshiba, Houston) Lineator outforms active filters (I have no axe to grind here as we also use active units). However, for VFD applications Lineator is on a par, if not better than active, at under half the price and none of the hassle (like $ 1500 a day commissioning charges !)

How does it work....Well, the output voltage waveform from Lineator is trapezoidal which forces the input devices in the drive front end to conduct for a longer period with a lower peak hence reducing the Ithd to ~ 5%. Simple as that !

Also, Lineator can be used on single and multiple drives. I have personally been involved in projects for group mitigation up 1500HP/1100kW. Works great. A 'fit and forget product'.

Talking about efficiencies....A standard 6 pulse VFD/Lineator is generally 3-4% more efficient than 12 or 18 pulse drives. For energy efficient applications this is important. Active filters have around 4-5% losses dependent on the carrier frequency.

Drives with active front ends (sinusoidal rectifiers) are fine if you need regen braking but they have a large EMI footprint and need large AC line reactors and high frequency passive filters to attenuation the input bridge carrier frequency (2-3.5kHz)so it doesn't interfere with mains switching devices. They are fine if you can afford them - ABB ans Vacon reckon AFE drives are 2.2 x 2.5 the cost of standard 6 pulse drives.

Why don't you guys talk to the major drives companies who have taken the Mirus Lineator onboard over the last few years. I understand there are almost 3000 installations now throughout the world in powers to around 2000HP/1500kW. Speak to ABB, Siemens, Vacon, Cutler Hamer, Square D, Toshiba, Control Techniques, Invertek et al or the users worldwide.

I also hear in the grapevine that the new Lineator can maybe also work with SCR fully controlled front end, AC and possibily DC. I have no corrobation of that though. You'll have to ask Mirus yourselves. The details are on the net.

There also a few Lineator variants I believe -

i) One which can be connected to the primary of an existing 12 pulse drive - this reduces the Ithd from ~10-15% (dependent on whether double wound or polygonal autotransformers) to around 2-4% similar to 18-30 pulse drives (in plain English - "not quite as good as 36 pulse, but not far off". I understand the physical size of this units to around 30% of the kVA size of drive.

ii) For application where three phases mains is not available there is a variant (never tried one yet) which permits the connection of three phase VFDs without the usual derating and provided harmonic mitigation to around 8% Ithd.

iii) There is aslo I believe (again never used one yet) there is a TransLineator (a transformer with a integrated Lineator obviously for applications where voltage transformation is required and/or galvanic isolation {e.g. on ships with IT networks (insulated neutrals - standard drive EMC filters cannot be used here so the isolating transformers acts as the common mode noise filter for the drives).

Anyway, rather than gossip on the net like a bunch of "old sweetie wives" {Scottish expression} talk to some of the drives companies or users (people at the front end) about their experiences with Lineator OR talk to Mirus themselves. I have personally found the Mirus guys very approachable and up front. They won't spin you a line - afterall the Lineator is guaranteed to do what they say it does or you get your money back - that's deal.

Anybody want any clarification or additional info re Lineator, active filters, active front ends, etc. drop me an email.

Regards
ICEMAN

P.S. Now I have a question for you guys which you can happily discuss over the garden fence......What do you all know about DUPLEX REACTORS ?

 

[edison123]

Thx ICEMAN. I contacted Mirus more than 10 days ago thru' website e-mail and yet to receive any reply. Would appreciate, if you can mail me your info on mirus products and any available test results to the e-mail id listed in my profile.

 

[jbartos]

Comment: I agree with the ICEMAN LINEATOR assessment based on the literature I received from MIRUS (including test results).

 

[Skogsgurra]

ICEMAN,

It may be that the LINEATOR is well known and widely used in the US. The situation over here in Europe is that most manufacturers (I have talked to centrally located development departments in Erlangen, Pacy, Wales and Helsinki) and nowhere did they know about this wonder. These people are actively looking for something as good as the LINEATOR, but are using AFE or conventional filters in applications where low mains harmonics are needed.

The reason could very well be complete ignorance from their part, but I find it hard to believe that deveplopers have not heard about something as good as this seems to be. As long as the LINEATOR guys cannot (or will not) show how they do it, most engineers will not believe their claims.

I am as much against "garden fence gossip and crap" as you are. But as long as we cannot be given the facts, we can only exchange our thoughts and try to understand by discussing the matter here in our little "garden". I welcome a link to a site where the working principles are described. That would end the "gossip" effectively.

You are welcome to produce that link. Here.


And, sed2developer: What is your thinking here? I have understood that you are actively involved in Ithd mitigation and that you are closely connected to Erlangen.

 

[jbartos]

Comment:
AFE is better than LINEATOR since it permits reverse energy flow from the load to the supply (4-quadrant VFD).
I do not think that many engineers "believe", they go by engineering theory, analysis/calculations, data sheets, specifications, test results, reliability results, etc.

 

[ICEMAN]

Hi guys, sorry about the missing words and spelling yesterday. I tried to do it in a hurry over breakfast before taking to the hills.

The 'sweetie wife' comments were to solicit a response. It worked !

Re 'edison123' - if Mirus haven't replied to your email interest I would phone them. If you want to email me direct use mailevans@tiscali.co.uk but I'd go to Mirus in the first instance.

Re 'jbartos' - there is lot of test data from Lineator out there. I suggest you try Toshiba Houston, ABB US, Siemens Canada, Alstom UK, Vacon Finland, etc. Whether the drive companies will give you them is another matter. Some seem to not want to advert their harmonic performance without mitigation.

Now the biggie -'skogsgurra'. You sound like a 'scanda-hooligan' so you must work for ABB. I agree with you that most European seem to not to look at Lineator as an option. This is mainly due to ignorance but it is also due to the fact that they still hide behind the fallacy of 12 pulse drives. In the absence of a sensible and coherent EU harmonic standard which can be policed drive companies fall back on the 12 pulse. Read any spec and the answer to drive harmonics is 12 pulse drives - they don't say whether that should be double wound or polygonal and as you know there is a big difference in performance between the two. Also, compared to 12 pulse Lineator may look slightly more expensive. (The fact that it rarely now complies any national harmonics standard is often overlooked). That may so on the suface but once the real additional cost of 12 pulse drives (not just the transformer is taken into account) Lineator is more attractive. In the UK Lineator is now used by water companies, one of the most conservative users. I have seen applications where large installations of both Vacon and ABB drives are fed from common Lineators.

As for Erlangen - the Siemens drives team is large and the large I heard Siemens were wanting to push their active filter for drive applications (as were Group Schneider). As far as I am aware Siemens in Canada use Lineator all the time so why don't Siemens in Europe ? Maybe it's the 'not designed here' syndrome !

Wales - well having a Welsh name help here. The only two drives companies of any real size in Wales are Control Techniques (Emerson in the US) and Invertek Drives {CT spin off}...both have purchased Lineator and I was talking to a guy last week who bought a 600HP Lineator/CT drive package from CT some time last year. It is installed on a cable ship in the Arabian Gulf and is seemingly working brilliantly. So much so that he was looking for Mirus telephone so he could look at buying more.

Skogsgurra is right though. There is not enough information in Europe about Lineator. You have got to remember that the Europeans are 5-10 years behind North America in harmonics and power quality issues. It's a matter of education but the European market is much more fragimented by the US. They don't have the benefit of IEEE 519 (2001) to focus minds ! The UK has G5/4 (2001) which, after working with '519' is a wee bit more complex. So complex that the application guide has not yet been issued, 3 years 5 months after the recommendations were issued.

I think the problem re Mirus publishing 'how it works; is more to do with the fact that Lineator is a patented technology which has had its 'copiers' in past. One is named after a movie and the other is based in another scanda-hooligan country called Denmark. Of course, both parties will deny it. Actually, from what I know of it Lineator is a simple but very complex product in design and is very hard to copy correctly.

OK. I have dealt with Erlangen, Wales; Helsinki is ABB as Vacon are in Vaasa and a big user of Lineator. Well, I was informed last week that a 1200HP Lineator was on it's way to Ireland for use with ABB drives. I think you should talk to Alf-Kare Adnanes of ABB Marine - I believe they are waiting to test a Lineator very soon. If you do work for ABB you;ll find him in the ABB phone list. If you are not with ABB email me at the address given above and I'll tried and get the details to you.

Actually, talking about gardens...it would be really nice with summer upon us to all gather in garden with a few beers and 'blether' {Scottish word for informal discussion} on the subject of drives and harmonics.

OK. I've kinda of wandered here....Another difference between MTE and Mirus is regarding the application with generator derived supplies (either standby or constant). Lineator has aroud 15% of the kVA tied up in kVAr whereas Matrix is 30%. That makes a big difference as to whether it works successfully on generators.

A year or so ago I make up a chart listing the salient differences between Matrix and Lineator. Email me at the address above if you want a copy.

A final comment re 'jbartos' and AFE. I agree AFE are for regen drives. Wunderbar - what use is that on a centrifugal pump ? On a crane fine but for a single quadrant drive (87% of the drives market !)? As I said yesterday :-

i) The EMI footprint of AFE drives are 100% more than 6 pulse therefore additional cost in attentuating that.
ii) Reliability is reduced. I was on a project in the UAE some years ago. Water pumping application. The client bought 5 x 900HP drives soley on the basis that they were 'sexy' as far as harmonics were concerned. They all failed at least twice in the first two years. When an AFE drive fails the pump fails and you do not water supply. If an active filter (parallel device) fails (I know these well too) the drive carries on, albeit with harmonics (which on a square law load are approx. proporation to the cube of the load). If a Lineator fails (serial device){never heard of any failures yet mind you} it can easily be bypass }automatically or manually} - in both case system integrity is maintained and everybody is relatively happy.

ii) Cost is a big issue. As I said yesterday ABB and Vacon reckon that an AFE version of their 6 pulse drives are between 2.2.- 2.5 more than 6 pulse drives. That's slightly more on average than active filters applied holistically. I believe the ratio of a standard 6 pulse drive/Lineator is in the region of 1.5-1.7.

If 'sed2developer' or 'skogsgurra' want to email me I'll what more I can give on operation of Lineator but I have to agree with you this kind of information is very scare.

OK. That's me done my bit again BUT to date I have no replies re DUPLEX REACTORS and if the man is from ABB he will no that ABB recently used them on some 40MW cycloconverter main propulsion drives. Let's have some feedback here please.

ICEMAN

 

[jbartos]

Comment on the previous posting:
1. I have already received the MIRUS provided test results for LINEATOR as I indicated in my previous posting.
2. Presently, LINEATOR has some limit in HP, say between 1000HP and 1500HP or so. If there are VFDs for motors over this HP limit, they still have to use phase shifters, AFE or very good harmonic filtering on their input side.
3. AFE is sometimes used just for its low input harmonic content, not necessarily for 4-quadrant VFD applications only.
4. There are numerous applications where 4-quadrant drives are beneficial, e.g. elevators, electrical vehicles, any motor propelled load with high inertia being frequently reversed, etc.