Hi All,
I have a project that I'm powering with a 'brick' external power supply, the supply provides +5 volts @ 7Amps max and +24V @ 2Amps max. The problem is, the power supply can be connected and disconnected with the power switch set the on position. The connector is not a hot swap type; I'm using an eight pin Din connector. The power supply is used to power a motor controller; +5volts used for control logic and the +24volts is used to drive stepper motors. I know that someone will connect the external power supply with the power switched on. I'm looking for some type of circuit protection to lower the chance of damaging the electronics in the controller. Does anyone have a simple method of protecting the controller electronics?
Power Supply 'Hot Swapping'
You cannot prevent all accidents from happening. It will require some prudence on the part of the operator.
Install an E-stop and LED indicator in series with the DC voltages.
We had a similar circumstance in my last job wherein we hot-plugged Molex connectors to our drives and sometimes we would blow the onboard CPLD chip. We fixed the problem by installing E-stops and never had the problem again.
Install an E-stop and LED indicator in series with the DC voltages.
We had a similar circumstance in my last job wherein we hot-plugged Molex connectors to our drives and sometimes we would blow the onboard CPLD chip. We fixed the problem by installing E-stops and never had the problem again.
-
Chris Smith
- Posts: 4325
- Joined: Tue Dec 04, 2001 1:01 am
- Location: Bieber Ca.
The best you may be able to do is to install a large Cap at the circuit level input to hold the vlotage for a long while inbetween transients, and a Zener diode to ground in case the voltage in, rises or spikes above a certain level.
Also low voltage spike protectors like the MOV can help.
Nothing is 100%?
Also low voltage spike protectors like the MOV can help.
Nothing is 100%?
If you are mainly concerned about plugging the power supply in while the device's switches are on, why not use a triac (or SCR) and a momentary switch to turn it on? Thus, once the triac has had power removed, it won't turn on until the gate voltage is non-zero (see the datasheet for correct amount). The momentary switch is used to turn the TRIAC on and it will stay on until power is removed.
You won't stop some one from plugging in the power and hitting the on button but at least it's a 2 step process and you can lable the button.
You won't stop some one from plugging in the power and hitting the on button but at least it's a 2 step process and you can lable the button.
-
Robert Reed
- Posts: 2277
- Joined: Wed Nov 24, 2004 1:01 am
- Location: ASHTABULA,OHIO
- Contact:
The oldest safety device of all - the 'interlock'. If you have two pins available on your connector, short the two on the downsteam connector. Then run two wires from the other connector back to your power supply. From this point you have many options for it's power up, power down, such as AC power,Delay relays etc. Don't know if you require timed or sequential operation, but again that could be started wih the interlock making contact.
I haven't worked out all the permutations but here are the basics as I see it.
Assuming the supply does not have to slew large currents (rapid supply current change) then the simplest would be an LC filter. A Capacitor C across + and Ground tends to support voltage sags and dampen voltage spikes and the series inductor L dampens large surges in current. I might also put a resistor across the common node to regulate supply to supply surges.
Snce super caps have dropped so much in price lately, I might start there otherwise a beefy electrolytic is whats called for (47,000Uf, 35V). As for the inductor, I am not sure where to start, certainly not very large but large enough to limit surges. MAybe 20-100 turns of #12 around a 1/2" pipe to start. Smaller wire if current is not that big.
Depending on the supplies, there may also be sense or control lines to regulate or inhibit the output. An inhibit line can be connected to an RC circuit to hold it off for a moment after plugging in so it only activates after a solid connection is made. A sense line can also be used to ensure the output matches the presently inserted supply to reduce supply to supply current surges that would cause voltage spikes.
Ferrite beads around the control cables or inputs will further inhibit current spikes that cause voltage glitches.
The major failure mode to engineer out would be if the supply voltage sagged while some inputs were still powered to full (logic 1) value. This would cause large currents to enter the input protection circuits on those pins and if sustained can cause Electrical Overstress (EOS) damage or device latchup (if it is a susceptable CMOS device) which also can lead to EOS or temporary malfunction. Ground bounce can cause a similar EOS damage to logic 0 inputs
Secondarily, large overvoltage on the power pins of an IC device can cause EOS damage due to junction breakdown or gate oxide damage.
Assuming the supply does not have to slew large currents (rapid supply current change) then the simplest would be an LC filter. A Capacitor C across + and Ground tends to support voltage sags and dampen voltage spikes and the series inductor L dampens large surges in current. I might also put a resistor across the common node to regulate supply to supply surges.
Snce super caps have dropped so much in price lately, I might start there otherwise a beefy electrolytic is whats called for (47,000Uf, 35V). As for the inductor, I am not sure where to start, certainly not very large but large enough to limit surges. MAybe 20-100 turns of #12 around a 1/2" pipe to start. Smaller wire if current is not that big.
Depending on the supplies, there may also be sense or control lines to regulate or inhibit the output. An inhibit line can be connected to an RC circuit to hold it off for a moment after plugging in so it only activates after a solid connection is made. A sense line can also be used to ensure the output matches the presently inserted supply to reduce supply to supply current surges that would cause voltage spikes.
Ferrite beads around the control cables or inputs will further inhibit current spikes that cause voltage glitches.
The major failure mode to engineer out would be if the supply voltage sagged while some inputs were still powered to full (logic 1) value. This would cause large currents to enter the input protection circuits on those pins and if sustained can cause Electrical Overstress (EOS) damage or device latchup (if it is a susceptable CMOS device) which also can lead to EOS or temporary malfunction. Ground bounce can cause a similar EOS damage to logic 0 inputs
Secondarily, large overvoltage on the power pins of an IC device can cause EOS damage due to junction breakdown or gate oxide damage.
Who is online
Users browsing this forum: Ahrefs [Bot], Google [Bot], haklesup and 51 guests