Question about fusing a power supply
On a product I am building, I have a power supply 120AC to 26VDC unregulated and switching 5 volts. I have seen power supplies short out in the past and smoke, so I was considering some protection against an unlikely event such as this. Would it be a good practice to include a fuse to one side of the transformer main input? Assuming so, how would I determine the value of the fuse? My thoughts are to measure the peak (power up) and average current limits, and use a fuse an amp higher. I am not sure whether fast or slow makes the most sense. The idea is for safety primarily, but I also don't want to be getting calls that the gadget has stopped working and find out the fuse was blown when there was not a real problem, I would assume slow blow is best suited for this concern. Any suggestions appreciated.
Comments
Leon
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tom Sisk
http://www.siskconsult.com
·
Leon
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle
On a similar subject, I started giving thought to a device that the Prop could turn off under certain circumstances:
The enclosure for the boards has a temperature sensor IC, so if something is overheating the enclosure will rather quickly change it's temp, so I could set a temp threshold to a value where the enclosure would never exceed, and if it does then have the Prop turn off the power. On any fault detected by the Prop (i.e. excessive current on the motor driver due to shorted mosfet/motor etc) it can stop the AC also. This is an idea of the extra protection in case of fuse failure. I have no ideas yet on such a device.
Depending on the design. that might not work as you'd like. An overheating component's heat can be·highly localized at first. A carbon resistor, for example, can fry in place and blow open. Although the heat can be intense enough to "charcoal" the board underneath, the amount actually radiated into the enclosure might not even register on the sensor.
If you know which components might overheat, you could put the sensors directly on them. Then again - and this might seem obvious - it would be better to stop the problem before it starts: in the design... give yourself enough duty cycle, headroom, load capacity of the components in the first place. Better to spend a bit more on components upfront than to use time, money, and aspirin.
- H
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I hacked a Genie garage door opener to make a gate opener. To make a long story short, Genie used an opto/chopper wheel on the back of the 120vAC motor to detect a drop in motor RPMs signifying an obstruction. This worked really well.It was very fast in detection and immediately reversed the motor.It also detected zero speed on the motor before reversing it. This is important with split single phase motors.(Signal phase reversible motors)
My point: RPMs are instant , Heat comes with time.
This is assuming the load is a motor.
_____I try to stay in the AC_________$WMc%____________________out______
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
The Truth is out there············___$WMc%___···························· BoogerWoods, FL. USA
Post Edited ($WMc%) : 10/20/2009 1:09:12 AM GMT