# How does a diode-connected MOS device 'clamp' the voltage

Hi,

I'm having some trouble understanding how the diode-connected device M3 maintains a fixed voltage difference between VDD (source) and node X (drain)? In other words, why is it called a diode?

I know that diode-connected devices are always in saturation if they are turned on. But who says that their Vgs will be equal exactly to their Vth. Can't Vgs be greater than Vth for a diode connected deivce?, so then the Vdd to Vx difference will no longer be a constant Vth.

• If Vgs at 1mA is 3V and Iss = 2mA thus the voltage at node X is Vdd-Vgs = Vdd - 3V, So if you change Vdd value the voltage at node X will follow the change in Vdd. Vx = 12V - 3V = 9V or Vx = 10V - 3V = 7V
– G36
Jul 26 '20 at 15:34

## 2 Answers

Can't Vgs be greater than Vth for a diode connected deivce?, so then the Vdd to Vx difference will no longer be a constant Vth.

That's not the point here. The point is, for a given voltage difference between the gates of M1 and M2 there will be defined currents through M1 and M2 (sharing Iss). This current through M1 sets the $$\V_{gs}\$$ of M3 and thereby the voltage drop from $$\V_{DD}\$$ to point X. So for a given input the voltage the voltage on point X is always $$\V_{DD} - V_{gs}\$$.

• Ah! I see, so for a defined bias current of M1, the VGS of M3 will be constant and node X is equal to VDD - Vsg. If VDD changes, node X will change or if the M1 bias changes, node X will change. But it will always obey Vdd - Vgs. Jul 26 '20 at 15:51
• May I ask, why is it called a "diode"-connected device? I would have expected a fixed "Vth" drop like a diode? Jul 26 '20 at 15:51
• Because you get a behaviour similar to a diode: A two-terminal device with a voltage drop following the IV-characteristic of a diode. Jul 26 '20 at 16:00
• Awesome. Thanks very much. I think I understand now. Jul 26 '20 at 16:09

I'm having some trouble understanding how the diode-connected device M3 maintains a fixed voltage difference between VDD (source) and node X (drain)? In other words, why is it called a diode?

The operation of the "diode-connected" M3 can be explained by the help of the negative-feedback principle as follows.

The output drain-source voltage Vds3 (the voltage of the node "X" referenced to Vdd) is compared with the threshold voltage Vth and amplified by M3 (Vth serves as the input voltage in conventional negative-feedback circuits). The output drain current changes in an opposite direction so that Vds3 is kept almost equal to Vth.