In the subthreshold region, gm = Id/mVT where m is the body-effect parameter (it is not the ideality factor n of diode I-V equation). 'm' is defined as the m = Delta(Vg)/Delta(Surface Potential), which may be viewed as the Incremental gate voltage that must be applied in order to produce the incremental change in the surface potential. m = 1.2 - 1.4. We control the surface potential to control the inversion charge. m is like the actual price we pay (i.e., gate voltage) to buy a product (surface potential). The price we pay includes a 20-40% sales tax, LOL (m=1.2 to 1.4)
Say, @14:00, why'd you take 2Av|Vin|? Why not using just Av|Vin|? Because that'd be the excess over the DC at the output node....................................... OR IS IT?
It has been a while but I am trying to make sure that VDD is big enough that the amplifier will not clip the signal. IF Vin is given in amplitude rather than peak to peak one would need to multiply it by 2 to get the voltage over dead you need. That or I put 2\times to make sure I had extra headroom. I lost my original notes in a computer mishap.
Here isa video about an opamp desing. Not exactly what you wanted though, but maybe it will give you some ideas:ua-cam.com/video/I2aOC0fcI_g/v-deo.html
Here is a circuit with the spice model. The important part is set the model level to 12 which is the EKV model. IT is for a 2n7000 that I extracted myslef:www.dropbox.com/s/fp9tk8l6nf9as8s/vds_ekv7.asc?dl=0
I followed this article:ekv.epfl.ch/files/content/sites/ekv/files/pdf/icmts96.pdf and made an Lt-spice model of it (Level 11 I think) The model that comes with ltspice for that transistor does not have threshold conduction and so will not work.
In the subthreshold region, gm = Id/mVT where m is the body-effect parameter (it is not the ideality factor n of diode I-V equation). 'm' is defined as the m = Delta(Vg)/Delta(Surface Potential), which may be viewed as the Incremental gate voltage that must be applied in order to produce the incremental change in the surface potential. m = 1.2 - 1.4. We control the surface potential to control the inversion charge. m is like the actual price we pay (i.e., gate voltage) to buy a product (surface potential). The price we pay includes a 20-40% sales tax, LOL (m=1.2 to 1.4)
Say, @14:00, why'd you take 2Av|Vin|?
Why not using just Av|Vin|? Because that'd be the excess over the DC at the output node....................................... OR IS IT?
It has been a while but I am trying to make sure that VDD is big enough that the amplifier will not clip the signal. IF Vin is given in amplitude rather than peak to peak one would need to multiply it by 2 to get the voltage over dead you need. That or I put 2\times to make sure I had extra headroom. I lost my original notes in a computer mishap.
@@davidwparentmeans you memorize things?
A very nice video, thank you for your work!
And sure I would love to see more!
ua-cam.com/video/I2aOC0fcI_g/v-deo.html
Hello sir, how the term 2AvVin is coming ?...
It works for a Folded Cascode Design?
Sorry for the slow response. It works for everything really. I do not have an example though.
Here isa video about an opamp desing. Not exactly what you wanted though, but maybe it will give you some ideas:ua-cam.com/video/I2aOC0fcI_g/v-deo.html
Hi. Thanks! Do you have the spice model?
Here is a circuit with the spice model. The important part is set the model level to 12 which is the EKV model. IT is for a 2n7000 that I extracted myslef:www.dropbox.com/s/fp9tk8l6nf9as8s/vds_ekv7.asc?dl=0
which formula are you using to find the Is and N ?
I followed this article:ekv.epfl.ch/files/content/sites/ekv/files/pdf/icmts96.pdf and made an Lt-spice model of it (Level 11 I think) The model that comes with ltspice for that transistor does not have threshold conduction and so will not work.
@@davidwparent Thankyou
Thanks!