Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!


  1. Jul 5, 2012 #1
    I'm a bit of a noob on all of this, but can anyone help me (or point to a great reference resource) on pro's and con's of each in designing a power management circuit?

    Specifically, I'm looking for what kinds of voltage and frequency applications each manufacturing process is used nowadays... I.e. LDMOS power amps are heavily used in wireless infrastructure systems due to their higher operational frequency (1GHz+)... How do bipolar and CMOS devices compare?

    the rise of BCDMOS is particularly interesting - if I understand it right, LDMOS here is used for high-rel power, bipolar is for higher accuracy analog functionality (voltage reference?), and CMOS is just for general integration.

    while the general trends are pretty well known (CMOS for logic and some analog, LDMOS for basestation PAs, GaAs for handset PAs, bipolar for large-scale applications), just would like some professional opinion..
  2. jcsd
  3. Jul 5, 2012 #2
    I am no expert in LDMOS nor did I use it before. From my understanding, one important feature of MOS transistor is it's output and input capacitance.

    In RF matching, it is very common to use LC matching network where a capacitor is used in parallel with the input and output of the transistor for impedance matching and to form a moderate Q filter. Actually a lot of the big RF power transistors have chip capacitor built into the transistor, it is no longer a monolithic device in any stretch. They are actually a hybrid circuit with the transistor in the middle and a lot of capacitors in parallel. The cap and the bonding wires form the LC network for the last stage of impedance matching. If you look in RF Power Amp for Wireless Communication by Steven C. Cripps page 79 to 81, there is a good explanation how it become the last part of the matching network.

    MOS transistor gives you the free input and output capacitor!!! There might be other reason that I don't know of, but this is one big reason already, chip and wire as described above is very expensive to make.

    When come to CMOS logic, MOSFET inherently is much superior in speed compare to BJT. The main thing is MOSFET don't have saturation problem like BJT, that's the killer for BJT logic. All the schotky don't quite do it.

    I am not expert, I think the main problem with the MOSFET is the input capacitance that impose heavy drive requirement. In the older days they couldn't make the MOSFET small enough and the input capacitance is very high. Now, they can make very small CMOS and the capacitance is much lower, they can make very fast logic circuit. Also, the structure of the MOSFET is much simpler that require less process to fab on the chip. They are basically something like a rectangular block with an oxide cover on top then a metal gate. BJT need more processes to fab the base and emitter. That's money.

    As for precision analog circuits, BJT is hands down better. It Vbe is much more predictable compare to the gate Vth of MOSFET and the drift is much bigger. I did circuit design that require very stable timing and I had to use all BJT devices. The big disadvantage of MOSFET, MESFET and the like is they all have high flicker noise at low frequency. It is just very not useful in low speed precision low noise circuit design.
    Last edited: Jul 5, 2012
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook