Pages: [1]   Go Down
Author Topic: Are MOSFETs really this good?  (Read 514 times)
0 Members and 1 Guest are viewing this topic.
Offline Offline
Full Member
***
Karma: 5
Posts: 207
View Profile
 Bigger Bigger  Smaller Smaller  Reset Reset

I've been looking for MOSFETs to build a couple h-bridges for stepper motors (rated at 0.3A), and I'm really surprised at how tiny of a MOSFET I need to drive this motor. For the p-channel MOSFETs I'm looking at these http://www.irf.com/product-info/datasheets/data/irlml2246pbf.pdf
After doing some calculations it looks like I'll get about 80m-ohms drain to source with -12v gate to source at 70 degrees C. So I should be able to push the recommended max of 2.1A through it all day without heatsinking.

Is this all correct or am I missing something? I just can't believe that this tiny thing can handle 2.1A all day long. I know for a fact that I'm leaving out switching losses, but I honestly have no idea how to calculate that. I also don't imagine it'll have a big effect when I have such a big overhead.

EDIT: A question about flyback diodes. I know it's generally recommended to include separate ones and I usually do, but what's the theory behind this? It seems to me like the internal diode in the MOSFET should be able to handle the back EMF. Maybe someone more knowledgeable than me will be able to make something of the winding inductance in relation to back EMF voltage, so here's the motor product page: https://www.sparkfun.com/products/9238
« Last Edit: July 20, 2013, 12:48:15 am by bobthebanana » Logged

United Kingdom
Offline Offline
Tesla Member
***
Karma: 224
Posts: 6614
Hofstadter's Law: It always takes longer than you expect, even when you take into account Hofstadter's Law.
View Profile
WWW
 Bigger Bigger  Smaller Smaller  Reset Reset

I've been looking for MOSFETs to build a couple h-bridges for stepper motors (rated at 0.3A), and I'm really surprised at how tiny of a MOSFET I need to drive this motor. For the p-channel MOSFETs I'm looking at these http://www.irf.com/product-info/datasheets/data/irlml2246pbf.pdf
After doing some calculations it looks like I'll get about 80m-ohms drain to source with -12v gate to source at 70 degrees C. So I should be able to push the recommended max of 2.1A through it all day without heatsinking.

Is this all correct or am I missing something? I just can't believe that this tiny thing can handle 2.1A all day long. I know for a fact that I'm leaving out switching losses, but I honestly have no idea how to calculate that. I also don't imagine it'll have a big effect when I have such a big overhead.

You should design for worst-case values, not typical values. For calculating Rds(on), start with the worst-case value on the datasheet (135 millohms @ Vgs=4.5V) and then use the graph in Fig. 4 to calculate the value at higher temperatures. Also bear in mind that 12V is the absolute maximum Vgs for that mosfet and you should design your circuit to use a lower value.

EDIT: A question about flyback diodes. I know it's generally recommended to include separate ones and I usually do, but what's the theory behind this? It seems to me like the internal diode in the MOSFET should be able to handle the back EMF. Maybe someone more knowledgeable than me will be able to make something of the winding inductance in relation to back EMF voltage, so here's the motor product page: https://www.sparkfun.com/products/9238

It's perfectly OK to use the mosfet body diodes as the flyback diodes in an H-bridge. However, even better is to turn the corresponding mosfet on, so that the mosfet conducts the current instead of its body diode. This gives lower voltage drop and hence lower power dissipation. The snag is that you need a short delay between switching one mosfet off and switching the other on, to avoid shoot-through. This is why it is generally better to use a mosfet-based H-bridge chip than to roll your own H-bridge, because the required delay is taken care of for you.

