2ctiby

modified 9 years ago

CMOS shoot through....explained below

141

03:10:31

Look at the circuit on the right....

The two mosfets working in a complementary way together is called CMOS.

In this setup here I have supplied the gates of the top and bottom mosfet there so that the two gates are never conducting at the same time. If they were to conduct at the same time, then current could flow from top to bottom and in a real circuit that may well blow a component....That is called 'shoot-through'.

If you look hard enough, you will see that the top pmos closes just before the bottom nmos starts to open.
You will also see that the bottom nmos then closes just before the top pmos starts to open.
There is no overlap........ The bottom ON duration is enclosed within the top OFF duration.
You can see that on the scope if you study it for a while.

Some may find it easier to visualise by looking at the two bottom left lights where the one on the left is nested within the one next to it.

When gates are rapidly moving in and out it is easy to miss the shoot-through which may be occuring, but it is important.

We have the opportunity on EC to simulate mosfet gate activity......

In the very top left mosfet settings..... turn the VTO from my default -8v up to -10v 
See the LED blowing at the bottom of the right circuit.
Look at the scope traces, then turn that VTO right down and look at the nicely nested traces...no shoot-through.

The trouble with too much gap time is that circuits can be sluggish, so aim to keep these timings tightly nested but without shoot-through. Here is some more help with adjusting mosfet gate timing activity: 

See here  http://everycircuit.com/circuit/5521145715752960   

and here  http://everycircuit.com/circuit/6676866922708992

published 9 years ago

hurz

9 years ago

Messed a little your circuit to show you are far away from presenting a solution. Dont worry. Dont be too embarrassed. http://everycircuit.com/circuit/5184187747532800

2ctiby

9 years ago

Your pointless messing is complete nonsense... You obviously have no idea about shoot-through as can be seen on your mosfet circuits. You are a deluded idiot Hurz.

hurz

9 years ago

As normal, you are lost and dont know how to solve the issue. Keep on learning at start from the basics.

2ctiby

9 years ago

Any sensible reader will now see your childish stupidity in areas which you know nothing about, such as this.... You are finally showing just how deluded you are. It's time to take a dose of your own medicine...You are devoid of stating points of reason here because the concepts are beyond you....Go ahead state some valid point here...we are waiting.

hurz

9 years ago

As normal, anything you have to tell the EC community which is about electronic, or are you just a jerk? I like my demonstration of your circuit. Its one of the best shoot through you build there.

nikisalli

9 years ago

Hurz is right. As you can see from his modifications on your circuit there is a peak of current that means for a moment your mosfets are active at the same time

nikisalli

9 years ago

And he's not a deluded idiot. He is very stubborn

2ctiby

9 years ago

@nikisalli We as a community have avoided conflict with you for a long time because we have seen that you are a child who needs a father figure and have taken Hurz for that role. It is time however, that you see faìr reason rather than following a delusional mentor. ...Look for truth.

2ctiby

9 years ago

@Hurz We await academic discussion instead of your dillusional side-tracking insults. You are washed up now and deservedly so.

hurz

9 years ago

I just watch the show you cellebrate here. Circuit + description doesn't fit together as I have demonstrated, which is for a nasty person like you much more then you should expect. Actually I did that for the part of community who respect and like my knowledge in electronic. Now close the door and sit down and use your brain for the technical problem you like to solve. Again start at the basics you do not know and do not start with semiconductors. At least not to many on one worksheet. And better stop to act like grumpy granny.

2ctiby

9 years ago

Stupid reply Hurz....still waiting for an academic discussion on any relevant point here.. which point do you wish to make on this presentation of mine?

nikisalli

9 years ago

Why you two have to make this comedy here. Hurz is right cuz i see a bad current spike. If hurz was wrong i would have said "hurz you're wrong here cuz there's no current spike"

hurz

9 years ago

Niki seems to understand that. Isn't that much enough motivation for you to think hard about? BTW, what is your problem in non academic discussions and persons. You hate kids and even try to threat them? You hate unstudied people. They are scum for you? Learn to love and repect humans first before learn about electronic.

2ctiby

9 years ago

@ nikisalli You have not understood the whole concept of what I was saying...read it all carefully again...good luck.

2ctiby

9 years ago

@ Hurz You are a peculiar idiot...academic question now please about the original presentation rather than your usual diversions

nikisalli

9 years ago

I read it all again and from what i see hurz gave you a motivation on why your circuit doesn't fit in your description and you started to insult him. I can give an explanation too on why he is right: you are basically driving your mosfets with a pulse longer then the other But with two pulses like that your fets will blow up! To protect your fets from shoot through you have to phase shift two pulses not making them one longer then the other! Here's the solution: http://everycircuit.com/circuit/6082951060389888

hurz

9 years ago

Good boy 👍 😘

2ctiby

9 years ago

@nikisalli ... Now this is more like it...well done you are considering possibilities and presenting sensible arguments without pesonal hatred etc.... Your presentation is very valid in my opinion...your bottom mosfet has a different time delay to achieve the required non-shootthrough and that is an admirable method. Now consider the case where we cannot start with two separate pulses like yours...My presentation starts with just one source which then splits that in to two different voltage limits and separates the gate operation accordingly to achieve a similar non-shoothrough situation. The key point being that one way or another, we need to find suitable methods to stop shootthrough rather than being oblivious to it (my presentation only allows a controlled shootthrough at certain specified voltages by the way). Well done...this is what EC discussion should be about, without hostility.

nikisalli

9 years ago

Now help me with my insect's buck converter 😜

2ctiby

9 years ago

Wish I could niki, but my electronics knowledge is very limited and certainly less than your advanced presentations.....we all have strong and weak areas. I have so much to learn, but I can try and help with mosfets as best I can.

nikisalli

9 years ago

@2ctiby i have a method that is often used in half bridge smps like atx from computers with ics like ka7500 or tl494. I recostructed their internal circuit here: http://everycircuit.com/circuit/6656988078931968

nikisalli

9 years ago

@2ctiby i was talking with hurz that is helping me with a buck converter for my hexapod robot

2ctiby

9 years ago

That looks so good with the neatly placed on/off impulses. I started to get in to these logic gates etc last year, but I left them off and now I am so behind when it comes to things involving them. Perhaps you could help me if I get back in to them sometime. I can certainly see their usefulness replacing transistors as switches, but they are a bit restricted to on/off compared to the adaptability of voltage/current decisions available with transistors do you think?

nikisalli

9 years ago

Obviously logic parts does not have voltage/current capabilities that transistors have but their good to use from time to time and you can find ics with all of them in place like ka7500 and tl494 in this case to solve shoot through problem

2ctiby

9 years ago

Yes, and getting back to my original presentation...if you really want to get rid of the insignificant few remaining micoamps shootthrough, then highlight the resistor there on the right to see that on the scope...then turn the top mosfet VTO (the one just to the right of the top left) down from -9 to -8 but you may still need to play around with fine tuning to get rid of the last microamp if you wish....Whatever shows in the Hurz link is not significant because that circuit has been changed from mine....My demonstration was dependent on the changing voltage of the ac at the left....Hurz has removed that. Hopefully you now have a better awareness of shootthrough and methods of adjusting for it...which was the aim of my presentation.

nikisalli

9 years ago

The method you're using is wrong. You don't have to use a Sine wave. And anyways this is not practical because in a real circuit you don't have fets with adjustable VTO and you don't have sine waves

nikisalli

9 years ago

As i said before with your method you make one pulse longer then the other but you need two phase shifted pulses not two pulses one longer then the other

2ctiby

9 years ago

I have already agreed that your time delay of pulses is a good method for avoiding shootthrough (although you can't link two cmos gates together in that case), but my main point with this demonstration is to show that VTO variations can produce shootthrough due to overlap. Some ICs have a built in design to prevent shootthrough happening, but with EC we need to be aware of it happening when designing our CMOS circuits, otherwise we may build the circuit using our mosfet schematic only to find that a component gets blown. This presentation is just to draw attention to and briefly describe shootthrough, it does not consider other timing issues of gate charge and pull-off delay etc.

hurz

9 years ago

[BLOCKED]

add comment in editor