Will fuse blow....with repetitive overcurrent pulses? 2

[schnell]

Hi

We are using the 500mA fuse of this series (Littelfuse 452 series)

FUSE DATASHEET:

http://www.littelfuse.com/data/en/Da...454_Series.pdf

The current flowing in this fuse will be as follows

CURRENT WAVEFORM:

http://i55.tinypic.com/2ed9hf8.jpg

its peak is 1.2A
its ON for 6.2us, then OFF for 31us...repetitively

The current average value is 204mA.
The current RMS value is 494mA.


This current waveform will flow continuously,
-will the fuse blow?
-or will we encounter occasional nuisance tripping?

the datasheet explains what would happen for single overcurrent events....but not for repetitive overcurrent pulse streams.

 

[IRstuff]

Most of the fuse literature discuss the ability to survive current surges, which are termed "non-repetitive," but the design examples clearly use applications where the surges are repetitive, albeit not to the same degree as yours:

http://www.littelfuse.com/data/en/A..._Application-Note-Slow-vs-Fast-Auto-Fuses.pdf

http://www.cooperbussmann.com/pdf/e117a412-79c0-436e-943e-969a1f80ad6e.pdf

There are some guides and discussions available on the web:

http://www.sandc.com/webzine/2003/012003_1.asp

http://www.eeel.nist.gov/817/pubs/spd-anthology/files/Repetitive surges.pdf

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize

 

[waross]

One of IRstuff's links has the following information which is probably applicable to your application.

Thermal cycling
Thermal cycling applies to all fuse technologies and in fact to all
materials but may be considered especially important to fuses
with the active mechanism of diffusion pill technology. While the
diffusion pill design makes for a very effective time-delay fuse, what
happens if the overcurrent does not last long enough to blow the
fuse? Each time a short overcurrent pulse flows through the fuse
the element will heat up, and some tin-copper diffusion will take
place, permanently changing the characteristics of the element.
With enough pulses the fuse will eventually fail. If there is some
question as to whether a particular fuse type will survive in a pulsed
application, then it is best to perform an I2T analysis. I2T (“I squared
T”), expressed in ampere2 seconds is a measure of the total energy
dissipated in the fuse during a pulse. This involves calculating the
ratio of the applied I2T to the fuse’s nominal melting I2T, then
looking at the table below and Fig. 4.

LPS, IR

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[VEBill]

Is there a good reason to use a fuse so close to the actual (RMS) current? Why not use a 1 amp fuse?

Depending on the specific application, sometimes there's no good reason to keep the fuse so tight to the edge. But sometimes there is... It depends on the foreseen failure modes, safety and planned repair schemes.

 

[Comcokid]

The fuse curves and I2t of your really seem to indicate you will have no problems.

However, you could have trouble anyway. A few years back I used a fuse like this but higher current rating. I found that RoHS soldering temperatures modified the fuse blowing charcteristics, effectively turning a slow-blow into a super-fast acting. I'm noticing that the datasheet is dated 2009 and now containg a very exacting reflow profile now!

You could put an inductor in series with the fuse, or a capacitor parallel (depending upon how the fuse is used) to reduce the peak current.

 

[schnell]

hi

please may i expand a bit..............

The fuse is on the output of the PSU board, which feeds the stepper motor driver board.
the stepper driver board USED to contain an MC34920 stepper driver IC, which contains 4 h-bridges on-chip.....to drive our two stepper motors.

as luck would have it, MC34920 has just gone obselete.

So we are now (intending to be) using 2 pieces of NJM2673 stepper driver IC.

The MC34920 is a chopper driver, but synchronizes all the switching in all 4 of its H Bridges so that the rms current is low.

the NJM2673 is not sync'd, and there is every chance that they will all switch together and give the current waveform above.

I cannot find another IC like MC34920...and my company like NJM2673 becauuse its very cheap.

Anyway, the reason why the fuse is set at 500mA.........

....this is so that if a h-bridge goes short, then the stepper motor current will rise to [Vbus/phase resistance], and this could catch fire due to overheating.......so thats why our approvals engineer set the fuse at 500mA.

The stepper motor is a Crouzet 15.5 ohms per phase and artound 15mH....(bipolar)
-its rated at 5.6V so the rated phase current is 361mA.

The input bus voltage is 36V.

to make matters worse, their is a 5V SMPS running off the 36V too......and this supplies 5V at 230mA to LEDs, stepper driver control pins, etc.