Have you considered basing your design on a mosfet-based stepper driver chip, such as the A4988? These chips allow you to use a relatively high supply voltage so as to build up the coil current quickly, but have built-in current sensing and PWM current limiting.
Logged

Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

Offline Offline
Full Member
***
Karma: 5
Posts: 207
View Profile
 Bigger Bigger  Smaller Smaller  Reset Reset

Yea, I'm on the fence between making my own driver and using a real stepper driver, but I'm ordering components off Digikey and they don't seem to have any cheap stepper motor drivers. I was also a bit concerned about the fact that those MOSFETs were only rated for 12v Vgs, so I think I'll look for some other ones. And I guess I won't be using external diodes. That's also a good idea about using the MOSFET as the flyback diode, maybe I'll try that. Thanks for all the info.
Logged

United Kingdom
Offline Offline
Tesla Member
***
Karma: 224
Posts: 6614
Hofstadter's Law: It always takes longer than you expect, even when you take into account Hofstadter's Law.
View Profile
WWW
 Bigger Bigger  Smaller Smaller  Reset Reset

Digikey does sell the A4988 and related stepper drivers, see http://www.digikey.com/product-highlights/us/en/allegro-microsystems-a498x-motor-driver-family/1444.
Logged

Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

0
Online Online
Shannon Member
****
Karma: 206
Posts: 12167
Arduino rocks
View Profile
 Bigger Bigger  Smaller Smaller  Reset Reset

Of course they are that good, so long as you get modern devices, not the old versions such as the
IRF610 and such (which are hopeless by modern standards - people still sell them though).

Discrete MOSFETs use vertical current flow and have extremely low Rds(on) values (1 to 20 milliohms for
lower voltage devices is the sort of range).  With vertical current flow all MOSFETs on the die must
share the drain terminal, which isn't useful.

MOSFETs integrated onto chips (such as the A4988) are DMOS which have lateral current flow
and much higher on-resistances (0.2ohms being the lowest I've seen, 0.3 to 0.6 ohms is typical).

Thus H-bridges for more than an amp or two need to use discrete MOSFETs or you'll need big heatsink /
forced-air cooling on the chip.  The current goes through 2 devices so the power dissipation = 2 I^2 R

If you do use a discrete MOSFETs its normal to use all n-channel devices and bootstrapped high-side
drivers - n-channel MOSFETS have lower on-resistance due to increased carrier mobility of electrons
compared to holes (which carry the current in a p-MOS device).

Where MOSFETs fall down is on the sensitivity of the gate to feedback currents via the gate-drain
capacitance - the higher the voltage and current being switched the more careful you have to be
with the gate-drive to keep things under control.  For voltages above about 100V these days everyone
uses IGBTs which are far more robust.


[ My current project uses MOSFETs rated at 200A, 400A peak.  smiley ]
Logged

[ I won't respond to messages, use the forum please ]

Offline Offline
Full Member
***
Karma: 5
Posts: 207
View Profile
 Bigger Bigger  Smaller Smaller  Reset Reset

I've never understood the logic in using N-channel FETs on the high side of the h-bridge. If I'm using a 12v supply and I want to get 12v Vgs on the high side FETs for low resistance, wouldn't I need 24v relative to ground since the source would be at 12v when the motor is on?
Logged

United Kingdom
Offline Offline
Tesla Member
***
Karma: 224
Posts: 6614
Hofstadter's Law: It always takes longer than you expect, even when you take into account Hofstadter's Law.
View Profile
WWW
 Bigger Bigger  Smaller Smaller  Reset Reset

The reason for using N-channel mosfets on the high side is that you can get better (e.g. lower Rds(on)) N-channel mosfets than P-channel mosfets. If you are using discrete power mosfets instead of an all-in-one H-bridge chip, this normally means that you are switching higher currents than an all-in-one chip can manage; in which case you need very low Rds(on). There are plenty of H-bridge driver chips around that generate the gate drive voltage for the upper mosfets.

As you only want to switch 0.3A for your stepper motors, and all-in-one chip seems to me to be the obvious solution.
Logged

Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

Pages: [1]   Go Up
Jump to: