Jump to content


Photo

Soft Starter Ratings. - How are they derived?


  • Please log in to reply
18 replies to this topic

#1 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 18 November 2003 - 06:36 PM

Hi

I am interested in finding out how designers determine the ratings they put on to their soft starters. There are many different ideas and techniques, but in the real commercial world out there, it is important to both retain reliability, and to retain commercial competitiveness.

I will post my ideas and experience shortly.

Best regards,

#2 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 25 November 2003 - 09:50 AM

In private discussions with others, I have found many theories and practices on how to rate the starter based on the operating and starting characteristics of the motor and load.
If the rating of the starter is too high, then the SCRs will become overheated and fail early. If the ratings of the starter are too low, the starter will be too expensive and will not sell. As a customer, how do we determine whether the ratings are appropriate?

The miniumum starting current that will start a motor and load is a function of both the motor characteristics and the load characteristics. In any event, the start current will be several times the rated current of the motor. Rarely, the start current will be much below 250% of the rated motor currentand can be over 600% for some motor/load combinations (although at this current you could consider DOL starting!)
The starting time is a function of the motor and load characteristics, and in particular, the inertia of the load is important.

During run, there will be a current passed through the SCR pairs which will result in power being dissipated. This power must be got rid of using some form of heatsink.
During start, there is a high overload current flowing for a relatively short period of time.

At one end of the design strategy, the starter rating is based on the maximum continuous current that the SCR and heatsink can operate at. The overload during start overloads the SCRs, and because it is only for a short time, the SCR will not fail immediately. To determine start current/time ratings, some designers select a semiconductor fuse to protect the SCR (based on I2t ratings) and then use the fuse overload curves to rate the starter.

At the other end of the design strategy, designers take the maximum start current and select SCRs and Heatsinks that will allow this current to flow continuously.

There are other strategies that lie between these two extremes.
Any thoughts?
Best regards,

#3 BigMax

BigMax

    Intermediate Member

  • Full Member
  • PipPipPip
  • 43 posts
  • Location:Melbourne, Australia

Posted 25 November 2003 - 11:39 PM

Hi Marke,

Here's my suggestion.

You can attempt to interpret the commonly used AC53a/b ratings for non-bypassed/bypassed conditions and extrapolate those ratings according to the motor FLC and load type you are attempting to start.........in my opinion this method is nonsense and the correct starter selection is sometimes not obvious.

It seems to me that an obvious and simple alternative would be for softstarter manufacturuers to select SCRs and heatsinking arangements that match (or exceed by, say, 10%?) a standard class 10 thermal overload trip curve and state the maximum FLC rating at which this trip curve applies. In this way, a closed loop softstarter could 'protect' both the motor and itself from overload.

For heavy duty applications, class20 trip curve ratings could also be given.

So, all that would be required is to consider the motor FLC, and load type (usually standard class10) and choose a starter to suit.

What do you think?

Cheers!

BigMax

#4 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 26 November 2003 - 08:49 AM

Hello BigMax

Long time no see!!
I think that that proposal is an excellent idea if it can be implemented.
Any ideas as to how this can be achieved??

The big problem would be to get a time constant that was appropriate. Perhaps one of our Power Electronics experts can make some suggestions.

Keep the thinking cap on,
Best regards,

#5 BigMax

BigMax

    Intermediate Member

  • Full Member
  • PipPipPip
  • 43 posts
  • Location:Melbourne, Australia

Posted 07 December 2003 - 11:34 PM

Hi all,

Having spent some more time thinking further about soft starter 'ratings', the more I'm convinced that the answer to the question posed by Marke should be simple, although I'm approaching from a slightly different direction I suspect.

Which other (correctly sized) motor starter (ie star/delta, autotransformer, etc) have you had experience with which may fail if, for example, the load jams (ie conveyor) and a start is attempted? Or, occasional overload occurs during normal operation (ie crusher). Of course, under these circumstances the thermal overload should simply trip, preventing permanant damage to anything.

Why should a softstarter be any different? Why should we accept that such conditions may exceed the 'ratings' of the SCRs and probably not be covered under warranty? In my opinion, born out of occasional negotiations involving SCR failure and silly softstarter selection methods, the weakest link in any starter / motor chain should be the motor, not the starter.

