|
N Mosfet switch ....How to set the Rds₍ₒₙ₎ in EC
The schematic looks to be working nicely.....but it's not:
If you place an ordinary switch in EC there are no settings for it...you don't need any, but some components do need settings, such as a transformer...you would not ever think about placing the default transformer in to EC to work as you want without altering it. Unfortunately, many EC users do just that with a Mosfet. (and I did...but we can learn).
In an earlier article, I described how to set the Transconductance using the KP mosfet setting and how to alter the VTO for Vgs₍ₜₕ₎ http://everycircuit.com/circuit/6359306484318208 and http://everycircuit.com/circuit/6653380093280256
The Width, Length and Lambda settings also alter the mosfet gate bar action in EC which effects the transconductance extra to the KP. The Lambda can be thought of as a fine tuning to the channel Length modulation http://everycircuit.com/circuit/6148524605702144 ....in most cases you may never use it, so I suggest that you leave it and never think about it again for now.
Turning the Width clockwise as far as possible is a good starting position to get the gate bar working in EC and turning the Length anti-clockwise is likewise the best starting position. The KP is also best started fully clockwise. Turn VTO to 3V as a good starting position. (or 2v if you have a Vgs of 3.3 as with a pi)
Now whenever you come to use a N mosfet in EC with those adjustments, the gate bar is most likely to move fully in and out sharply....which is usually what you need to happen. You can adjust any of those for advanced circuits in an effort to get the gate bar working fully as a switch.
When you buy a mosfet from a shop, you can not alter those settings....the alterations here in EC are an aid to getting that bar moving fully under differing conditions...eg to simulate different Vgs requirements etc.
If your mosfet gate is only moving half out or stuck or sluggish, then it probably needs altering to work fully....That is the object in EC rather than thinking about any specific setting value. If you fail to get it moving fully, then your current flow is not likely to be realistic. If the gate bar has a sluggish turn-off then review your pull-down resistor.
With that in mind you can then happily view your mosfet on EC knowing that your shown current flow is showing the action of a real mosfet. A real mosfet has a tiny resistance built in which by the nature of V=IR and P=IV means that wasted heat power will be produced there. There is no 'real heat' produced on EC, so that resistance Rds₍ₒₙ₎ is not really needed on EC....instead, just aim to get the gate bar working fully with max current flow.
Although the Rds₍ₒₙ₎ is not important in EC, it is important in a real mosfet. We need to calculate how much heat will be dissipated and know when a heat sink is required. For interest sake, we can play with creating a mosfet Rds₍ₒₙ₎ which can help us to understand what is happening visually and give us practice with the real heatsink calculation....just remember, we don't need it on EC...better to aim for max current flow with good gate bar movement, and use a calculation for real heat.
Now look at the schematic...
The load resistor is 8Ω, the supply for that is 10v, so the current should ideally be 10/8 = 1.25A
But see now the current is shown as 1A....that wastage is too great.
Tap on the wire just above the mosfet drain and now see the volt reading at the bottom line of oscilloscope.
See how there is a 2v drop across the mosfet (drain source). 2v / 1A = 2Ω ie there is a hidden 2Ω resistance in the mosfet.....we can call this our Rds₍ₒₙ₎ (R_esistance at d_rain s_ource whilst turned on)
Real mosfets have an Rds₍ₒₙ₎ of around 0.05Ω as per its datasheet, so let's alter this mosfet to that:
Look at the schematic again....it is really a simple resistance divider with R1=8 and R2=0.05 cp R2=2v above.
If we call the mid point voltage Vm, then Vm=[(R2/(R1+R2)) x (Vsupply)]
So Vm=(0.05/8.05) * 10 ie Vm=0.0621v ie 62.1mV is the required volt drop across the mosfet.
Alter the KP mosfet setting carefully whilst looking at the drain wire volts (to source zero) on the bottom left of the oscilloscope...(forget the 10v seen alongside...that relates to the Rds₍off₎ which we have no interest in here). When it holds at 62.1mV you are done. You have now set that hidden Rds₍ₒₙ₎ to 0.05Ω and simulated your datasheet spec. As a double check....V=IR so 0.0621/1.24 = 0.05Ω where the current through load and drain source is now seen as ≈ 1.24A if you take a look. If you can't turn the KP dial any further, then use the Width or Length dials....it's the bar movement we are concerned with remember, not any specific dial value for this.
So there we have it....the mosfet bar looking good, a full current flow and a Rds₍ₒₙ₎ set to whatever we wish.
Now alter the supply to any volts you wish, and likewise the load R1......use the same method to create a Rds₍ₒₙ₎ of 0.06 or whatever your datasheet states.....good messing....keep an eye on that gate bar movement.
Now see here for real mosfet heatsink decision http://everycircuit.com/circuit/5293393395843072
|
