Myth Busters 3 – Myth: “You must have a gate resistor”

[charliesixpack]

I don't waste my time trying to help anyone anymore. All I get is insults, irrelevant "facts" and complaints about how what I said "won't work" (even though I've been doing it for eons).

I've argued your point on this one.  I have designed off chip drivers on ICs and had to consider this very problem for reliability on systems selling hundreds of million dollars.  I can't get upset when some kid living in his mom's basement tells me his intuition proves me wrong.

[gpsmikey]

It's like those green EE kids fresh out of school. They want to power an LED, run the numbers through Ohm's law and then start searching for the proper resistor... to three decimal places, then wonder why they can't seem to find a 121.294 ohm resistor (and if an old timer tells them that it's not so critical and that a 120, a 100 or even a 150 will work just fine), they don't want to hear it. After all, what can a guy who knows what a 5U4 is possibly know about those "silicon thingies"?

I'm so old not only do I know what a 5U4 is, I have an adapter downstairs I built years ago that adapts an 80 to a 5U4 socket !!   

And yes, your comment about a limiting resistor to 3 decimal places is right on.

mikey

[groundFungus]

We had a 7 seg display that showed the integer temperature for a process.  A new engineer came on board and wanted the temperature displayed to 3 decimal places.  We told him that the measurement was not that accurate.  He replied that his model calculated temperature to 3 places.  We added 3 digits to the display and rigged a random number generator to the 3 places.  He was happy.

[Boardburner2]

Krupski, its an old thread

I only revived it to discover what bandwidth scope was used.

I have the chance to play with a 2GHz scope over the next couple of weeks and wondered if it is worth repeating some of the experiments.

[larryd]

It's always worthy if it is free.

Make sure the scope probe has good band width too.


.

[Boardburner2]

It's always worthy if it is free.

Make sure the scope probe has good band width too.


.

It is a rental. We needed a 1 GHz scope for a month.
Got a 2 GHz for the same price (what they had available) with 1 GHz probes, they seem to be missing or locked up though.
Lunchime job i think. Is it worth doing  ?

IT is a tek scope with the funny connectors though.

[larryd]

Personally, I have used ccts. with and without a series resistor over quite a few years.
I wouldn't jump up and down saying a resistor is needed, but if it makes someone feel better to do it their way, all the power to them.

It's always nice to compare things.
Having a high and low quality scope looking at the same signal might be interesting.

OTH a series resistor is dirt cheap, oh what to do? 

.

[krupski]

Krupski, its an old thread

I only revived it to discover what bandwidth scope was used.

I have the chance to play with a 2GHz scope over the next couple of weeks and wondered if it is worth repeating some of the experiments.

Why repeat the experiments? I think they showed reasonable data. In fact, they proved "the point"... yes there is an overcurrent, but it is of such short duration that it simply does not matter.

In fact, adding a gate resistor only serves to slow down the switching time of the power mosfet, leading to extra power dissipation (i.e. heat) which kinda defeats the purpose of using a mosfet.

Also, what nobody seemed to mention (although I may have just missed it) is that the output drivers for the AVR pins, every time they switch, go through a very short period where they are BOTH conducting and placing virtually a short circuit across the Vdd supply (hence the need for bypass capacitors).  Nobody seems to worry about THOSE short duration, high current spikes, but OMG a current spike caused by not using a gate resistor will ruin, degrade or smoke the AVR.

Really??? 

[krupski]

Personally, I have used ccts. with and without a series resistor over quite a few years.
I wouldn't jump up and down saying a resistor is needed, but if it makes someone feel better to do it their way, all the power to them.

It's always nice to compare things.
Having a high and low quality scope looking at the same signal might be interesting.

OTH a series resistor is dirt cheap, oh what to do?  

.

True for the hobbyist, but no engineer will spec a useless component for a mass produced product where the extra 1 cent is multiplied by millions of devices.

[larryd]

"True for the hobbyist, but no engineer will spec a useless component for a mass produced product where the extra 1 cent is multiplied by millions of devices."

A SMD resistor is 0.2 of a cent at 10,000 pieces.

You are correct with this, but we are a home brew kind of place.

Didn't think you were concerned about comerial manufacturing numbers.


Engineers: allow vehicle ignition switches to fail resulting in death, allow faulty air bags manufacture causing death, allow software to circumvent pollution laws, allow a bridge to fall down. 


I just don't care about this damn resistor thing and I think you are there too.
.

[Boardburner2]

Why repeat the experiments? I think they showed reasonable data. In fact, they proved "the point"... yes there is an

Us mortals do not get to play with this sort of equipment normally.

Also, what nobody seemed to mention (although I may have just missed it) is that the output drivers for the AVR pins, every time they switch, go through a very short period where they are BOTH conducting and placing virtually a short circuit across the Vdd supply (hence the need for bypass capacitors).

Shoot through ?
Also spice models would suggest a short high current pulse on the transition.
I suspect breadboard layout problems will make a nonsense of the measurement results anyway.

Resistor may slow down switching but around here i think it best to suggest it as it gives some protection to the outputs from the inevitable silly mistake that the newbies (and others) make.

[gpsmikey]

Us mortals do not get to play with this sort of equipment normally.

Resistor may slow down switching but around here i think it best to suggest it as it gives some protection to the outputs from the inevitable silly mistake that the newbies (and others) make.

Agreed - another point I have not seen discussed is RFI from the switching transients.  Depending on the switching frequency, slowing the rise/fall times does increase the time the FET is in the middle or "linear" region, however, if you are using a low switching frequency (like I have for a heater I control - 2,5 second cycle time), slight heating is not an issue, but by rounding off the edges of the switching, you drop the RFI way down (I hate "tic .. tic .. tic" on a nearby radio on the AM band).  Same thing applies with RS-232 etc - slowing the rise/fall times can give you better distance, but you have to trade that with what baud rate you can work with.  Fast rise/fall times can give all sorts of unwanted ringing depending on the antenna ... I mean wiring 

[alka]

Agreed - another point I have not seen discussed is RFI from the switching transients. 

This is key..

This is the reason that every single chinese speed controller using atmega's built to the lowest possible price point has gate resistors. To slow down the rise /fall time and reduce the voltage spike /  EMI. The resistances used are in the range of 50-100 Ohms so I don't think they care about the current limit.

V(stray) = L * di/dt.

For three phase brushless controllers in a small footprint they are essential.

One other thing .. the gate resistor needs to be large enough to be compatible with the reverse recovery time of the free wheel diode otherwise this can cause overvoltage stress on the mosfet/igbt.

[Smajdalf]

Hi, thanks to reviving this old topic - it was very interesting to read.

But I don't understand one thing: why is everybody so afraid about driving (MOS)FET without gate resistor but does not care about other capacitive loads - such as pins of any other ICs, piezospeakers, long wires...? For example when I connect 10 of anything to SPI bus the CLK will be loaded with similar (or even greater) capacitance than single discrete MOSFET has and is switching with much higher frequency than typical Arduino PWM.

[krupski]

"True for the hobbyist, but no engineer will spec a useless component for a mass produced product where the extra 1 cent is multiplied by millions of devices."

A SMD resistor is 0.2 of a cent at 10,000 pieces.

You are correct with this, but we are a home brew kind of place.

Didn't think you were concerned about comerial manufacturing numbers.


Engineers: allow vehicle ignition switches to fail resulting in death, allow faulty air bags manufacture causing death, allow software to circumvent pollution laws, allow a bridge to fall down.  


I just don't care about this damn resistor thing and I think you are there too.
.

Even at the hobbyist level, I would not use a gate resistor because:

• It's not necessary
• It consumes board space
• It costs me 3 cents
• It consumes my time to solder it onto the board
• It slows down the switching time of the mosfet, leading to increased power dissipation

With regard to the part of your quote that I bolded... agree completely. We're beating a dead (and now decomposing) horse to death.

As of this post, I'm done with this thread.