Offline
Newbie
Karma: 0
Posts: 24
|
 |
« on: May 06, 2012, 08:37:19 am » |
Hello everyone, I am trying to control a dc motor in both the directions using mosfets wired in an h bridge, but no matter how much i try, i cannot get it to run. I am using 4 N channel mosfets (IRF540n) and wired it up according to this schematic - http://modularcircuits.tantosonline.com/blog/articles/old-h-bridge-secrets/part-1/Please help
|
|
|
|
|
Logged
|
|
|
|
|
Massachusetts, USA
Offline
Tesla Member
Karma: 96
Posts: 6372
|
 |
« Reply #1 on: May 06, 2012, 09:03:02 am » |
What is the source voltage are you trying to control? Did you read the section on "P- VERSUS N- CHANNEL HIGH-SIDE SWITCHES"? How are you generating the "Source voltage plus five volts" needed to close the high-side N-channel switches?
|
|
|
|
|
Logged
|
|
|
|
|
Offline
Edison Member
Karma: 9
Posts: 1001
|
 |
« Reply #2 on: May 06, 2012, 12:33:33 pm » |
That schematic is wrong, and your mosfets are wrong.
If you use two N-channel "logic" mosfet at ground and two P-channel "logic" mosfets at +5V you can have a bridge for 5V. If the voltage is higher, you need more electronic components.
|
|
|
|
|
Logged
|
|
|
|
|
0
Offline
Tesla Member
Karma: 71
Posts: 6624
Arduino rocks
|
 |
« Reply #3 on: May 07, 2012, 11:21:19 am » |
The schematic isn't wrong, its just not complete - it doesn't address the H-bridge driver chip that is needed (such as HIP4081, FAN7388, etc).
|
|
|
|
|
Logged
|
|
|
|
|
0
Offline
Tesla Member
Karma: 71
Posts: 6624
Arduino rocks
|
 |
« Reply #4 on: May 07, 2012, 11:24:54 am » |
What is the source voltage are you trying to control? Did you read the section on "P- VERSUS N- CHANNEL HIGH-SIDE SWITCHES"? How are you generating the "Source voltage plus five volts" needed to close the high-side N-channel switches?
The MOSFETs in question won't switch at 5V, they need 10V.
|
|
|
|
|
Logged
|
|
|
|
|
Offline
Newbie
Karma: 0
Posts: 24
|
 |
« Reply #5 on: May 08, 2012, 04:40:52 am » |
thanks all for the replies. @Mark - Would i be able to use the same schematic if i give a 10 volts input? How do i give such an input with the arduino?
|
|
|
|
|
Logged
|
|
|
|
|
0
Offline
Tesla Member
Karma: 71
Posts: 6624
Arduino rocks
|
 |
« Reply #6 on: May 08, 2012, 05:45:52 am » |
H-bridge MOSFET driver chips take logic levels and output full 10V gate signals and generate high-side gate supplies too (with external diode and capacitor) They also drive the gates fast enough for efficient high-frequency PWM (which the Arduino cannot really do)
Or you can use a couple of high-side-low-side driver chips. Basically you need 2 high-side drivers and 2 low-side drivers, but the combined high-side-low-side driver usually offers shoot-through prevention (this is definitely worth having). There are many many MOSFET driver chips available, some for complete H-bridge.
For an easy life at just 5V, 2 n-channel and 2 p-channel logic-level MOSFETs will be much easier to drive directly (you have to ensure there is no shoot-through though - get this wrong and 2 MOSFETs may explode - wear eye protection, use a current-limited supply, etc).
---
"Shoot-through" - when the high-side and low-side switching device are turned on at the same time, causing a direct short of the H-bridge supply through the two transistors. If this happens for any length of time the devices tend to melt/explode. If it happens for a fraction of a microsecond on every PWM pulse the devices tend to get very hot.
|
|
|
|
|
Logged
|
|
|
|
|
Offline
Newbie
Karma: 0
Posts: 1
|
 |
« Reply #7 on: November 02, 2012, 08:53:56 am » |
Old, to make this work put 1k resistor1 to Q2 base and 1k resistor2 to Q4 base and join R2 to Q1 base and join R1 to Q3 base, to move forward/back connect Q1 base/Q2 base to input.
|
|
|
|
|
Logged
|
|
|
|
|
|
|
0
Offline
Tesla Member
Karma: 71
Posts: 6624
Arduino rocks
|
 |
« Reply #9 on: November 02, 2012, 10:57:21 am » |
OK, here's my understanding:
Basically the FET equation is: Ids = k/2*(Vgs-Vth)^2
What this means that as the gate voltage rises to the threshold voltage the device starts to conduct, but only a tiny fraction (say a millionth) of its full on current. As the voltage rises higher the drain current should rise as the second power of Vgs-Vth (for an ideal FET).
In practice Vth is measured at a standard small current (250uA for a big power MOSFET), and its a highly variable parameter - you'd never rely on it being a particular value for instance. Its the gate voltage at which the device turns _off_, rather than on - on means fully on for a switching device, note.
Power MOSFETs don't obey the ideal equation by any means though, there is a plateau in the gate voltage as it charges up, shown in every datasheet - this is where the real switching action is. As the drain current ramps up from a few percent of full load up to 100% of full load the gate charge increases for almost no change in gate voltage. The effective drain-source resistance is dropping all the time this happens as the channel gets deeper (the channel is the region of induced space-charge between source and drain through which current can flow. The huge extent of the channel in a power device involves a lot of space charge - and the gate charge has to balance this to allow the channel to grow.)
Typically a normal power MOSFET will have Vth from 2 to 4V, will have a gate plateau voltage around 5 to 6V, and is fully on at 12V (though 10V is usually the voltage used in the datasheet to allow for power supply and driver losses in real designs.)
A logic level MOSFET will have Vth around 0.5V, plateau around 2 to 3V, and fully on at 5 to 6V (quoted at 4.5V often). Its vital with logic-level MOSFETs to drive the gate down to 0V to fully switch it off given the low Vth.
|
|
|
|
|
Logged
|
|
|
|
|
0
Offline
Tesla Member
Karma: 50
Posts: 6546
Arduino rocks
|
 |
« Reply #10 on: November 02, 2012, 11:32:43 am » |
|
|
|
|
|
Logged
|
|
|
|
|
Georgia
Offline
Newbie
Karma: 0
Posts: 28
I get by with lots of help from my friends
|
 |
« Reply #11 on: November 03, 2012, 02:16:23 pm » |
ON my blog evmaker.tumblr.com i have a drive schematic for my motor controller. It is to drive igbt but will also drive mosfet just fine. It is for a 3 phase hex bridge but would wok a H bridge just don't need the extra two drive circuits. The first on that I built ended badly because of a sketch error. The blog shows the hex inverter running my truck have to look back a few pages as the past month is about drive 2.0. The new schematic will be up in a few as I am modifying the way the leds in the optocouplers are driven. I will be hooking the leds of the optos in each half bridge together parallel but opposite polarity as one pin is pwm high the cathode will ground to the opposing pin held low. This will prevent damaging shot through caused by code screw ups. Kind of a play on charlie plexing. The schematic is the drive controler not the H bridge it's self.
|
|
|
|
|
Logged
|
|
|
|
|
|