mosfet load switch

[andy257]

Ive dug this circuit back out to have a play with. I have added a 5V switching regulator to the output of the circuit. However i think this circuit is causing havock with my low voltage detection.

what i want is to have is a 12V to 9V output, but also have a steady 5V output tapped off from the same 12v to 9V output.

The trouble i am finding with this is it screws up my calculations and in some cases it prevent my TL431 from complelty turn off the mosfet (which get very hot due to load current). I think its either the capacitors in conjunction with the 240K hysterisis resistor or the actual switching reg itself.


Is there any way i can isolate the two circuit so neither effect the perfromance of each other?

 

[MikeMl]

What do D1, D2, C1 and R6 represent?

I have an idea that the circuit and switcher are fine, but it gets into some funny state due to the your weird source??

 

[andy257]

Hi Mike,

C1 is a supercap, 5 Farad or 10 Farad depending how much energy i require to power the load. D1 and D2 are sschottky steering diodes to prevent the capacitor discharging into the PSU and to bypass R6 in order to get maximum current flow out of the capacitor. R6 is used in the charging phase to adjust charging rate.

Ive made this circuit up on breadboard and its driving me crazy. It works fine with a resistive load but when i attach the small board computer (SBC) the mosfet never turns off fully and i get very strange oscillation close to the cutt off point. This shouldnt happen because my hysterisis is around 1V. Ive tried this with only the mosfet and TL431 without the front end capacitors etc and used a variable power supply. i still get the same symptoms.

Can you help any

thanks

 

[MikeMl]

What the Effective Series Resistance (ESR) of the Supercapacitor?

 

[andy257]

The supercap is a BCAP0050 or BCAP0025. I am using 5 of them in series to get a 12V working voltage. The ESR of them is 20m ohms.

UPDATE: Circuit is working as intended now. I needed to reduce the 4k7 resistor from cathode to source to around 1k. I dont think the tl431 was gettting the required 1mA to stabalise. All seems fine now.

thanks mike

andy