Cheaper alternative to MIC5018 or MIC5019 boost MOSFET drivers?
Jan wrote 04/27/2019 at 11:36 • 1 pointHi guys!
For a recent project (https://hackaday.io/project/164742/log/162573-01-revision-v01-for-tests) I am using a MIC5018 to always fully drive my MOSFET to at least 10V at its gate. I want the lowest Rds_on possible at all times without using a logic level mosfet...
I searched for hours to find a similar device from a cheaper (chinese?) vendor. Those Micrel/Microchip ones are super easy to use with the integrated charge pump. Would be hard to make something cheaper from discrete parts I guess...
Do you know of similar devices? Help is much appreciated.
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.
Hey @Simon Merrett You're right, I need the voltage just for driving the MOSFET on/off. No fast switching is needed at all. Thanks a lot for your input!
Did a bit of research based on that and had a look at the https://en.wikipedia.org/wiki/Voltage_multiplier#Dickson_charge_pump.
The doubler wouldn't be enough at 5V in due to diode voltage drops and other inefficiencies but the tripler would be great. Unfortunately it needs two phase-shifted pulses, so two more I/O-lines would be needed (and I want to replace the Atmega328 with something less IO but muuuuch cheaper).
Another idea I had was using the super cheap https://datasheet.lcsc.com/szlcsc/Diodes-Incorporated-AP3012KTR-G1_C102618.pdf DC/DC chip with some 0603 passives which would work as a 5 to 15V tripler for around 0.2€ passives included at a scale of 10 boards.
Would have to calculate what is more expensive, more board real estate plus cheap parts or less PCB space and more expensive parts :)
Cheers, Jan
Are you sure? yes | no
https://www.electronics-tutorials.ws/blog/voltage-multiplier-circuit.html
Makes me think my pointer to the dickson doubler was a bum steer and a "conventional" tripler could be what you compare with the AP3012KTR. At the link above you can find a tripler that should work with a single square wave with three caps and three diodes. Any good?
Are you sure? yes | no
From a quick look at your project, I assume you don't need fast modulation on the MOSFET gate - just normally on and then switched off in relatively slow time (based on the fact that you have a 5s dip delay mode).
So how about a charge pump with discrete parts? If you can pulse your gate pin on the microcontroller then you can use a dickson doubler. Could work with a 5V regulator on your microcontroller or add more stages for lower Vreg/VCC.
For battery voltages above 10v, can you control the gate voltage using a resistor divider? Eg a 4S at 12-16.8V battery with a 10k resistor from gate to Vbat and a 2.2k resistor from gate to the microcontroller GPIO. Then for ON state put the microcontroller pin in high impedence mode (gate is pulled up to Vbat) and for OFF state put the microcontroller pin into an output mode and tie to ground (gate pulled to ~3V). You will need to find values this works for with the full range of voltages and I admit it doesn't sound workable for 3S and 2S voltages.
Are you sure? yes | no