If a soft starter is selected to match motor FLC and a thermal trip curve, it should not fail first under general overload conditions. The first manufacturer to market such a softstarter will have my approval ;c;

Cheers!

Bigmax

#6 jraef

jraef

    Posting Freak

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

Posted 09 December 2003 - 07:41 PM

BigMax,

They are out there, but this being a non-commercial group I will refrain from proseletizing (much) since I work for one of them. :P

An interesting side note however is that we have been losing sales on a regualr basis to the competitors who are extremely marginal on their ratings, to the point where we must knuckle-under or perish. All of our new development projects are based upon safely reducing the capacity of our design while attempting to be competitive to the el-cheapos. Even in our latest iteration we are still rating them to do the job IF PROPERLY SELECTED, but that is a departure from our previous philosphy of designing for the worst case. In fact we are designing for capacity just as you suggest, by using OL trip curves and NEMA MG-1 motor utilization criteria for duty cycling.

Just to give you an idea as to ratings philosophies, I did an analysis of overload capacity of our current product vs most of the major competitors here in the US. As a basis for comparison I used heat capacity by converting the overload energy and time to BTUs, since ultimately overload capacity is related to heat buildup at the SCR substrate and the heat sink's ability to move it away (I know this is oversimplified but stick with me here). Since ours provides the highest overload capacity on the market, 500% for 60 seconds at 50 C, I used it as the 100% mark. The next nearest competitor was rated for 500% for 30 seconds at 40 C, so when adjusted for temperature it has only 37% of our capacity. This is by no means an inadequate rating, but it gets worse. The current price and size leader in the market came in at having only 8% of the thermal capacity of our design, yet we are losing sales to them every day. That is not a typo, they have 8% of our overload capacity! The main reason why they are so small and cheap is that they have NO heat sink! When we point this out to the buyers, they say "Oh well, drop your price and I'll give you the order."

Price is, unfortunately, the primary purchase factor in too many instances, and the age-old concept of "you get what you pay for" is being largely ignored. I only wish there were more customers out there who think as you do. ;b;
"He's not dead, he's just pinin' for the fjords!"

#7 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 12 December 2003 - 08:57 PM

One of the big things for the interested buyer, is exactly how are the ratings derived.
If a manufacturer claims 500% current for 30 seconds, what does that mean and how is it derived?
At this stage, I am not aware of a "standard" that specifies how thes ratings will be achieved or calculated. We have the AC53a and AC53b ratings, but they are just a means of stating what the manufacturer (or sales department) believe the starter is good for, (or what they can get away with).

Too often, on reverse engineering this type of equipment, I have found that the rated current of the starter, is in reality very close to the maximum continuous current of the SCR heatsink combination (AC1 equivilent) and the starter ratings are based on the overload current curve of a semiconductor fuse that would protect the SCRs under a short circuit condition.
These ratings are usually quite easy to identify from the start current/time ratings given by the manufacturer. Typically, a starter rated to fuse curves will be rated at 600% for 3 seconds, 300% for 20 seconds and 200% for say 40 seconds. Essentially, there is a rapid decay in the starting current rating with time. (Semiconductor fuse curve)
From my experience, starters rated this way do not last in the field unless they are installed with a reasonable safety margin in rating.
jraef, I totally agree with you, and unfortuately, the commercial reality is that companies do not win on having the best technology, or the ideal ratings. Marketing plays a major factor as does price.
In this field, we are seeing new players come onto the market and all are facing a price down drive by the consumer. I believe that the consumer needs to know how the ratings have been derived to help them differentiate between products that are claimed to do the distance and those that will.
jraef, if you wish to describe how your company derives it's ratings, naming the company as well, that is OK by me. I consider that good information, not blatant advertising. Non commercial means no selling, information is fine and valuable.
Best regards,

#8 jraef

jraef

    Posting Freak

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

Posted 02 February 2004 - 08:13 AM

OK, here goes Marke.

