using mosfets for H bridge a few questions

[large_ghostman]

hi
i want to use some mosfets for a H bridge to drive some small motors <1A i want to try and use what we have already the H bridge will be switched on and off by a 5v pic probaly using pwm what figures do i need to look for in the datasheets to find a suitable one? i have done some reading and looks like i need a min voltage on the gate to switch them on but where do i find this figure in the datasheets? i dont have part numbers yet as i have found a few diff ones in the box:d so wanted to look up the datasheets but not sure what bits i need to know
thank you lg

 

[MrAl]

Hi,


If you look up the data sheet for one of the part numbers you'll see a gate voltage spec. If you cant find it, post the data sheet and we'll take a look. Once you find one you'll know how to do it with all of them.

 

[crutschow]

The value you are looking for is the ON resistance at a given source-drain current. The gate-source voltage (Vgs) required to achieve this tells you the voltage required to fully turn on the device. Standard MOSFETs require 10V but Logic Level types require only 5V or less.

There is also the Absolute Maximum Vgs which must never be exceeded, and the Vgs threshold voltage, which is the value at which the transistor is just starting to turn on.

 

[large_ghostman]

thank you for that ok for an example take a BF244B as an example found here http://www.datasheetcatalog.com/datasheets_pdf/B/F/2/4/BF244B.shtml
how do i find the values for this? by the way i know it says RF amplifier so probaly wouldnt work but i had the datasheet to hand

 

[MrAl]

Hi LG,

That's not a MOSFET. Do a search for IRF511 or something like that and see what it says for gate voltage.
You'll see plus or minus 20 volts max, but further down you'll see RDS(on) which shows
the gate voltage as 10 volts under "Test Conditions".

 

[large_ghostman]

The value you are looking for is the ON resistance at a given source-drain current. The gate-source voltage (Vgs) required to achieve this tells you the voltage required to fully turn on the device. Standard MOSFETs require 10V but Logic Level types require only 5V or less.

There is also the Absolute Maximum Vgs which must never be exceeded, and the Vgs threshold voltage, which is the value at which the transistor is just starting to turn on.

LOL ok thank you for telling me! there was me thinking i was smart i have downloaded the sheet you sugested and tommorow i will go hunting in the workshop and see what we have thank you lg

Hi LG,

That's not a MOSFET. Do a search for IRF511 or something like that and see what it says for gate voltage.
You'll see plus or minus 20 volts max, but further down you'll see RDS(on) which shows
the gate voltage as 10 volts under "Test Conditions".

 

[MrAl]

Hi again,


Most MOSFETs require 10 volts gate drive, but the logic level MOSFETs only require 5 volts. That's the short answer.

 

[large_ghostman]

hi if most need 10v can i use 12v and use a small transistor to turn a mosfet gate on?

 

[MrAl]

Hi,

You should be able to use 12v on most MOSFETs unless they are the logic type then you might want to check the data sheet.
They usually spec at 10v because that's considered the min drive voltage for many MOSFETs, although they usually take up to 20v.
12v should be a good level.

 

[wagnerlip]

So much important as to turn on the Power MosFET (charge the gate) with 10 to 12V and fast current (think about 100mA or more), is also the fast discharge when you want to cut off the MosFET. Any slow charge or discharge will make the transistor go into the linear region of the condution, and it will heat.

This is why people develop several solutions with one or two transistors to charge/discharge the gate, or even using a Power MosFET driver.

No kidding, you can destroy a Power MosFET easily if going constantly into the linear region with a high current load, due high temperature in the junction.

We also use a zener diode between gate and source to limit charging voltage on the gate. It is easy to destroy the gate with excessive voltage. Sometimes 18V between Gate and Source can destroy a transistor designed to work with a maximum of 12VGS.

The following is a simple N Channel driver, but it lacks the good discharge transistor, it uses a simple resistor. To speed up the discharge, R4 must be a low value, no higher than 150 or 180Ω, by itself it will constantly drain current from R3 and Q2 when it is feeding the gate. Also, for Q2 to be able to supply almost all +VBat to the gate, R3 must be much lower than R4, so, lots of current is wasted in this driver when feeding the MosFET's gate. As you can see, a zener diode is used to protect the gate against high VGS.

View attachment 64037

The following circuit uses a medium power MosFET (IRF510), but look at the R1 resistor, 10kΩ, when Q2 is not conducting, the maximum peak gate charging current is 1.2mA, this is ridiculous, the poor IRF510 will work in the linear region all the time, and will require a large heatsink. By the other side, Q2 does a pretty good job to discharge the gate, since the 2N3903 can sink hundreds of mA at once from the gate. The IRF510 has a medium "virtual capacitor" at the gate, it can store quite good charge. For this circuito to work nicely, R1 should allow 80 to 100mA flowing from +12V to the gate. This would only be possible, if R1 would be from 50 to 100Ω, but no more than that. Of course, when Q2 conducts, it will not only discharge the IRF510 gate capacitor, but will also sucks current from +12V via R1, so make sure to not exagerate. If R1 = 100Ω, Q2 will sink 120mA, with a possible peak of 400mA or more, based on the IRF510 gate capacitor charged. During the sinking, Q2 may dissipate from 100mW to 250mW, be careful to not use lower values at R1.

View attachment 64038

Below a terrible configuration for a Power MosFET driver, in totem-pole configuration.
Due Q1 being a NPN, the MosFET gate will be loaded with the input pulse voltage, less 0.6V, that's nasty.
If the MosFET requires 10 to 12VGS, the input pulse at the bases of the bipolar transistores must be at least that voltage plus 0.6V.
It doesn't make any sense use such configuration when driving the transistors with a microcontroller, powered by 5V, and the Power MosFET requiring more than 10V at the gate.

View attachment 64039

The following shows a better way to do it, when input is zero, Q1 is not conducting, R1 + R2 feeds current to the Q3's base, it conducts and feeds almost +VCC to the MosFET's gate. When input is high or floating, Q1 conducts, connecting R2 to ground, cutting off Q3 and turning on Q2, that will discharge MosFET's gate fast. This works nicely. R3 in the circuit is to avoid "ringing waves" between the transistors and the MosFET's gate, it must be a value 10Ω or lower.

View attachment 64040

Below another driver using NPN + PNP to drive the Power MosFET's gate. Note the use of reversed NPN (bottom) and PNP (top). The left transistor, a NPN, works as a level translator. Its base is connected to the microcontroller's VCC, and the control is done at the NPN emitter. This works very, very well when voltages feeding the NPN+PNP pair and the base resistor is way above the microcontroller logic VCC. When the input is zero, the NPN conducts, since VBE is practically VCC, then PNP conducts and feeds the maximum possible current and voltage to the MosFET's gate. When the input is high, close to the microcontroller VCC, the NPN cuts off, now is the NPN of the pair that conducts and sinks all the MosFET gate's charged capacitor. This is one of the best and fast drive, but, with a price, it doesn't work very well for high frequencies, when the square wave starts to looks like round in the corners, and the transistors pair get lazy and do not switch fast enough, feeding linear voltage to the gate, MosFET will heat.

View attachment 64041

The following is one of the best drivers you can built with discrete components.
When the input is high (above 3V), the 3904 conducts, also the 3906 conducts. This will load the MosFET gate's capacitor very fast, via the 1N4148 diode. As the VBE of the bottom 3906 is reversed (more positive at its base when compared to its emitter, caused by the voltage drop over the 1N4148 diode), the transistor does not conduct. The MosFET's VGS is limited by the 12V zener. You could include a 10Ω resistor between the 1N4148 and the zener to ensure the zener will not blow up if +V is way above it. When the input is low, the 3904 will cut off, the 3906 will cut off as well, now there is no more current to charge the MosFET's gate, but the MosFET gate's capacitor still charged with +12V. Now this gate's charge can't not go back via the 1N4148 diode, so the base of the bottom 3906 is at the ground level via the 1k5 resistor, while its emitter is at +12V from the MosFET gate's charge voltage. It makes the bottom 3906 conducts and discharge the gate in high speed and current. This is a fast self discharge system and a very effective way to drive the Power MosFET, at the cost of several components, but believe me, it worths build it.

View attachment 64042

 

[large_ghostman]

thank you wagnerlip i very much appreciate all the information! how would i configure it to a H bridge? any sugestions greatfuly recieved i am off school again this week with bad tonsils so i thought i might spend the time getting the motors working hope you dont mind if i ask more questions as i think of them
thank you very much lg