[robzy]

Hey guys,

I have a couple of questions about the following circuit:

(Assuming a transconductance of 1)

My questions are:

1) What would the current be before the first mosfet, in the middle of the two mosfets, and after the second mosfet?

2) What would the voltag be in between the two mosfets?

I think i am missing some theory that would help explain the characteristics of this circuit.

Thanks a lot,
Rob.

 

[amb]

Quote:

Originally Posted by robzy 1) What would the current be before the first mosfet, in the middle of the two mosfets, and after the second mosfet?

By "before", "middle" and "after" I assume you mean the drain of the top mosfet, the connection point between the source of the top mosfet and the drain of the bottom mosfet, and the source of the bottom mosfet, respectively.

Since the two mosfets are connected in series (they are cascoded), the current through all three points are identical.

Quote:

2) What would the voltag be in between the two mosfets?

10V - Vgs

By the way, since the bottom mosfet has a gate voltage of 5V, and most mosfets' Vgs(on) threshold is less than that, it's likely you've driven it close to, or into saturation (where the transfer curve flattens out and the current no longer increases even with more Vgs bias voltage).

 

[robzy]

Quote:

Originally Posted by amb Since the two mosfets are connected in series (they are cascoded), the current through all three points are identical.

And what current would that be? As i see it they are two seperate CCSs that are trying to draw different currents, how would the "battle" play out?

Also, is Vds always the same as Vgs?

Also, sorry, but i have one last question. According to most datasheets MOSFETs are nonlinear (transconductance-wise) for the currents that headphone amps operate that. Wouldnt that make them "not ideal"?

Rob.

 

[robzy]

[edit]: Double post, sorry.

 

[tangent]

Quote:

Originally Posted by robzy And what current would that be? As i see it they are two seperate CCSs that are trying to draw different currents, how would the "battle" play out?

I'm not sure why you're mentally modeling this as a CCS situation. It should settle on some constant value, but any variation at the gates will change everything. I'd be happier if I saw something at the gates that itself was constant, and not just an assertion that "this is always X volts".

Anyway...

This seems like a problem for SPICE. I'm probably just ignorant, but it seems to me that you'd need calculus to figure this out by hand. You have two interdependent nonlinear equations. Ick.

Quote:

Also, is Vds always the same as Vgs?

They should be equal only by chance. There's no force trying to keep them equal.

Quote:

MOSFETs are nonlinear (transconductance-wise) for the currents that headphone amps operate that. Wouldnt that make them "not ideal"?

Of course. This is why amp designers fuss so much about biasing. They're trying to pushing the transistor into the most linear section of its operating curve. (Unless they're trying to add creative amounts of distortion, that is.)

 

[amb]

Quote:

Originally Posted by robzy According to most datasheets MOSFETs are nonlinear (transconductance-wise) for the currents that headphone amps operate that. Wouldnt that make them "not ideal"?

All devices are non-linear, be it MOSFET, JFET, BJT, triode, pentode..., whatever. They are each non-linear in different ways (just look at the transfer curves of each). If you use a single device as a simple amplifier with no negative feedback, you would get lots of distortion. However, the use of negative feedback and other techniques such as cascoding, and carefully designing the bias points, you can remove much of the non-linearity and get very low distortion with any device type. There are other issues to consider besides linearity when picking devices, and there is a lot of religion surrounding them. Pick your own poison...

 

[Rescue Toaster]

Quote:

Originally Posted by robzy Hey guys, I have a couple of questions about the following circuit: Thanks a lot, Rob.

You need some feedback on the bottom mosfet for it to become a CCS. Easiest is to put a resistor on the source. Without some feedback that bottom mosfet will act like an opamp without feedback- a comparator or switch - on/off.

Here's Gary Pimm's battery-biased cascode mosfet CCS:
http://home.pacifier.com/~gpimm/Battery_biased_ccs.htm

This makes it clearer what vds is for the bottom mosfet. It's equal to B2 - the vgs of the cascode (upper) mosfet at the operating current.

vgs for q1 + the voltage across R1 & it's trimmer always adds up to B1. If the current goes down then voltage across the resistor drops, vgs increases and current goes back up.

All the CCS action is determined by Q1 and the equation B1 = vgs(q1) + Vr1. Q2's source voltage should be Vr1 + B2 - Vgs(q2). All of those are pretty constant, so we know vds for Q1 is very close to constant as well.