mosfet h-bridge design comments

[keplerforever]

Hi,

I'm a newby in electronic design and i would like to submit to you the following design for a h-bridge:

**broken link removed**

I'm planning to use four IRLZ 44 N-Channel logical level gate Mosfet. Each gate would be directly controlled by a TTL level circuit (not on schematics) providing to the h-bridge the following command: bracke, coast, CV, CCV

I would use either shotky diode like 1N5817 or 1N4001/2. Ground of TTL level and motor power supply would be common.

The intent of this design came because I have not been able to find after deep research on the web a Mosfet schematics that would fit my needs: control a 5 to 20V motor consumming maximum 10 A controlled by a PWM at 1 kHz in a very effcient way (ie without transistor). Until now the only schematics I could find was for motor supply above 12V, and in order to keep the design as simple as possible I would like to avoid mosfet driver

My request is therefore: is this design a smoky one, are there some pitfalls I should expect, and by the way is there a simple way to avoid shoot trough?

Thanks in advance for all your remarks and improvments

Olivier

 

[Analog]

Do you know how a MOSFET works? If so, then you would see right away that the top MOSFETs require Vgs + Vl to operate. Therefore you need a charge pump or some kind of driver to get those to turn on, as a simple logic level won't do the trick.

 

[keplerforever]

Thanks for your fast answer.

I do agree with you that i'm not the most knowlegable person about mosfet (I just started to dig into this field a couple of days ago). But after dozzen of schematics reading I got a little bit confused: I'm aware of this voltage accross G and S. But I also saw this kind of schematics in datasheets and in the book: "practical electronics for inventors p177".
My itend is to make the simplet design as possible, and make available to evryone the schematics and the pcb.
So how could I rework this schematics to make it usable. I do not ask you to make the work for me i'm just looking for hints and advices.

 

[Hero999]

The bottom transistors need a gate voltage of >10 to turn on and top transistors will need the gate voltage of 10V greater than the supply voltage to the motor to work. The maximum voltage between the gate and source should never exceed 20V.

 

[Sceadwian]

They make IC's called photovoltaic issolators that are exactly what you need to drive the high side lines of a FET circuit like that as simply as possible. They don't allow for extremely high speed switching but you can also PWM the low side if you need high speed control. They're moderatly expensive. Digi-key list one PVI5080N from International Rectifier for about 5 bucks for a two channel chip. They provide an 8 volt 10uAmp issolated voltage source from a 50ma input. In series their supposed to be rated for 10 volts, which should be more than enough to drive any FET out there.

 

[Roff]

The bottom transistors need a gate voltage of >10 to turn on and top transistors will need the gate voltage of 10V greater than the supply voltage to the motor to work. The maximum voltage between the gate and source should never exceed 20V.

As our OP pointed out, the IRLZ44 is a logic level MOSFET, and will turn on just fine with Vgs=4V (Rds=39 milliohms max). This still means that the gates on the upper transistors must go to at least 9V to turn them on.
Also, the gate to source breakdown voltage is +/-10V, not 20.

 

[i_build_stuff]

The circuit you have is extremely common, and would work fine if you drive it right. There are a few options you might consider:

-Pulse transformers: there are tiny transformers made specifically for driving high-side gates. You won't be able to get 100% duty cycle, but you could use double the normal voltage and cap it at 50%. This is probably not the best option for your application, but it's simple to use if you just want to see something working.

-Carrier drive: use a pulse transformer with some kind of rectifier / resistor / capacitor circuit, and use an AC signal to turn the switches on. This one requires more calculations, but it has a lot of potential when you do it right.

You could also attach a small floating power supply to the high side and use it to power a current-controlled or optically-coupled drive. Optocouplers are easier design-wise, but don't respond very fast (they're good for a couple kHz at most).

Avoiding shoot-through is usually done from the control side, by inserting a little delay between the time when each direction is on. Assuming you're using a digital controller, this is fairly straightforward to do.

One other option, IF your supply voltage is lower than the MOSFET gate-source ratings, is to use P-channel MOSFETs at the high side, and just invert the control signals going to them.

 

[Styx]

As our OP pointed out, the IRLZ44 is a logic level MOSFET, and will turn on just fine with Vgs=4V (Rds=39 milliohms max). This still means that the gates on the upper transistors must go to at least 9V to turn them on.
Also, the gate to source breakdown voltage is +/-10V, not 20.

you do mean the gate of the upper FET to goto 14V don't you?
IF the top-FET was hard-on then its source would be at ~10V (ie the DC-link) and thus for it to be hard-on that gate potential MUST be 4V higher... 14V

@OP simplify it all and change the top two FET's to P-type. valid solution if yr link is <=12V

 

[Sceadwian]

The photo voltaic issolators are probably the easiest sollution. It provides +5 to +8 volts per channel over whatever you connect the outputs Vss to using an input current of 10-50ma's. No muss no fuss no complicated drive circuits. When the input is disabled the issolated output will provide a relativly low impedance to the Vss side discharging the mosfet gate. A two channel 8pin dip costs like 5 bucks.

 

[i_build_stuff]

The photo voltaic issolators are probably the easiest sollution. It provides +5 to +8 volts per channel over whatever you connect the outputs Vss to using an input current of 10-50ma's. No muss no fuss no complicated drive circuits. When the input is disabled the issolated output will provide a relativly low impedance to the Vss side discharging the mosfet gate. A two channel 8pin dip costs like 5 bucks.

I hadn't seen those before, but they do look like they might be useful in some places.

Unfortunately, it also looks like they have horrible timing characteristics (even worse than optocouplers). The datasheets I looked at showed turn-on times from 300us to 2.5ms (!)--way too slow for any realistic PWM.

 

[Sceadwian]

In a large majority of motor applications both sides of the h-bridge don't have to switch. You can use the high side drivers to switch direction and then you just PWM the opposide leg of the H-bridge for motor speed control which could be logic drive to whatever speed you can manage..

 

[Styx]

In a large majority of motor applications both sides of the h-bridge don't have to switch. You can use the high side drivers to switch direction and then you just PWM the opposide leg of the H-bridge for motor speed control which could be logic drive to whatever speed you can manage..

Except a large majority of motor applications is to hook upto Sync machines and thus full PWM of an upper and lower is needed

If you are talking abt 12V motor-drive application then yer

big difference depending on the market

 

[keplerforever]

Thanks to all of you,
After reading all your valuable comments, and digging around I found that a easy solution could be done by using the maxim 620/621 to drive the mosfet