For a new project, I'm trying to create my own BLDC motor driver that is capable of position control. I already have working code from a previous project of mine, but I'd like to go through the experience of also prototyping and designing my own motor drive. For this, I am aware that I need 3 sets of N and P channel MOSFETs. But that's where the problems start.
Right now I'm looking at all the available MOSFETs and to be honest, I'm quite lost in which ones to select. It will be logic level driven (preferably < 3V3 but < 5V is also fine), Vmotor is about 12 Volts and the power supply is currently limited to 1.5A for the prototyping phase. Should the N and P channel MOSFETs match as closely as possible? Or doesn't that matter as long as they're beefy enough?
I do know the basics of MOSFETs, how they work and which values to look for in a datasheet. There are just so many of them and with my inexperience in motor drive design, I'm hesitant to just by a bunch and see which work out.
Once I have a little more experience, I'm looking to get it all surface mounted with a driver circuit so the MOSFETs don't have to be logic level anymore, but for now everything will be breadboarded.
The ratings of MOSFET's in inverter are dependent on the currents to handle at particular voltage levels to operate in Safe operating region. So mention the Switching frequency, input voltage, the required peak power to drive the application. According to the spec's the switches can be chosen.
Can you explain in detail the position control of the BLDC motor and share the code.
For a new project, I'm trying to create my own BLDC motor driver that is capable of position control. I already have working code from a previous project of mine, but I'd like to go through the experience of also prototyping and designing my own motor drive. For this, I am aware that I need 3 sets of N and P channel MOSFETs. But that's where the problems start.
6 n-channel MOSFETs and a FAN7388 work well. Using complementary MOSFETs is only viable for a limited
range of voltages and is not how this is done these days, everyone uses appropriate bootstrapped drivers.
The FAN7388 and FAN7888 are 3-phase drivers that I've used in several projects (alas SMT-only
which will affect some people). You can use 3 half-bridge drivers instead.
The datasheets for high-low MOSFET drivers are a great resource about how to do this stuff.
You'd probably be wise to start with just a half-bridge initially (fewer devices to fry!). Perhaps look at the IRS2001 and IRS2112 half-bridge drivers as they are available in DIP. Or the HIP4081A (full H-bridge, ie you need 1.5 chips for 3-phase!)
With MOSFET drivers you don't need logic-level MOSFETs, but you do need a clean 12V supply (free from
spikes and dips). A good safety precaution is add a resistor of a few kohms in line with each logic
signal to the gate driver, in case the driver gets fried and tries to push 12V back at the Arduino.
Your choice of MOSFETs depends on the voltage rating and the on-resistance rating (coupled with
how much heatsinking you use). For surface mount dual-MOSFETs are available in SOIC8 pinout.
[ Oh, one very nice feature of using bootstrapped high-low drivers with MOSFETs is the the main power to the MOSFETs and the power to the logic/driver can be completely independent - running with one on and the other off has no repercussions. Just common the grounds as usual. And the MOSFET voltage is not limited to specific values, you could run at just 1V or upto the max voltage the relevant parts are rated for.
6 n-channel MOSFETs and a FAN7388 work well. Using complementary MOSFETs is only viable for a limited
range of voltages and is not how this is done these days, everyone uses appropriate bootstrapped drivers.
The FAN7388 and FAN7888 are 3-phase drivers that I've used in several projects (alas SMT-only
which will affect some people). You can use 3 half-bridge drivers instead.
It took me a while to understand how this works, but it's actually pretty clever, thanks for pointing me towards it!
You'd probably be wise to start with just a half-bridge initially (fewer devices to fry!). Perhaps look at the IRS2001 and IRS2112 half-bridge drivers as they are available in DIP. Or the HIP4081A (full H-bridge, ie you need 1.5 chips for 3-phase!)
I've been looking at these solution and finally decided (after looking at the availability of the IRS chips in my country) to go for the DRV8313 (Datasheet). It's the same principle but only requires 3 PWM inputs.
As it is a SMT part, I decided to just design a breakout board for it. It should be a nice fit for the job, or did I miss anything?
I'll experiment with this one first before going full power (both on design and on the motor). those duo MOSFETs sound really nice, if the thermals can handle it!
I'll elaborate a little about the project I'm using this for btw. I'm trying to build at least part of an exoskeleton and these position controlled bldcs are going to be the motors for the joints. I do have some experience with those, since I took part in this project as an embedded control engineer.
