R/C Tech Forums - View Single Post - keyance zero extreme
View Single Post
Old 11-17-2003, 10:43 AM
  #11  
dpaton
Tech Regular
 
dpaton's Avatar
 
Join Date: Sep 2003
Location: Schaumburg, IL
Posts: 283
Default

Originally posted by Rory
i'm no electronics expert, far from it, but the way i see it there are two ways of thinking about it
1: the lower frequency will give more punch therefore somehow making the esc work harder (not heaps of logic there i know but oh well)
2: the higher frequency will mean the FETs are switching more (or something like that (again im no electronics expert and i'm not sure that FETs do switch but it rings a bell in my head) and would get hotter due to switching more often and of course working harder
I'll toss a little knowledge on this one.

Background:
FETs have 3 operating states: on, off and in between. That in-between state is called the linear region, and is what is used when FETs are designed into audio amplifiers. Most fets are practically lossless when they are all the way on. They ahve a very very low voltage drop, and dissipate a very small amount of current compared to transistors (0.01V or less vs 0.7V respectively). When designing a PWM system, the goal is to maximize power denstiy. You want to move the most charge from one place to another in the smallest box with the least loss posssible. If you need to vary the amount of charge transferred, you pulse the charge from full o to full off to full on, etc. The amount of time each pulse is either on or off will determine the effective DC voltage at the other end. The raw PWM signal isn't very good for driving motors however. It's rapid pulses generate heat and soem ugly magnetic losses when run though a coil of wire spinning in a magnet. That's why we have those big caps on out speedos. They are a special variety of capacitor, designed to run at very high frequencies and smooth out the PWM waveform so it looks like the DC our motors know and love.
Everybody with me? Good. Now on to the frequency stuff:

Lower frequency gives more 'punch' because it's closer to real DC, and thus is filtered much much better. The doen side of that is that it makes acceleration and braking a lot more chunky'. Remember the old Novak T-4? Compare it side by side to a GT7 and you'll notice how much smoother the new ESC is. It's a trade off. Personally, I like about 1500Hz on the gas and 7500hz on the brake, but that's me.

To finally get around to the question...

The higher frequency means the FETs switch more often, yes. When FETs are swiched properly, they do it nearly instantaneously, and go from off to on and back so fast that the time they spend in their linear region is pretty much nil. That's whyv the new generations of speed controls don't need heatsinks fo rmost things, unlike their predecessors. There is basically no power dissipated by a good fet if it's either all the way on or all the way off. If Keyance says higher frequencies generate more heat, I'd suspect it's one of the following possibilities:
1. Someone mis-translated the manual.
2. The FETs aren't being switched on and off fast enough
3. The FETs are very lossy, and dislike being switched.

My gut says it's #1. All else being equal, higher frequency switching is better for getting more amps through a small box without much heat.

-dave
dpaton is offline