First off, I agree with your assessment of ratings often being associated with fuse ratings. I work for Motortronics and we have never done that. I came here from another manufacturer who did, and I was excited to see how conservative our design engineers were. Their approach was to first do a heat run calculation based upon a 50degC ambient, with full rated current and no fans! (their philosophy was that the fans should only be used to dissipate for duty cycle). Then they took that design, added fans, and took it to 500% of the rating. If after 60 seconds the SCR junction temperature was maintained at 20% below threshold, they shut down and waited an amount of time dependant upon the duty cycle, then repeated the tests over a 1 hour period. The duty cycle was taken from the NEMA MG-1 design specs for the motor it weas intended to be used on, i.e. 10 starts per hour on smaller frames, 6 per hour on larger, 4 per hour on the 1250A frame etc. When it passed without component failure, they called it a day and released it to production (highly oversimplified of course).

Incidentally, our short circuit tests are done without fuses! We are, according to UL, the only RVSS manufacturer who has passed the "3 cycle test" at UL with solid state starters. This means they can be used behind any brand of circuit breaker at the rated interrupt rating of the breaker, without additional combination testing. The only other RVSS's that come close are tested with current limiting breakers, and so can only be used with those specific units or the same brand.

What does all this get us? Not much in the marketplace. When I get a chance to bring up these issues to a decent engineer it will get us the nod IF our price is equal, but more and more I am finding what I call "checkbook engineering" where design specs are ignored for the bottom line. I am today fighting for a project where the consultant wrote my design specification for 2ea 460V 700HP starters. Unfortunately Cutler-Hammer was 2/3 my price. When I pointed out that they do not come even CLOSE to my specs, the contractor petitioned the municipality and questioned the engineer's judgement, accusing him of unnecessarilly driving up the price. I even pointed out that the CH unit was rated at 850A, and the motors are 820A with a 1.15 service factor. In addition, if the user changes the current limit setting to 450%, the SAME UNIT becomes rated for only 650A! So far all this is falling on deaf ears because "a big company like Eaton would not make a product that couldn't do the job".

Sorry for the ramble, but I often feel like a lone voice in the wilderness. I enjoy fair competition from other quality manufacturers, and there are still quite a few out there, but we are all being forced into changes by a marketplace that I feel will ultimately suffer the loss in the long run. Hopefully just not until I retire!
"He's not dead, he's just pinin' for the fjords!"

#9 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 02 February 2004 - 06:31 PM

Hello jraef

Yes, that illustrates my point exactly!!

Here we have an example of a good belts and braces approach to engineering a solution that wil be very reliable, complies with all the standards etc and will last a very long time, but in terms of how the starter is rated, how is it differentiated to the customer??

We have the 947-4-2 spec suggesting how the starters should be rated, but there is no standard covering how these ratings are actually derived. So, under the 947 spec, the manufacturer can put down the figures that in his/her opinion are reasonable for that starter but with no substantiation or verification.

If we take the case where the starter ratings are based on the semiconductor fuse curves, then from the curves we can derive the start time and duty and if we try it out in the lab, it will work. The problem with this rating, is that at every start, the SCR is well overloaded and is operated in the region generally refered to as the "transient overload region".
Regular operation in this part of the curve is not recommended by the SCR manufacturers, and one of them suggests that operation in this region should be "limited to no more than eight times in the life of the device". In other words, everytime you operate into this part of the curve, you are degrading the life of the SCR.

From my experience, starters with fuse curve derived ratings have a short life unless they are used well below their ratings. Of course, if you have discrete steps in your model ratings and you always select the unit rated above the motor rating, you will have an inherent safety margin that provides a degree of derating. i.e. model 1 rated at 25 Amps, model 2 rated at 35 amps. For a 28 Amp motor, use model 2 and you gain a 7 amp buffer, and the life of the starter will be extended. For this reason, some manufacturers appear to get away with it except where there are difficult starting conditions, or a 35 A unit is used on a 35A motor!

So here we have a major dilema. Just what do the ratings applied by the manufacturer really mean? How are they derived?
In the cold hard commercial world, reallity is that price is important as jraef has shown. Over engineering the product will cost sales, under engineering the product will gain sales but will cost reliability and reputation. Unfortunately, a bad experience with a soft starter tends to taint soft starters rather than a particular brand. The market still needs to mature.
Where is the happy median? and How do we make it easy for the customer to ensure that what he gets is correct for the application?

Best regards,

#10 Guest__*

Guest__*
  • Guests

Posted 08 February 2004 - 07:20 PM

Interesting

My opinion is that the ratings and protection are different issues (although obviously they interact). A soft starter should be rated for its end use, at a realistic temperature. IEC standards set out some rating that can be used.

The protection should be that under stressed conditions(overtemperature, over current, excess starts) the soft starter acts sensibly and that does not include melt down. One approach is to monitor the soft starter and shut it down under over temperature - another is to over build the unit and not internally monitor. Both are valid approaches.

#11 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 09 February 2004 - 06:22 PM

Hello Anonymous

QUOTE

A soft starter should be rated for its end use, at a realistic temperature. IEC standards set out some rating that can be used.  

Good point. But how do we derive the ratings? based on a series of lab test where we puch it till it breaks and back off a little?


QUOTE

My opinion is that the ratings and protection are different issues  
Agreed.
QUOTE

The protection should be that under stressed conditions(overtemperature, over current, excess starts) the soft starter acts sensibly and that does not include melt down.
Agreed this is the ideal.
QUOTE

One approach is to monitor the soft starter and shut it down under over temperature  
This is more complex than it seems. Ideally, one should design to restrict the junction temperature of the device to allowable limits, but the junction temperature is a function of the case temperature, the power dissipated and the thermal resistance between the junction and the case.
The power dissipated is dependant on the current flow and so if the upper temperature limit is fixed, the maximum safe case temperature is current dependant. This can be done provided that there is a compensation system that monitors the actual current flow and converts this into a junction temperature rise above the case temperature at that current, and then determines the maximum case temperature for the protection system. This requires that the "processor" is programmed with the actual characteristics of the SCRs and heatsink used on that particular starter.

QUOTE

over build the unit  
This is along the line that jraef covered above and is technically very sound, but commercially hard work due to the cost.
The big question is, how does the customer determine whether the unit is rated by "over build", good protection and design, or by fuse curves?
Good comments,
Best regards,

#12 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 12 February 2004 - 06:20 PM

Hi
As has been shown in the above discussions, we have two conflicting requirements in rating a soft starter.

  • Starter reliability
  • Starter cost and saleability
      In the interests of ensuring absolute reliability, we can take the approach of using SCRs that will never be overloaded under any possible start condition, or in the interests of minimum price we can use SCRs that will just hold in for a period of time in most cases. The two extremes.
      The mechanism for determining just where we are on this design curve varies considerably between design engineers and companies.
      If we look at the base problem that we are faced with, the reliability under normal operating conditions (voltage disturbances excepted), are related to the SCR and heatsink assemblies. There is primarily, a thermal circuit here and we need to ensure that the junction temperature of the SCR does not exceed the maximum junction temperature rating under any circumstances. If we allow the junction temperature to rise above the maximum operating temperature, there will not be immediate failure, but as in the case of contacts in a contactor, switch or circuit breaker, the life will be rapidly degraded.
      If we use the semiconductor fuse curves to determine the starter ratings, then we are operating the SCR with junction temperatures well in excess of the maximum ratings everytime we do a hot start, and we are reducing the life expectancy of the device.

      An induction motor requires a high starting current relative to it's rated running current. Under full voltage conditions, this starting current is in the order of 600% to 900% of the full load rating of the motor. With a soft starter, we reduce the start voltage by using phase control. We also reduce the starting torque by the square of the voltage reduction. The minimum start current is determined by the torque required by the load to start successfully, and with common applications is in the order of 300% - 450% FLC.

      An extreme and safe design strategy, is to continuously rate the SCR Heatsink assembly at the locked rotor current of the motor. It should then accept any abuse that can be thrown at it for much longer than the motor, and be ultra reliable provided there are no voltage disturbances. The problem with this model is the cost. It is just not commercial and will not sell well in the market place.

      Looking at the thermal circuit, we have an SCR which generates heat. The heat generated is a function of the current passed through it, but it is not a linear relationship. We have a heatsink that is capable of absorbing, storing and dissipating heat, and we have thermal resistance between the junction of the SCR and the heatsink. As the SCR size increases, the thermal resistance between the junction and the heatsink reduces, reducing the junction temperature rise for a given heatsink temperature and current. Additionally, as the SCR size increases, the power dissipated by the SCR reduces for a given current. In other words, for a given heatsink temperature, the maximum current will increase with increasing SCR size.
      The heatsink temperature is a function of the heatsink thermal resistance and and the power dissipated. Reducing the heatsink thermal resistance will reduce the temperature rise of the heatsink and allow an increased current through the SCR.
      The thermal mass of the heatsink influences the rate of rise and fall of the heatsink temperature and effectively helps to smooth out the temperature profile at the heatsink.

      The Soft starter effectively has three modes of thermal operation.

      • OFF time. No power is dissipated. The heatsink temperature slowly drops.
      • START time. Very High power dissipation, the heatsink temperature rises.
      • RUN Time. moderate power dissipation, the heatsink temperature trends towards a level temperature.

          The instantaneous junction temperature is going to be maximum at the end of the start period (due to the rising heatsink temperature) The actual junction temperature will be a fucntion of the heatsink temperature and the start power dissipated times the thermal resistance junction to heatsink.
          For any given SCR Heatsink combination, start time, run time, OFF time, and start amplitude, Analysis can give us the maximum rated current that will not exceed the maximum junction temperature at the end of the start. The ambient temperature is of course an important parameter in this equation.

          I believe that Soft Starter ratings should be based on junction temperature analysis based on the hardware used and the starting and duty cycle employed.
          A starter so rated will be reliable provided it is operated within the junction temperature constraints and is not subjected to severe voltage disturbances. I believe that this strategy determines the MAXIMUM applicable rating of the device for long term operation. i.e. the most commercial position without reliability compromise.

          Best regards,

#13 jraef

jraef

    Posting Freak

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

Posted 12 February 2004 - 08:30 PM

Well said Marke, now if only there were a universally understood (or at least accepted) measurement system we could apply to this concept...
"He's not dead, he's just pinin' for the fjords!"

#14 GGOSS

GGOSS

    Senior Member

  • Moderator
  • PipPipPipPipPip
  • 407 posts
  • Location:Melbourne

Posted 03 March 2004 - 07:28 AM

Great discussion! I am somewhat disappointed in myself for not getting involved earlier.

IEC 60947-4-2 will need to be re-written eventually. Contributing information now to those involved in the writing the new standard may assist those of us who truly understand soft starters and their application to achieve a desirable outcome.

Throughout my 20 plus years experience in the field of electronic motor control I have represented manufacturers who use fuse curves to derive their ratings, others who adopt an extremely conservative approach and several that fall somewhere in between.

Irrespective of the methods used by those manufacturers to define their ratings, we experience very few SCR failures annually and on further inspection of failed devices find that approximately 80% of those are transient over-voltage related.

Having said all of that, current related faults do occur and in the main I believe it is because products in the field are being abused ie either through miss-application, excessive duty or poor set-up. How many times for example have you seen a simple TVR soft starter rated to say 3 x FLC, 15 seconds, 45C driving a highly inertial load and set up for a 2 second ramp with a start voltage setting of 80 or 90%? Or for that matter the very same soft starter enclosed without bypass contactor and/or any form of forced air ventilation, positioned by the banks of a creek or river in sitting in direct sunlight? We can certainly continue to discuss how we proceed to make products more industrially robust but will we ever be able to make them idiot proof?

IEC 60947-4-2 falls short of defining how soft starter ratings should be derived. It outlines the format for presenting ratings information and promotes this as a means of assisting the product user to 'compare' products of different manufacture. However it does not do either of these very well. The affect of controlled stopping (be it soft stopping or braking) on starter ratings is not discussed, nor is it considered in the ratings format. Also manufacturers who claim they understand IEC 60947-4-2 specify ratings such as AC53b: 3.5-15: 3585 and in the very same documentation claim their product is suitable for 4 starts per hour. As I was taught an hour was made up of 3600 seconds I struggle to comprehend how it is possible to achieve 4 starts per hour with an off-time of 3585 seconds between starts! If the manufacturers canít understand it, how on Earth can we expect consumers to?

I find AC53b products to be particularly amusing, not so much because their ratings are unclear but because in most cases AC53b rated starters are open loop voltage controllers that employ AC1 rated PCB mounted relays to bypass the SCR's. No consideration is given to the fact that the starter may be missapplied/used to drive a highly inertial load and that the motor may continue to draw overload current for several seconds if not minutes after the bypass relay contactor has closed. So relay failure is imminent in such applications, and when it does occur the starters is often rendered the soft starter un-serviceable. To my way of thinking in-built bypass relays (contactors) should only be employed on closed loop controlled current soft starters that transition from start to run as the current falls below say 110% of full load.

