A word of warning about mosfet gate drive voltage: to turn a mosfet fully on you must be more than a threshold voltage above BOTH the source AND the drain - otherwise the channel will not be inverted over the entire length and the FET will be operating in constant current mode instead of constant resistance. If the Rds_on is low and the current is fairly low, this is usually a non-issue, but if you are switching substantial current, it can be a problem as the drain will be a volt or two higher than the source in the fully-on condition. Also, the faster you can take the FET through the partially-on range, the lower the switching power dissipation will be. Generally this requires a voltage at least several volts above the threshold voltage because a) most of the current during the actual transition from off to on will be going to charge the "Miller capacitance" and you need some to still go into the actual MOS capacitor of the gate-source/substrate and b) the source lead inductance will combine with the very high di/dt to steal a volt or two from your gate drive voltage if you consider the actual voltage that the MOS capacitor sees inside the FET package. This means that usually FETs which are rated for Vgs_max of 20V are driven with 12V to 15V, and almost never less than 10V. Most gate drive ICs will go into an undervoltage lockout below 7 or 8V. Why do you care so much about how large the final gate charge is? Sean On Fri, May 24, 2013 at 6:26 PM, veegee wrote: > On 2013-05-24 4:21 PM, RussellMc wrote: > > What they say, but ... > > > > 6V gate drive is getting low if you can tolerate 18V. > > The datasheet for the MOSFETs I'm using shows that they fully turn on at > 6V, and still more than enough at 5V. I also don't want to charge the > MOSFET gate to high voltages if I can avoid it since the gate charge > increases linearly without lowering the Rds any noticeable amount further= .. > > > The "whole regulator" is worth tens of cents and you are supplying it > > not "just for the gate driver but "just for the MOSFETS and whatever > > they are doing". Is the take really so value-less? > > Even when required capacitors are added, cost is unlikely to be a good > > reason for doing this. Size may be. > > You're right, I've decided to use a linear regulator. > > > If you use an 18V zener (less whatever safety margin is required) then > > it needs to supply (just enough current at 16V so you'll be > > dissipating Imax x (2-16) =3D 8 x I_max Watts. > > A 10V zener in series will dissipate up to 10 x Imax Watts at any > > Voltage. I'ts more complex than that but essentially you'll disispate > > about the same either way and have ~=3D 16V min gate drive using a > > conventional zener. > > I don't think I could use a Zener in parallel for a gate driver IC > supply. The gate driver and the average current to switch the gate will > be below 100mA, but I'm relying on high pulse current when switching to > charge the MOSFET gate fast enough. I could use a big capacitor in > parallel with the Zener, but then again, I might as well use a 1.5A > linear regulator and be done with it. So that's what I'll do. > > -- > http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive > View/change your membership options at > http://mailman.mit.edu/mailman/listinfo/piclist > --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .