Hello, i wanna utilize these here: https://de.aliexpress.com/item/1005002737105856.html but you need to limit charge current, balance (which is included on the chinese board), and manage/control when to charge again to make it a "standalone" power supply
since you dont wanna include measuring circuitry of the voltage, one could just use a power switch IC (which limits current well) and a isolated microcontroller to recharge in a timebased manner depending on the max. current consumption
4,8V - 5,2V, 250F, 100mA = 15 minutes discharge time
4,8V - 5,2V, 250F, 300mA = 5 minutes discharge time
4,8V - 5,2V, 500F (2 banks), 100mA = 30 minutes discharge time
4,8V - 5,2V, 500F (2 banks), 300mA = 10 minutes discharge time
Any better ideas here or am i on the right track?
since you dont wanna include measuring circuitry of the voltage, one could just use a power switch IC (which limits current well) and a isolated microcontroller to recharge in a timebased manner depending on the max. current consumption
4,8V - 5,2V, 250F, 100mA = 15 minutes discharge time
4,8V - 5,2V, 250F, 300mA = 5 minutes discharge time
4,8V - 5,2V, 500F (2 banks), 100mA = 30 minutes discharge time
4,8V - 5,2V, 500F (2 banks), 300mA = 10 minutes discharge time
Any better ideas here or am i on the right track?
i made a small circuit with a AP2552W6 (which is 3,3V and 5V combatible) and an 3V kemet 2A 230V signal relay (both available at lcsc.com)
its not complete yet, im kinda unsure what todo with the fault pin
there is a socket for my regulators, to regulate the possibly 5V down to 3V for the relays, for 3,3V input there is the possibility to use a jumper inplace of the regulator
also there is another 2x3 pin header for setting the charge current, i used resistors that roughly equal 1,8A, 800mA, 400mA, for 2A,1A or 500mA regulators
then there is a 2x2 pinheader that 1. lets you set a jumper to make the circuit always active, for when you just wanna hook up supercapacitors to a power supply, and 2. the other two pins are EN and GND ... to control this board via a external signal
its not complete yet, im kinda unsure what todo with the fault pin
there is a socket for my regulators, to regulate the possibly 5V down to 3V for the relays, for 3,3V input there is the possibility to use a jumper inplace of the regulator
also there is another 2x3 pin header for setting the charge current, i used resistors that roughly equal 1,8A, 800mA, 400mA, for 2A,1A or 500mA regulators
then there is a 2x2 pinheader that 1. lets you set a jumper to make the circuit always active, for when you just wanna hook up supercapacitors to a power supply, and 2. the other two pins are EN and GND ... to control this board via a external signal
i could add a LED to both, fault and en pin i guess
probably also a good idea to add a optocoupler to the EN input..
probably also a good idea to add a optocoupler to the EN input..
Last edited:
If you see any errors please let me now, i added PC817 for EN pin isolation, and 1 EN led and 1 fault led (well its an "OK" led.. i could also leave the EN led out because of this...)
1. i added a 1N4007 flyback diode on the relay coil
2. switched the AP2552 against the AP25562 counterpart, i saw on the mouser datasheet that the AP2552 is not recommended for new design anymore..
3. changed the ground scheme, unfortunaly because of the one trace still suboptimal but i stitched a bridge over it
could someone say something about my previous ground scheme vs the current one? i feel like the new one is better and less errorprone lets say..
i think i had the led wired wrong before, since its wired now in such way that it only light on fault condition of the power switch IC i added a charge led
and major upgrade i had completely forgotten about... i exchanged the regulator slot for a simple resistor for 5V for the 3V relay
i think its getting close 🙂
does someone has a great idea to make a under and over voltage "detection" ?
there are some things that could be done
1. micrcontroller on the supercap side measuring directly with ADC
2. microcontroller measureing voltage with adc over a linear optocoupler amplifier circuit (galvanic isolation)
3. TL431 as undervoltage and overvoltage detection via comporator mode on the supercap side
4. time based charging mode via microcontroller, no parts on the supercap side, lets say, each 20 minutes for 1 minute
5. measuring current with a hallsensor module like this one: https://www.aliexpress.us/item/1005006222090940.html and calculating the remaining charge
currently im tending towards either 4. or 5. solution... it makes it possible to still measure the current and therefor the remaining voltage, if done right, fairly well i think without any parts beside the hallsensor in the super cap path
do you have other ideas? 🙂
there are some things that could be done
1. micrcontroller on the supercap side measuring directly with ADC
2. microcontroller measureing voltage with adc over a linear optocoupler amplifier circuit (galvanic isolation)
3. TL431 as undervoltage and overvoltage detection via comporator mode on the supercap side
4. time based charging mode via microcontroller, no parts on the supercap side, lets say, each 20 minutes for 1 minute
5. measuring current with a hallsensor module like this one: https://www.aliexpress.us/item/1005006222090940.html and calculating the remaining charge
currently im tending towards either 4. or 5. solution... it makes it possible to still measure the current and therefor the remaining voltage, if done right, fairly well i think without any parts beside the hallsensor in the super cap path
do you have other ideas? 🙂
just found this circuit https://www.electronics-lab.com/project/optically-isolated-analog-input-module-for-arduino/ , the finished board is rather expensive with 40 euro
hmm i just had a thought, while the supercapacitor is charging it will trigger the current limit... does somebody know what happens to the attached load if the supercap essentially trys to charge with 1,8A with a 1,8A current limit? essentially there would be nothing left for the load right?
in theory the power switch could be also switched out by a resistor this would actually also allow for a more effective DC filter ....
it would be probably around 10-20x lower corner frequency, but in the end its 0,0001Hz (6 ohm) vs 0,002Hz (0,2ohm, your usual wire resistance...) with a 500F pack
i kinda like the resistor because we can get rid of the power switch... but it would definitely waste alot of power while charging, tho im also not too concerned about that...
downside is waste of power and the power resistor would need to be adjusted for charge current, voltage etc, but its probably still possible to calculate one resistor for 3,3 and 5V use
____
i think with this kind of pack because of this:
its also safe to say that there needs to be relay after the supercap for when its initally charging to switch the load on or off or the power rail will be probably unstable or unuseable ... i think once the caps reach 4,8V (for 5V use) and power cycle between 4,8 and 5,2 then the load still gets enough current even if we charge the caps (because current still wants to flow to ground even if the supercap trys to "buffer" the current, right?)
so i might rethink the design and add a second relay and a resistor instead of power switch
this would actually allow for a higher voltage charger too instead of just 5V (because of the power switch), that was also one thing that still kinda bothered me
it would be probably around 10-20x lower corner frequency, but in the end its 0,0001Hz (6 ohm) vs 0,002Hz (0,2ohm, your usual wire resistance...) with a 500F pack
i kinda like the resistor because we can get rid of the power switch... but it would definitely waste alot of power while charging, tho im also not too concerned about that...
downside is waste of power and the power resistor would need to be adjusted for charge current, voltage etc, but its probably still possible to calculate one resistor for 3,3 and 5V use
____
i think with this kind of pack because of this:
its also safe to say that there needs to be relay after the supercap for when its initally charging to switch the load on or off or the power rail will be probably unstable or unuseable ... i think once the caps reach 4,8V (for 5V use) and power cycle between 4,8 and 5,2 then the load still gets enough current even if we charge the caps (because current still wants to flow to ground even if the supercap trys to "buffer" the current, right?)
so i might rethink the design and add a second relay and a resistor instead of power switch
this would actually allow for a higher voltage charger too instead of just 5V (because of the power switch), that was also one thing that still kinda bothered me