Fuse Resistance

[Brandon Loyd]

Using values 45, 51, 59.

NETA MTS says to investigate fuse-resistance values that deviate from each other by more than 15 percent. How would you calculate this?

For contact and bolted connection resistance NETA is very clear saying to investigate values...by more than 50% of lowest value. Easy math 45 + 50% = 67.5. All values are within 50%. No problem.

For fuses, would I add 15% to the lowest value, deduct 15% from the highest value, or find the average then go +/- 15% from the average?

I can't seem to find a definitive answer anywhere, so hoping someone here with NETA/ANSI expertise can help. Thanks in advance!

 

[synchro]

For fuses, would I add 15% to the lowest value, deduct 15% from the highest value, or find the average then go +/- 15% from the average?

I can't seem to find a definitive answer anywhere, so hoping someone here with NETA/ANSI expertise can help.

I don't know what specific standards NETA/ANSI might have relating to this, but here's my take on this:

Don't take the average (i.e., mean) of the resistance values as a reference point. The average can be skewed off too much by any "outliers" that depart significantly from the other values. Instead, take the median of the set of resistance values, which is the middle value when the values are ordered in sequence from low to high (or high to low), and then use it as your reference point. If the number of values is even, take the average of the two middle values. Then determine how much each value deviates from this median value on a percentage basis. Identify which if any resistances are more than 15% above or below the reference value.

 

[topgone]

If I may add, you don't compare resistance values of different fuses. Those 15% variance is applicable to fuses of the same fuse manufacturer.

 

[steve66]

To me, 15% is 15%. It's not + or - 15%.

Just take the highest minus the lowest and divide by the lowest and multiply by 100.

(59-45) = 14.

14/45 * 100 = 31%.

59 is 31% off from 45 and I would say they warrant investigation.