It is currently Sun Feb 28, 2021 5:59 am



Post new topic Reply to topic Go to page 1, 2  Next
Author Message
ekstra   ara
 Post subject: My team just finished a new Arc Protection device
PostPosted: Fri Sep 24, 2010 1:40 pm 

Joined: Fri Sep 24, 2010 1:24 pm
Posts: 5
Location: Denmark
Hi All,

I was happy to discover this forum. There is not much information about arc flash protection equipment on the net, so it was really nice to find this site.

My team and I have just finished developing a new product for Arc Detection. It is a simple modular device which can optically detect arcs through photo sensors and/or fiber optics. The device is able to issue a trip signal to the breaker in less than 0.5 ms and it can also combine the optical detection with a three phase current measurement (done through 5A CTs). The current measurement will work on peak values and can trip in less than 0.8 ms.

The aim was to make the device easy to install and simple to use, so it can be comissioned with just a flat-tipped screwdriver. Despite being a simple "box" we have provided more advanced users with configuration through USB and Internet Explorer. The device can also do event-logging and Java script based data analysis.

We would really like to know what the experts on this forum thinks of our new baby :-)

http://www.selco.com/News/News%20archive/Arc%20Protection%20-%20D1000.aspx?pageoffset=

Best regards

JASE


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Fri Sep 24, 2010 4:33 pm 
Plasma Level

Joined: Mon Jun 02, 2008 11:58 am
Posts: 1103
Location: Charlotte, NC
You are 10 years too late, ABb has had this for years and many others have followed suit.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Sat Sep 25, 2010 1:20 am 

Joined: Fri Sep 24, 2010 1:24 pm
Posts: 5
Location: Denmark
Hi Zog,

Thank you for your comment. You are quite right, Arc Detection is not exactly a virgin market. Still I do beleive that we can offer some advantages with our new device. We have been on the market with our previous D0100 series for about 15 years, so we are not without experience :)

I know of the ABB REA and the TVOC-2. The REA seems to be a system (a collection of devices) where the user will have to install additional modules for additional sensors etc. Also, as far as I know see the REA does not provide logging capabilities and data analysis.

ABB also has the new TVOC-2 system. The TVOC-2 looks like a much more complete device, but I would expect it to be a bit costly. Do anyone now the approx. price of the TVOC-2?

Our aim is to provide a simple device that will have everything built-in and which is very simple to install. Also, I think we have added a number of features which are not provided by the competition - at least not in the price range around 1500 USD.

By the way... I think that our new D1000 is also the fastest device on the market, as we can trip on light in less than 0.5 milliseconds.

Anyway... I don't want to missuse this nice forum for advertisement. It was just to get som technical feedback from somone who knows and appriciates the arc flash business.

Best regards

JASE


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Sat Sep 25, 2010 5:20 am 
Plasma Level

Joined: Mon Jun 02, 2008 11:58 am
Posts: 1103
Location: Charlotte, NC
jseedorff wrote:
Hi Zog,

Thank you for your comment. You are quite right, Arc Detection is not exactly a virgin market. Still I do beleive that we can offer some advantages with our new device. We have been on the market with our previous D0100 series for about 15 years, so we are not without experience :)

I know of the ABB REA and the TVOC-2. The REA seems to be a system (a collection of devices) where the user will have to install additional modules for additional sensors etc. Also, as far as I know see the REA does not provide logging capabilities and data analysis.

ABB also has the new TVOC-2 system. The TVOC-2 looks like a much more complete device, but I would expect it to be a bit costly. Do anyone now the approx. price of the TVOC-2?


It says ABB on the side, so that means it works well and is expensive :)

jseedorff wrote:
Our aim is to provide a simple device that will have everything built-in and which is very simple to install. Also, I think we have added a number of features which are not provided by the competition - at least not in the price range around 1500 USD.
Have a link to more technical details?

jseedorff wrote:
By the way... I think that our new D1000 is also the fastest device on the market, as we can trip on light in less than 0.5 milliseconds.


Like I told the ABB guys, your relay can be faster than light, but you still are bound by the opening time of the breaker.

jseedorff wrote:
Anyway... I don't want to missuse this nice forum for advertisement. It was just to get som technical feedback from somone who knows and appriciates the arc flash business.

Best regards

JASE


Yep. sort of toeing the line here but keeping the sales pitches aside I think a technical discussion would be allowed.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Sun Sep 26, 2010 2:02 am 

Joined: Fri Sep 24, 2010 1:24 pm
Posts: 5
Location: Denmark
Hi,

Here are some technical details.

The device is based on microprocessor circuitry, which provides the advantage of various programmable options, PC based configuration, time-stamped event logging etc. The microprocessor will sample all six sensors every 125 microseconds, but the IGBT based trip circuit will add another 200 microseconds before trip. Thus, the fastest possible trip time (not counting the external breaker) is 325 microseconds. However, this is based on one sample only. The default trip time is less than 600 microseconds, which is equal to three consecutive positive samples. The trip time can be adjusted in the PC based configuration from 325 microseconds to 2 seconds.

The microprocessor circuit is fully reliable but in the unlikely event that it (or its software) should fail, we have provided a redundant solid-state trip path. The solid-state trip path, which has onlt OP-AMPS and discrete components, provides instant trip. The solid-state trip path is only there as a fail-over safety and provides no programmable features.

Trip happens only if the light on any of the six sensors stays above the adjustable level for the complete period of the trip time. The trip level can be easily set on a dial located on the front facia of the unit (10.000 to 25.000 lux). The front facia also has its own light sensor for easy testing. We can provide point sensors as well as fiber optical sensors. The point sensor has a range of about 3 meters (half a orb viewing angle), while the fiber sensors can "see" 2 meters around its length. The length of the fiberoptical sensor is 8 meters (+ wire based transmission leads for the fiber transmitter and receiver). The transmission cable to the sensors can be up to 100 meters. Any combination or six sensors can be used.

The sensors are (shielded) wire based and carries supply for a small amplifier located inside the sensor. The amplifier ensures a proper signal to noise ratio so that we get no issues with EMC. The sensors also have a built-in health check circuit that ensures proper function all the way trough the loop. A small red LED fashes inside the sensors to check the reception. The LED will also provide remote visual indication of operation, trip and sensor health. The remote LED indication is a nice feature when comissioning the system.

The unit also provides a built-in 3-phased overcurrent / short-circuit function. The current trip works with external 5A CT's and it can be configured with two levels and delays to cover both "slow" overcurrent and "fast" short-circuit.

The output is based on a fast reacting IGBT circuit with high drive capability. The unit will monitor both the state of IGBT "contact" and the coil of the external relay. The trip circuit can operate with external supply in a wide range (DC or AC).

The unit can be supplied by DC or AC and it can even charge a 24 VDC lead battery (for backup).

Various discrete I/O's are provided for trip reset, service mode, trip output etc. There are also push buttons on the front to cover these functions.

Configuration is done trhough USB. The device will work as a USB "mass storage device", so no drivers are required. Internet Explorer is used for configuration, which means that no special programming tool is required. The device will also provide event logging with zoomable graphs showing trip situations, intensity over time etc. RS485 MODBUS-RTU is available for SCADA or remote HMI.

Best regards

JASE


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Mon Sep 27, 2010 5:15 pm 
Arc Level

Joined: Tue Jan 13, 2009 5:00 pm
Posts: 542
0.8 ms to detect overcurrent? 1/21 of a cycle at 60 Hz? Suppose your arcing fault occurs at a zero crossing. How can your OC relay differentiate between the faulted and unfaulted condition just 1/21 of a cycle later?

On the optic side, illumination from the arc will also be sinusoidal. You are suggesting detection times for events that will occur in nearly complete darkness if they happen to occur near the crossing.

And as Zog already suggested; this speed increase is happening at the wrong end of the trip circuit. A millisecond here or there makes little difference when breakers trip times are in the order of several cycles.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Tue Sep 28, 2010 7:48 am 

Joined: Tue Sep 28, 2010 4:21 am
Posts: 7
But when do you have an arc fault?

Disclaimer: I am also from the team jseedorff mentioned. Please tell us to can it if we get too carried away, but it is great to find a group of people who obviously know a lot about a subject which we often need to convince people even exists.

Stevenal and Zog are right, of course - the speed of the detecting equipment is normally negligible compared to the breaker which actually clears the fault. For better safety, we need a breaker which is an order of magnitude faster than the current breed. Perhaps solid state electronics will get there some day - then we will not need fancy pyrotechnical bolting shorts or plasma chambers as add-ons, but will just interrupt the current properly at the breaker if a fault happens.

Nevertheless, something needs to detect the arc fault before any of the devices can be operated.

It is impossible (or at least very risky in terms of downtime) to react before the arc fault is sufficiently large to distinguish it from normal operation. Thus, we measure the reaction time not from the inception of the arc, but from the moment it is powerful enough to be detected. But after this, why add more burn time to the arc fault than needed?

To Stevenal: You are right, we can't react the moment a wrench hits the busbar at the zero crossing - but then again, the fault has not yet appeared. No current is running, and no energy is released (yet). However, we do actually detect overcurrent in less than a millisecond, and we can exploit that the instantaneous power, current and light emmision in the arc fault are all proportional. Thus, we hold that it is valid to look at the instantaneous light and current and react if these both exceeds some limits, which will determine the minimum size of the arc which the system will detect. A short somewhere downstream will not produce light, and a photo flash will not produce current - thus, we will not trip the entire system unless it is actually necessary. Normal OC relays follow BS 142 / IEC 255 inverse time characteristics, which means that bolting the company standard OC circuit on an arc fault relay will make it much slower - in the same league as the breakers. We do not even try - we just look at the instantaneous value and react if it is too high.

The consequence of this decision is that the portion of the first arc cycle which is below the threshold is pure delay when looking at the complete arc fault event. On the other hand, if the current and thus the light and the power has not yet grown above the threshold, neither has the integrated energy - and an algorithm trying to sound the alarm would run a huge risk of tripping when it was not necessary. It is a question of balancing this possible delay against the need to avoid nuisance trips, which is the reason for the OC part of the device

I guess that you just pressed one of my buttons ;-)

/Niels


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Tue Sep 28, 2010 8:23 am 
Arc Level

Joined: Tue Jan 13, 2009 5:00 pm
Posts: 542
Thanks Niels,

I agree with your statement about the fault not yet appearing if the wrench hits near the zero crossing. However, the testing models we use (IEEE 1584 for example) start timing at fault initiation wherever it hits the waveform. Claims of 0.8 ms gives us the wrong values to enter into equations or software. We need manufacturers to give us detection times from fault initiation, or we need a different model.

The problem of course does not just occur around the zero crossing. Since the sources we have are inductive in nature, current rise will be gradual even if the wrench hits at the peak of the waveform.

Also note some of the circular reasoning involved here. If detection times are measured from the instant that faults are detectable, we ought to be able to cut those detection times to zero.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Tue Sep 28, 2010 10:45 am 
Arc Level

Joined: Mon Jan 18, 2010 11:35 am
Posts: 525
Location: Wisconsin
jseedorff wrote:
The trip level can be easily set on a dial located on the front facia of the unit (10.000 to 25.000 lux).


An adjustable trip level based on lux, how do we determine the desired level? If we are using the current sensing feature, does the lux setting matter?
If intensity is a factor, then is locating the sensor(s) properly also a concern?


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Tue Sep 28, 2010 11:53 am 
Plasma Level

Joined: Mon Jun 02, 2008 11:58 am
Posts: 1103
Location: Charlotte, NC
niho.selco wrote:
Stevenal and Zog are right, of course - the speed of the detecting equipment is normally negligible compared to the breaker which actually clears the fault. For better safety, we need a breaker which is an order of magnitude faster than the current breed.


I am working on that :)


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Wed Sep 29, 2010 3:01 am 

Joined: Tue Sep 28, 2010 4:21 am
Posts: 7
Stevenal: Okay, I understand the issue, but it is not easily solved. Taking everything into account, we are looking at up to half a cycle between the initiation event and the first energy release, which is necessary for a reliable detection - and this "up to" is completely stochastic. However, as far as I can see, the NFPA at least specifies arc burn time in the calculations, which must be the time from first to last energy release? (Which would be from detection to breaker tripped, as the NFPA assumes constant arc size equal to the worst case current).

The numbers Jakob sent are worst case numbers for the delay from energy release to current in the trip coil. Adding the delay from initiation to energy release would make the marketing people unhappy - but I will try to work the discussion into the documentation. It seems that adding a section on IEEE1584 or NFPA 70E, specifying that the arc burn time must be calculated as the sum of every delay, especially the breaker, would also be a good idea.

As for the reaction time of zero, our best case is just above 200 us for both light and current, but it would be completely meaningless to specify this in the normal documentation - or even just whispering it to marketing ;-)

Zog: That sounds really exciting. I could not find any details in your (recent) earlier posts - are you able to share more detail?

JBD: Yes, that is right, the settings need to reflect the placement of the sensors. 5 kA arc current at 3 meters direct view distance corresponds to about 12 klux. Larger arc faults give more light. The light intensity falls with the distance squared, but in cabinets, the entire interior is pretty well illuminated due to internal reflections. You will have to spend a sensor per compartment, though. There is some time dependency due to occlusion by smoke and debris if the arc keeps burning, but that is not a problem in the detection phase - and (knock on wood) should not be a problem in installations.

The current sensing is primarily for reducing nuisance trips due to photo flashes, welding or the like. It inhibits trips on light alone, if so desired. We recommend setting it to slightly above normal maximum load. However, it can also be used for near-instant short circuit trips.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Fri Oct 01, 2010 5:08 pm 
Arc Level

Joined: Tue Jan 13, 2009 5:00 pm
Posts: 542
niho.selco wrote:
However, as far as I can see, the NFPA at least specifies arc burn time in the calculations, which must be the time from first to last energy release? (Which would be from detection to breaker tripped, as the NFPA assumes constant arc size equal to the worst case current).


NFPA calculations are from IEEE. The IEEE equations were developed from arc fault testing. Exposure time was measured from fault initiation to fault clearing. Those who do this type of testing are welcome to jump in here.

I'm sure the displeasure of those in the marketing department will be offset by those in legal.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Mon Oct 04, 2010 2:12 am 

Joined: Tue Sep 28, 2010 4:21 am
Posts: 7
I am not sure that I understand the need for the fault initiation time, if we define it as potentially different from the first energy release. But first of all, I am unsure if we actually disagree on what is appropriate, or just on what the wording of the standard means? I must admit that I have never performed an arc flash study, so please correct me on this!

The IEEE calculations find the worst case arc incident power from the available voltage and current using a formula derived from statistical analysis of a number of tests made with integrating calorimeters. The standard incident power is then transformed to incident power at a given working distance using the inverse square law. Last, the power is multiplied by the clearing time to give the energy delivered. That is, the energy delivered to the skin (or hopefully, PPE) of the unfortunate soul in front of the equipment is approximated by the bounding rectangle, assuming that the arc will burn at maximum power from beginning to end. This is a worst case assumption, neglecting ramp-up and variations.

In my opinion, measuring burn time from energy release to clearing is correct according to both the calculations and to the physics of the situation - the wrench dropped on the busbar does not cause any injuries during the zero crossing. Thus, the time that should be plugged into the equation is detection time plus clearing time, provided that the detector is sensitive enough to catch the first energy release.

I have difficulty understanding why the standard would have people guessing on the initiation time? The arcs we have used for testing have been initiated using a piece of thin copper wire between the electrodes, and in this case initiation time and first energy release are the same. My guess would be that the IEEE tests have been conducted in approximately the same way - you cannot wait for the cat to run into the switchboard every time ;-)


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Mon Oct 04, 2010 4:10 pm 
Arc Level

Joined: Tue Jan 13, 2009 5:00 pm
Posts: 542
niho.selco wrote:
The arcs we have used for testing have been initiated using a piece of thin copper wire between the electrodes, and in this case initiation time and first energy release are the same.


No they are not the same. You have described the setup, but not the initiating event. Time will be measured from upstream contact closure. Since I have never heard of anyone monitoring the waveform to hit a particular point (Why would they if they are trying to emulate a real world random event?), the closure will occur randomly on the waveform. From that point the current ramps up, since current cannot change instantly in an inductor (or in a non-zero X/R source). These experiments were used to derive the IEEE linear equations. These equations will happily take your microsecond inputs to give low IE values, but these values are not realistic for the real world random events we are dealing with.

"Time of energy release" I find to be a pretty squishy concept. How do you know when it occurs? When you can measure it? If so, all you can do is compare one instrument to another, since there is no absolute to measure from. Contact closure is a lot firmer concept.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Tue Oct 05, 2010 8:56 am 
Sparks Level

Joined: Wed Jun 02, 2010 12:59 pm
Posts: 70
Location: Milwaukee WI
Selco

I work for an OEM that sends machinery to all the populated continents. Pretty sure Australian customers are using SELCO D0500 and associated components.

When I looked at the SELCO literature and ratings, I did not see any UL or CSA approvals. That limits the use of the product in North America.

Apparently the D0500 works well in Australia, CE ratings are acceptable there.

If UL and CSA approvals are in the works, I would like to know status.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Fri Oct 08, 2010 5:20 am 

Joined: Tue Sep 28, 2010 4:21 am
Posts: 7
First of all, we of course agree that using the time from detection to output activation for the burn time is wrong and dangerous, and I hope that noone will do so. The clearing time is already explained in our manual as the sum of detection and breaker operating time, but I will make sure that it is stressed sufficiently.

Thank you for explaining the IEEE definition of initiation - I think I see your point now: With that methodology, they will have measured the closing delay of the breaker and subtracted this fixed offset (of about 80 ms) from the coil activation time, of course, but the timing is still random relative to the waveform, and this lowers the average energy measured. Thus, the average power in their arc power equation is lower than the "real" power of the arc when it is actually burning, as they integrate over the entire time, and as some proportion of this time, the arc is not burning. The protection class may thus potentially be found too low when using timing based on energy output instead, as the average power is higher in a burning arc than found using IEEE1584. Is this correctly understood?

However, the calorimetric error using energy detection time instead of switch closing time must have an upper bound of ((one half period / IEEE integration time) * IEEE arc power), as far as I can see? One half period because this is the worst case current lag. It may be that there is a ramp up period, too, but I believe that the energy output should be quite limited until it is detected. More on that below.

Now, I could have written a long rant about how the IEEE guys should have done their work in order to support my view, but I will spare you - and spare myself the embarassment ;-) - The IEEE1584 is the standard methodology, and we will of course use that. Instead, I will write a long rant on arc sizes and detection times.


We react within one millisecond after the sensor is exposed to 10 klux. Let us for simplicity put the breaker clearing time at 99 ms - a quite slow switch. Thus, the time from the sensor is hit by 10 klux till the fault is cleared by the system, is 100 ms. 1)

Now, let us find the arc power which will just delivers 10 klux of visible light to the sensor. Let us put the sensor distance at 1 meter - a good mean for a 60 cm standard rack. The work distance is the standard 18 inches (0.46 m). The user is thus, from the inverse square law, exposed to almost five times the power per area compared to what the sensor sees - and his skin is heated by the entire spectrum, while the sensor only reacts to some wavelengths.

Now for the trickiest part: what is the radiated output of an arc, and how is the radiated energy distributed on the spectrum? For further discussion of that, see 2). For now, let us start at the luminous efficacy (portion of radiation which is radiated as visible light) for a standard arc lamp, of 30-50 lm/W. Please note that this is not the efficiency (lm/electrical W) but the portion of the emitted radiated energy which is visible (lm/radiated W), that is, the portion of the energy hitting the calorimeter which is visible to an optical sensor in the same position. We do not need to know the electrical arc power - we just convert backwards from light levels to IEEE1584 calorimeter numbers for the total radiated energy and calculate what total radiated power a user at 18 inches is exposed to at an arc delivering 10 klux visible light at 1 meter:

10 klux @ 1 meter = 47259 lux @ 46 cm = 47259 lm/m2 // geometry
47259 lm/m2 @ 30 lm/W = 1575 W/m2 // arc lamp luminous efficacy
1575 W/m2 = 158 mW/cm2 = 38 mcal/cm2/s = 0.004 cal/cm2 in 100 ms.

To counter the squishiness, we extend the exposure time at full power by the worst case rise time (one half period, or 10 ms at 50 Hz). This will more than compensate for the energy delivered while the arc is burning at less than full power and thus below the detection level for the sensor. Furthermore, let us apply a safety factor of 10 to the luminous efficacy - after all, the arc fault has not been optimized for optical output. This yields:

47259 lm/m2 @ 3 lm/W = 15753 W/m2 // arc worst case luminous efficacy
15753 W/m2 = 1575 mW/cm2 = 376 mcal/cm2/s = 0.04 cal/cm2 in 110 ms.

