Author Topic: Power MOSFET questions, why use an H bridge for an amp for example?  (Read 1793 times)

0 Members and 1 Guest are viewing this topic.

Offline Chris WilsonTopic starter

  • Super Contributor
  • ***
  • Posts: 1081
  • Country: gb
  • Race car engineer, dog lover, hoarder.
I am into LF amateur radio communications, and have built a 1kW Class D amp for none linear digital transmissions around 136kHz. The big power is needed due to the gross inefficiency of any antenna a normal citizen is likely ever to be able to erect, size and height wise. Even 1kW gives well under a Watt of ERP...

My amp uses four MOSFETS, two in parallel each side driven by a dedicated driver chip. I see the odd LF amp design using what's described as a H Bridge MOSFET array. The one I most recall used rectified 240 volt single phase mains to run it. My question is why use an H Bridge rather than paralleling more MOSFETS in a conventional format? Does it overcome voltage or current limitations? Is it due to adding more capacitance with paralleled devices?Does it allow a linear amp to be built?

I realize a meaningful reply might be quite lengthy, so I hasten to add that have Googled it, but most stuff related to industrial motor control.

Secondly, in a Class D none linear amp, how is the gate resistor value chosen, and is there ever a case for not using one at all?

Thirdly, am I alone in thinking the terms Drain and Source seem back to front and the names should intuitively be reversed so a Drain was called  a Source and vice versa? :)
« Last Edit: November 25, 2016, 12:16:36 pm by Chris Wilson »
Best regards,

                 Chris Wilson.
 

Offline danadak

  • Super Contributor
  • ***
  • Posts: 1875
  • Country: us
  • Reactor Operator SSN-583, Retired EE
Re: Power MOSFET questions, why use an H bridge for an amp for example?
« Reply #1 on: November 25, 2016, 01:58:33 pm »
Code: [Select]
Thirdly, am I alone in thinking the terms Drain and Source seem back to front and the names should intuitively be reversed so a Drain was called  a Source and vice versa? :)
Conventional vs Electron flow nomenclature ?

https://www.mi.mun.ca/users/cchaulk/eltk1100/ivse/ivse.htm


Regards, Dana.
Love Cypress PSOC, ATTiny, Bit Slice, OpAmps, Oscilloscopes, and Analog Gurus like Pease, Miller, Widlar, Dobkin, obsessed with being an engineer
 

Offline danadak

  • Super Contributor
  • ***
  • Posts: 1875
  • Country: us
  • Reactor Operator SSN-583, Retired EE
Re: Power MOSFET questions, why use an H bridge for an amp for example?
« Reply #2 on: November 25, 2016, 02:02:54 pm »
An H Bridge doubles the pk-pk V swing across the load for a given
supply voltage. Hence quad the power in the load.

This is an oversimplification, but basic result. Frequently used in Audio
work for LV amplifier applications.


Regards, Dana.
« Last Edit: November 25, 2016, 02:09:52 pm by danadak »
Love Cypress PSOC, ATTiny, Bit Slice, OpAmps, Oscilloscopes, and Analog Gurus like Pease, Miller, Widlar, Dobkin, obsessed with being an engineer
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21686
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Power MOSFET questions, why use an H bridge for an amp for example?
« Reply #3 on: November 25, 2016, 02:10:02 pm »
Your old circuit has a push-pull configuration, which is very inefficient due to the transformer.  This incurs a minimum propagation delay through the transformer, which inevitably means power lost through reactive (switching) energy.

No transformer has zero leakage and zero capacitance, so it is unavoidable.  One can make a very good transformer, but one cannot eliminate this aspect of the design.

Half bridge is the next best, as it doesn't require a transformer.  You still incur the physical length between transistors (and therefore equivalent inductance and capacitance), but this can be much shorter than the windings of a transformer.

When the stray inductance is smaller than the circuit's voltage, current and speed (when expressed as an inductance: 1 H == 1 V.s/A), the circuit will be overdamped, and will not exhibit significant overshoot.  In this case, the stray inductance is small enough that it can be ignored.

Because old transistors were limited on speed (perhaps 100 or 200ns switching time), old advice was to minimize inductance, to achieve this condition.  Modern transistors are almost never operated so slowly, and therefore this advice is obsolete.

You can still maintain that condition, but be mindful of where it came from, and what the consequences are.  Don't just parrot obsolete advice!

The downside to a half-bridge is, it still needs a center-tap.  Instead of a transformer (that PP requires), it's the power supply.  For an AC application, this is normally a capacitor divider.  The load is cap-coupled from the half bridge, to ground (but not just ground as such, but +V as well, for symmetry).

For a DC application (like a class D audio amplifier), a bipolar supply is required.

Or, we can double the circuit, so the load is driven by positive and negative voltages simultaneously.  Now the current always flows straight through the supply, with no center tap needed.  Downside?  Twice the transistors.

The actual amount of silicon required is the same, for a given power level.  The half bridge delivers half the (total) supply voltage to the load, so requires twice the current for the same power output.  Full bridge delivers full voltage and full current, with the downside that two transistors are connected in series at any given time.

Most power supplies in the 1kW range (which includes your transmitter, as it's nothing more than a switching power supply, with an RF filter on the output, instead of a rectifier) use H-bridge, because the cost of gate drivers is fairly modest, while the bulky coupling capacitors are eliminated (or at least, very reduced).  Other considerations (supply bypass capacitors, output transformer, etc.) remain constant.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline Chris WilsonTopic starter

  • Super Contributor
  • ***
  • Posts: 1081
  • Country: gb
  • Race car engineer, dog lover, hoarder.
Re: Power MOSFET questions, why use an H bridge for an amp for example?
« Reply #4 on: November 27, 2016, 02:03:56 pm »
Great info from all the contributors, thanks for the detailed replies and your trouble, I now have a much better understanding of why the H Bridge could be a lot more efficient, again, thank you all very much indeed!
Best regards,

                 Chris Wilson.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf