For high inrush DC current there is a solution with a P-FET as polarity protection. Adding a zener between the pulled down gate and the source pin, it would protect from high current too, as the zener is limiting the gate voltage. Bourns has an excellent part, the TBU-CAxxx series for limiting high inrush current. Fast acting, robust, though more expensive than a few passive parts.
Thanks! It's so calming to watch your videos. As a matter of fact, my current occupation is no longer connected to analog design (making software now). But still enjoying watching you! Want to make some DIY analog design as a hobby as soon as I will feel like doing it. And videos like this are helping me to gather motivation!
@@FesZElectronics I am not thinking about something in particular. Varistors have the reputation to not age well so I wonder if like rectifiers there is some active method to limit inrush current. Many dc regulators allow or embed soft start features so maybe there is a way to make some general circuit a bit more elaborate to mitigate the inconveniences of the NTC or the relay solutions presented at the end of the video.
I think you are right. The diode does have some small junction capacitance. However, the large peak current will probably not occur in real life, or at least not to that extent, if you consider the diode terminal resistance and inductance...
Interesting topic, I found schottky diodes very useful at inductive spikes, faster/more efficient than MOV - but a combination with the MOV more energy can be absorbed.
Its true that in certain high power application multiple protection elements are used, the exact combination being based on reaction speed and power handling capability. Its not uncommon to see a capacitor in parallel with a TVS or some other diode, and GDT (arrestor). Depends on the application and the specific pulse that needs to be handled.
This somewhat reminds me of automotive transient protection circuits, LTspice has all the ISO 7637-2 & 16750-2 waveforms, can you add this one to the list. Thanks.
This was very interesting because I'm currently having issues with the switch for the garbage disposal barely closing. I wonder if this in some form can meet the US electrical code.
For the capacitive loads, how about placing after the switch an inverse diode to ground, then a series inductance in parallel with a resistance and then a zener to limit the voltage spike on the circuit, if necessary? Ok, the voltage on the circuit will increase, but just a little, right? The circuit will remember a boost converter, but it will only "work" on one cycle.
Smart idea! But inductive elements are usually expensive and big. I think the P-MOS solution someone mentioned in another comment will win. And in practice for loads like a few 100's W a NTC resistor I encountered most often.
@@LukaszTNT I saw the comment with the P-MOS solution and I didn't get how it can limit the inrush current. In the moment you apply voltage, there is the time to charge the gate of the mosfet and the resistance drops so quickly that I think it's possible to have a still very high inrush current to charge up the capacitors. And if the effect is due to the on resistance of the mosfet, just replace it for a miliohm range resistor.
Got lazy with mine… have a seperate 5V psu, always on driving a raspberry PI running Octoprint, so I just connect the raspberry PI to one of those SSR to control the 24V PSU for the printer. Normally those SSR’s do have integrated snubbers, so you dont have to worry about all those kind of things.
Useful video, but why not to use some common industrial modular contactor with NC and NO switches? They are rated for high voltage and current are relatively cheap and quick to mount (no soldering required). Besides, your printers power supply should have its own protection from inrush current. Of course contactors are quite noisy, but probably still more quiet than a 3d printer.
I never thought of using that type of contactor; the switches and relay where some components I already had around, that is why I wanted to use them. Anyway, regarding the printer supplies built in inrush protection, I did not check it in detail, but the box has a 10A rated switch - so I guess the inrush limiter will protect that, however the extra switch I used was only 1A rated... The relay that will keep the printer on is 10A rated, but I wanted to use a low current switch for turn on.
Maybe I don't see what's going on here. Can you explain how the circuit @0:56 works... The mains switch and a relay are in parallel. If the mains switch is a physical/mechanical device that is normally on, it will pass current, regardless of the state of the relay switch. How does the relay interrupt the mains switch?
I guess what I did not fully highlight is that both switches (for turn on and turn off) have just one stable position - Open for the mains, Closed for the relay driving; while you are pushing on them they change state, but once you let go they revert - like a push button. So the Mains switch turns on the supply, but the relay keeps it on as long as the relay is supplied
@@FesZElectronics OK: The mains switch is a momentary type that only passes current while depressed. Makes sense now. BTW: For relay bounce/oscillation... Years ago, I ordered 100, 1N4001 diodes and the vendor shipped 1000 instead. I offered to return them but they said "keep it". I'll be using those for relay over-shoot for the foreseeable future.
Hello Fesz! I always appreciate your videos! :D One question: at 12:30 you say that the protection works for both DC and AC. but I think that some AC current should be passing throught the capacitor (depending on the capacitance value). I am not sure, though
Indeed, its not perfect for AC circuits, but, if the AC frequency is far smaller than the transient, then its ok. For example in 50/60Hz circuits, with a small enough capacitor, the impact will be minimal on the "off" state.
As of my understanding, a diode and zener back-to-back will still permit a negative voltage equal to the zener voltage (minus the another diodes voltage drop). For AC it is fine as long as the negative glitch is within the AC limits.
For AC, you don't normally know when the power will cut out - on the positive or negative halfwave. So the inductive kickback can have any polarity. The Diode-Zener combination will only work on AC if you can ensure the turn off always occurs when the AC is positive. Dual zenner does not have this issue - it will protect regardless of the polarity.
Great video, i like the details. Thank you
your videos are very useful
For high inrush DC current there is a solution with a P-FET as polarity protection. Adding a zener between the pulled down gate and the source pin, it would protect from high current too, as the zener is limiting the gate voltage. Bourns has an excellent part, the TBU-CAxxx series for limiting high inrush current. Fast acting, robust, though more expensive than a few passive parts.
This is a very interesting topic !
Thanks! It's so calming to watch your videos. As a matter of fact, my current occupation is no longer connected to analog design (making software now). But still enjoying watching you! Want to make some DIY analog design as a hobby as soon as I will feel like doing it. And videos like this are helping me to gather motivation!
Nice explanation to protect a switch.
With the capacitive load we can protect the switch using a serie inductance with freewheeling diode
nice video Fesz. The topic is very interesting as it's often overlooked. Will you cover more complex active methods in a future video?
I did not think about that. What sort of active methods are you thinking about - is there something specific you tried in the past?
@@FesZElectronics I am not thinking about something in particular. Varistors have the reputation to not age well so I wonder if like rectifiers there is some active method to limit inrush current. Many dc regulators allow or embed soft start features so maybe there is a way to make some general circuit a bit more elaborate to mitigate the inconveniences of the NTC or the relay solutions presented at the end of the video.
12:18 that might not be a simulation error but rather the diode capacitance.
I think you are right. The diode does have some small junction capacitance. However, the large peak current will probably not occur in real life, or at least not to that extent, if you consider the diode terminal resistance and inductance...
@@FesZElectronics And even if it does, the duration is so small it's practically negligible.
Interesting topic, I found schottky diodes very useful at inductive spikes, faster/more efficient than MOV - but a combination with the MOV more energy can be absorbed.
Its true that in certain high power application multiple protection elements are used, the exact combination being based on reaction speed and power handling capability. Its not uncommon to see a capacitor in parallel with a TVS or some other diode, and GDT (arrestor). Depends on the application and the specific pulse that needs to be handled.
This somewhat reminds me of automotive transient protection circuits, LTspice has all the ISO 7637-2 & 16750-2 waveforms, can you add this one to the list. Thanks.
We can also put a inductor in series to limit the current with a capacitive load
This was very interesting because I'm currently having issues with the switch for the garbage disposal barely closing. I wonder if this in some form can meet the US electrical code.
For the capacitive loads, how about placing after the switch an inverse diode to ground, then a series inductance in parallel with a resistance and then a zener to limit the voltage spike on the circuit, if necessary? Ok, the voltage on the circuit will increase, but just a little, right? The circuit will remember a boost converter, but it will only "work" on one cycle.
Smart idea! But inductive elements are usually expensive and big. I think the P-MOS solution someone mentioned in another comment will win. And in practice for loads like a few 100's W a NTC resistor I encountered most often.
@@LukaszTNT I saw the comment with the P-MOS solution and I didn't get how it can limit the inrush current. In the moment you apply voltage, there is the time to charge the gate of the mosfet and the resistance drops so quickly that I think it's possible to have a still very high inrush current to charge up the capacitors. And if the effect is due to the on resistance of the mosfet, just replace it for a miliohm range resistor.
Got lazy with mine… have a seperate 5V psu, always on driving a raspberry PI running Octoprint, so I just connect the raspberry PI to one of those SSR to control the 24V PSU for the printer. Normally those SSR’s do have integrated snubbers, so you dont have to worry about all those kind of things.
Useful video, but why not to use some common industrial modular contactor with NC and NO switches? They are rated for high voltage and current are relatively cheap and quick to mount (no soldering required). Besides, your printers power supply should have its own protection from inrush current. Of course contactors are quite noisy, but probably still more quiet than a 3d printer.
I never thought of using that type of contactor; the switches and relay where some components I already had around, that is why I wanted to use them. Anyway, regarding the printer supplies built in inrush protection, I did not check it in detail, but the box has a 10A rated switch - so I guess the inrush limiter will protect that, however the extra switch I used was only 1A rated... The relay that will keep the printer on is 10A rated, but I wanted to use a low current switch for turn on.
Maybe I don't see what's going on here. Can you explain how the circuit @0:56 works... The mains switch and a relay are in parallel. If the mains switch is a physical/mechanical device that is normally on, it will pass current, regardless of the state of the relay switch. How does the relay interrupt the mains switch?
I guess what I did not fully highlight is that both switches (for turn on and turn off) have just one stable position - Open for the mains, Closed for the relay driving; while you are pushing on them they change state, but once you let go they revert - like a push button. So the Mains switch turns on the supply, but the relay keeps it on as long as the relay is supplied
@@FesZElectronics OK: The mains switch is a momentary type that only passes current while depressed. Makes sense now. BTW: For relay bounce/oscillation... Years ago, I ordered 100, 1N4001 diodes and the vendor shipped 1000 instead. I offered to return them but they said "keep it". I'll be using those for relay over-shoot for the foreseeable future.
Hello Fesz! I always appreciate your videos! :D
One question: at 12:30 you say that the protection works for both DC and AC.
but I think that some AC current should be passing throught the capacitor (depending on the capacitance value).
I am not sure, though
Indeed, its not perfect for AC circuits, but, if the AC frequency is far smaller than the transient, then its ok. For example in 50/60Hz circuits, with a small enough capacitor, the impact will be minimal on the "off" state.
As of my understanding, a diode and zener back-to-back will still permit a negative voltage equal to the zener voltage (minus the another diodes voltage drop). For AC it is fine as long as the negative glitch is within the AC limits.
For AC, you don't normally know when the power will cut out - on the positive or negative halfwave. So the inductive kickback can have any polarity. The Diode-Zener combination will only work on AC if you can ensure the turn off always occurs when the AC is positive. Dual zenner does not have this issue - it will protect regardless of the polarity.