What is the worst arc that the optical system will not limit the burn time for? Worst case is an arc which delivers 9999 lux, extending the burn time to the slow fuse clear time (Let's say 2 seconds), . The arc now delivers 20 times the above energy to the user, as the burn time is 20 times longer, exposing him to 0.75 cal/cm2. This is of course bad, but it is still below the threshold for second degree burns according to NFPA 70E - still in category 0.

The other bad part is of course the really big arcs, which will still burn for 100 ms (or 110 ms calculated from the initiation event). For these, prevention, distance, PPE and/or much faster switches are needed to avoid harm to the user.


The above calculations show that optical arc detection equipment, correctly installed and used, can limit the burn time of arc faults of all dangerous sizes to approximately the opening time of the breaker or the reaction time of the arc quenching device.

Gee, I guess that all here already knew that... But now I have also calculated that arcs which are too small to be detected are classed in Category 0, even after 2 seconds of burn time. The maximum ramp-up energy before detection is therefore quite small, and can be calculated or ignored.

Phew - that got out of hand. Please do not hesitate to correct me on this - I am not an NFPA 70E-expert.



Appendix to this monster post ;-) :

1) As an aside, in my test setup, the energy source is a high voltage capacitor initiated by a spark gap, so I have used very short-lived arcs (< 1 ms) with pretty high peaks (> 20 kA), but low total energy - category 0, certainly. Also ordinary welding gear have been used. I do unfortunately not have regular access to powerful switchgear which I can abuse at will, even though that is also a part of our testing.

2) There are two difficulties in this calculation. The first is the total radiated output of the arc. This is pretty much covered by the IEEE measurements. The other is the spectral emmision of the arc combined with the spectral sensitivity of the sensor. The arc itself is not a black body radiator, and consists of peaks corresponded to exitation of the materials involved - but overlaid with black body radiation from the heated surroundings.

We have done some spectroscopy on arc light to find out how to design our sensors. The details are confidential, but the bottom line is that the power spectrum is pretty close to the visible spectrum. It extends into both IR and UV, but luckily, copper-copper and copper-iron, which are most likely materials, both peak prominently in the visible spectrum around 500 nm, corresponding to bluish white light. Our equipment was not able to measure further than 1200 nm, where we would likely find another peak in the um range from heated secondary gasses - the surroundings. However, in the beginning of the arc lifecycle, the surroundings are not yet heated, and the majority of the radiation is from excited ions.

The spectrum must lastly be folded with the quantum efficiency of the used sensor to find the real sensitivity to arc light output. This corresponds roughly to the luminosity function for the human eye. The calculations here are thus not completely accurate, but they are also not that far off - and are certainly within the safety factor.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Fri Oct 08, 2010 5:25 am 

Joined: Tue Sep 28, 2010 4:21 am
Posts: 7
AB: You are right, we are selling a fair number of our older systems in Australia. And yes, we are pursuing both UL and CSA for the new system, but it takes months to get them. You can PM me if you want more details - this is getting a bit off topic, I think?


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Wed Sep 07, 2011 1:49 pm 

Joined: Fri Sep 24, 2010 1:24 pm
Posts: 5
Location: Denmark
Just a follow-up

The SELCO D1000 has now become the Littelfuse PGR-8800.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Thu Sep 08, 2011 7:48 am 
Sparks Level

Joined: Wed Jun 02, 2010 12:59 pm
Posts: 70
Location: Milwaukee WI
Press release Littelfuse & Selco

Littelfuse acquisition of Selco.

http://www.littelfuse.com/company/press-release/Littelfuse-Announces-Acquisition-of-Selco.html

I would still like to know status of US/Canada approvals. UL submission numbers are a good and useful start.


Top
 Profile Send private message  
Reply with quote  
 Post subject:
PostPosted: Thu Sep 08, 2011 8:03 am 

Joined: Tue Sep 28, 2010 4:21 am
Posts: 7
We are still working on the cULus approval, unfortunately. It seems to take some time to get from overseas. The current estimate from UL is early november 2011. I will post a UL submission number when I can get hold of it. Meanwhile, you can contact any sales office at Littelfuse/Selco directly for more information - and we can continue the juicy technical discussion on this board :-)


Top
 Profile Send private message  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 28 posts ]  Go to page 1, 2  Next

All times are UTC - 7 hours


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Jump to:  
cron
© 2019 Arcflash Forum / Brainfiller, Inc. | P.O. Box 12024 | Scottsdale, AZ 85267 USA | 800-874-8883