DC Inrush current limiting circuit ideas

[budm]

https://www.badcaps.net/forum/attach...3&d=1538991621
Your schematic showing I.E. IN1P input: when switch is in A positition the audio will flow through C123 and when it is in B position the audio will flow through 1uf cap instead, same for other inputs, I do not see any reverse phase input routing in Bridge mode.

[Dannyx]

The top B1 point is connected to the bottom B1 point. I just didn't draw wires because it would mean crossing over the entire circuit making a mess, so I used "indexes". In position B, after the 1uF cap, the audio also goes to IN2N. Same goes for the negative IN1N: when in position B, it also feeds the "positive" pin of the other channel down over at B2.

[Dannyx]

It all seems to be working now, except for one issue: when the signal source (the input "jack" so to say) is not plugged into anything, so the GND pin is floating, the speaker buzzes. If I plug the jack into a source, the buzzing goes away, so it's likely some noise from the adapter going into the signal line, since power GND and signal GND is the same thing. I tried separating the two GNDs, but it sounds rather distorted.

EDIT: now it's the stereo "mode" that doesn't want to work for some reason >.< very faint sound in the "bottom" channel.....

[Dannyx]

Sorted - one of the input caps on the non-functional channel decided to no longer conduct...probably due to the repeated contacts with the hot soldering iron...

Also a couple of pics of the "end product"......yeah, I know Enclosure time !

[Dannyx]

Originally Posted by AdrianM

D2 discharges C1 through the load when supply removed to reset delay for next time. Component values give roughly 0.5 seconds - if that's too short or long adjust C1 x R1. But watch out how long the Mosfet is in the linear region. May dissipate power if too long.

Back at the inrush portion: would C1 be electrolytic or something else ? I don't see a polarization mark on that schematic. Wanting to get this thing off the ground ASAP, I dug through my parts bin and found some WIMA caps which say 0.1 and them 100 under that....don't know what to make of this...is it 1 pico or something ? I suck at these

[mariushm]

Honestly if you're dealing with low voltages (<~35v) just get a LDO to get 5v at 50-100mA ,a 5v relay and a 6-8 pin microcontroller like a pic10/pic12 or a attiny something... they're like 30 cents excluding programmer.

You can get such relays which only need 25-30mA to stay energized, which can be easily provided by a mcu pin (but you can parallel 2-3 i/o pins for safety).

mcu has internal oscillator so you can count a few seconds in the micro,or optionally you can use a resistor divider or a potentiometer on an input pin and use adc to measure the value (say between 1-10s for voltage between 0v and 4v or whatever)

You're looking at minimum 5 components .. ldo, input and output cap (which could be ceramic depending on ldo), mcu, relay and a diode. optionally add potentiometer / 2 resistors to adjust time

No hassle with electrolytics which could degrade from heat, no zener diodes, no 555s that need 10 auxiliary things to set up etc etc...

Also if you want smallest power consumption (if those 20-30mA for relay are important) you could potentially use a bistable relay but you'd have to use some bigger capacitor to have enough "juice" to send a final pulse to the relay to reverse to off state when power is disconnected and the mcu can measure input voltage a few times a second and react accordingly.

[Dannyx]

Yeah, I would've certainly used a microcontroller if this were anything substantial, but as you can tell by looking at those pics...minimal care went into this

[Dannyx]

I built the circuit and it works. There is however one small caveat: when power is removed and reapplied in quick succession, a spark still occurs since the diode doesn't discharge the cap fast enough apparently so the transistor remains on just enough to conduct and allow inrush. True, I did deviate from the original design a bit: the capacitor I used is 220uF electrolytic instead of 100uF.

[budm]

The discharging rate is dictated by the resistance of the load. Did you check what the delayed on time is?

[Dannyx]

Can't really measure the on time - it's pretty instant.

[budm]

You have scope, right?
One thing is missing is the inrush current limiter, all you have is the delayed on switch (MOSFET) being switching on and off, you still need resistor to limit the current and when the delayed on MOSFET is turned on then it will bypass the current limiter resistor. The MOSFET will be in the linear region when the Gate as Gate Voltage rises until it reaches fully on state.. so I cannot tel how much current it will really limit the current to during that time period, if the MOSFET can take it then you do not need the resistor.
Check the inrush current with and without this existing setup and see what you have.

[Dannyx]

No, I don't have a scope sadly....would've made this a breeze. Well I can tell it's doing something, since it doesn't spark at all when the cap is indeed fully discharged and I do a "cold start". The transistor I BELIEVE is a B80NF10. Had it in my drawer. I believe I may have another answer as to why the transistor is being kept open even after power-down: the caps of the amp itself are discharging back into the circuit (sort of taking place of the supply which is now disconnected), since there's nothing prohibiting them from doing so. A series diode should solve this hopefully. It will cause a slight Vdrop of its own, but should be good enough. A SB5100 should suffice...

[budm]

The amplifier circuits will draw down the current from the filter cap when power supply is disconnected from the amplifier, so it will depend on how fast the filter caps will be discharged by the amplifier circuits, just check the Voltage on the filter caps to see how long it will take to discharge the filter caps.

[Dannyx]

Well guys, it's been 2 years since we've discussed this....a lot has happened since then, but it just so happens I came back to this inrush limiter idea for another audio project which would benefit from such a thing.

It just occurred to me that if you have a large load, say a couple of large caps rated at considerable voltage, a single FET would probably not be able to handle the inrush current. It would probably get hot or get destroyed immediately, so is it a good idea to parallel several of them where M1 is in the bottom right of the schematic ? That's probably not the best way of limiting large inrush currents - it's probably stepping into territory where a "resistor+relay" setup would be more robust and better suited.