The other thing that I find extremely amusing if not alarming is the fact that pressure to reduce product physical size, reduce price, increase standard features etc., etc., are items that I feel are more-so driven by manufacturers wanting to secure an advantage over their opposition than by the market itself. For example when I first became involved with soft starters I could barely lift a 5hp unit (slight exageration)and now can walk down the street boasting that I have 90hp in my pants. Although this might impress the ladies it does absolutely nothing for the product installer who has to connect his 50mm2 power cables to a product that is barely large enough to contain the power terminals. I spend a lot of time on the front line listening to product users and looking at products that are packed with features that are not being utilized. The classic of-course is the starter which sits in a pump shed (consisting on one pump only) 30km from the next nearest anything. It was specified and purchased because it has a comms port but will most likely never be connected to any form of network!

One day in the not too distant future Iím sure soft start manufacturers will come to understand that their most serious competition does not come from other soft start manufacturers and that reducing size & price whilst adding features to gain an advantage is not the answer.

I for one like the solution offered by BigMax because it gives consumers something that they can relate to. Having said that I also agree with both jraef and marke that a consensus for deriving soft starter ratings must firstly be established.

My apology for going off on tangents.

Regards,
GGOSS

#15 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 03 March 2004 - 10:01 AM

Hi GGOSS

Yes I agree with your comments.
I have however, seen the results of incorrectly rated units. I know of one installation where it is a very arduous duty, and in six months, six different starters failed on the one machine. The seventh starter was designed for the duty and was still operating happliy seven years after installation.
The "fuse" ratings will not result in and imediate failure, just a shortening of life. It is rather like the number of operations on a contactor. Reduce the rating and you will get more operations before a failure.

When I get a chance, I will illustrate the variations in rating for a given SCR heatsink assembly and this will help to illustrate the point. It would certainly be benificial in the market place if all ratings were derived the same way. The user can then compare apples with apples. At least with contactors, you have a measure of relativity when you look at the rated number of operations.

How about a few suppliers / manufacturers putting their hands up andshowing what their ratings mean and how they were derived.

Have a good day,
Best regards,

#16 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 03 March 2004 - 10:12 AM

Hi GGOSS


QUOTE

IEC 60947-4-2 will need to be re-written eventually. Contributing information now to those involved in the writing the new standard may assist those of us who truly understand soft starters and their application to achieve a desirable outcome.

Agreed. How do we get the right information to the right people?? Perhaps they could join in this discussion.
I for one would be happy to add my half cents worth to any standards commitee.

Best regards,

#17 GGOSS

GGOSS

    Senior Member

  • Moderator
  • PipPipPipPipPip
  • 407 posts
  • Location:Melbourne

Posted 03 March 2004 - 11:06 PM

Hello Marke,

It appears we agree but a coming at it from slightly different angles.

There's no doubt in my mind that it would be benificial if all ratings were derived the same way. As you quite rightly state, this would allow customers to compare products on a more 'apples for apples' basis and although I believe that was one of the objectives for IEC 60947-4-2, it has fallen short of delivering this.

The fact is that you do not need to comply with a standard to release a product to the market, it just helps your case when it is being specified. So even if the majors agreed to agree, other smaller manufacturers will look for avenues to make their products appear more competitive.

The angle that I was comming from relates more so to those people who are marketing soft starters as commodity items. I firmly believe this failure to educate the market can only damage the case for soft starters.

Where the right product is used in the right application, fear of premature failure can almost eliminated. But there's not much you can do to eliminate the idiot factor!

The market for soft starters comprises four distinct segments. The ability of manufacturers and/or suppliers to provide the right product (features, price, duty cycle) for each segment together with clear information about the application of the products is essential. For example saying to a customer that a starter is rated for 3.5 x FLC, 15 seconds generally does not mean a lot. Saying to that same customer the starter is rated for centrifugal pump applications and explaining that anything less than 3.5 x FLC, 15 seconds will compromise service life tends to bring the message home much more clearly.

Although I can sympathise with jraef (been there), my view has changed over the years. If the customer is not prepared to take the advice of an expert pre-purchase, then how can we expect that very same person to be reasonable when he/she encounters problems (whatever they may be) post sale. Quite frankly I prefer to let and in fact sometimes encourage those people to purchase from our opposition. It reduces heart-ache, and allows me to maintain personal credibility as well as company credibility.

As far as defining ratings go, I agree that using junction temperature as the reference point is the most sensible approach.

Regards,
GGOSS

#18 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 06 March 2004 - 04:55 AM

Hello all

To illustrate the variation of thermal ratings with starting conditions, (junction temperature limited to 125 degrees C under all normal conditions, I have taken an SCR module with a single fan cooled heatsink and run some calculations for this combination.


  • Continuous current (No overload) equivilent to AC1 Max rated current = 189 Amps
  • Start current = 300% for 30 seconds, 120 second cool down before restart. Max rated current = 101 Amps
  • Start current = 450% for 30 seconds, 120 second cool down period before restart. Max rated current = 73 Amps
  • Start Current = 600% for 30 seconds, 120 second cool down period before restart. Max rated current = 58 Amps.
  • Rating based on continuous LRC of 600% Max rated current = 31.5 Amps

      These junction temperature based ratings are all calculated with an ambient temperature of 45 degrees C
      If the ambient temperature is reduced, then the ratings can go up, and if the ambient temperature is increased, then the ratings must come down. Similarly, at a higher altitude, the ratings will reduce.
      i.e. Let us consider the 300% start current device.


      • ambient temperature = 30 degrees C, Maximum Current rating = 114 A
      • Ambient Temperature = 40 C, Maximum rating = 105 Amps
      • Ambient Temperature = 45 C, Maximum rating = 101 Amps
      • Ambient Temperature = 50 C, Maximum rating = 96 Amps
      • Ambient Temperature = 60 C, Maximum rating = 87 Amps


          Unfortunately, at present I do not have the fuse curves for this SCR, so I can not quote the fuse based start currents, but I can show the variation in current rating with start time.

          Once again taking the case of 300% start current, 45 C ambient and 120 second cool down before a start (one start per hour)


          • 5 Seconds start time Max rated current = 106 Amps
          • 10 second start time. Max rated current = 105 Amps
          • 20 second start time. Max rated current = 103 Amps
          • 30 second Start Time. Max rated current = 101 Amps
          • 45 second start time. Max Rated current = 98 Amps
          • 60 Second start time. Max Rated current = 96 Amps.


              From the above, it is important to relate the start conditions to the rating if the maximum junction temperature is not going to be exceeded.

              Best regards,

#19 marke

marke

    Posting Freak

  • Moderator
  • PipPipPipPipPipPip
  • 2,604 posts
  • Gender:Male
  • Location:Christchurch, New Zealand

Posted 14 March 2004 - 07:12 PM

Another scheme for determining ratings, is to take the average power input from the maximum continuous rating, and calculate the long term I2T and then calculate the start and run currents such that the long term I2T equals the continuous run I2T. The Run current must be reduced to make allowance for the start current squared times the start time portion of the period.

i.e. If we take the case of a starter rated at a continuous current of 189 amps, then we can calculate the I2T for one hour at that current as 128 million A2S (189 x 189 x 3600)
Now let us assume that we have a start current of 300%, 10 times per hour with an OFF time of 120 seconds before each start. We now have a rated current of 163 Amps. This is higher than the junction temperature limited rating of 101 Amps! The reason is, that this approach assumes that the thermal resistance between the junction of the SCR and the heatsink is zero, and that the thermal time constant of the heatsing assembly is significantly greater than the start cycle time. Additionally, the power dissipated is not a function of current squared only.
A rating based on 600% start current for 30 seconds, 10 times per hour with a 120 second cool down period yeilds a rating of 99 amps. The junction temperature limited rating is 58 amps.
As you can see, this approach yields large anomolies in the actual ratings that should be applied. With the I2T approach, the junction temperature will be well above the manufacturers specification and you can expect severe degradation of the life of the semiconductors.

I2T method
continuous run = 189 A
300% start = 163 Amps
600% start = 99 Amps

Junction temperature method
Continuous run = 189 A
300% start = 101 A
600% start = 58 A

Best regards,




1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users