Designer’s Guide Community

shaikss_rf

Junior Member

Offline

Posts: 22

NMOS Bulk connections - Source -Bulk voltage
Nov 20^th, 2013, 3:33am

Hi,

Rf voltage is fed as input to NMOS source. RF input has both positive and negative voltage.
Bulk of NMOS is connected to most negative voltage in the complete circuit.
Now, the RF voltage has negative component.
So, how should I connect the bulk of NMOS?

How will the positive VSB(source- Bulk) and negative VSB (Source-bulk) impact?
What and how should be the bulk current for NMOS?

RF input is fed to source of NMOS and gate is being fed by control voltage of 1V. Bulk is grounded. Now, when I check the VSB, its has both positive and negative cycles. What is the impact of this behavior on the overall design? When I checked the bulk current, it is around 400uA when the width of MOS is 40um.

Pls help me out.

Note: All discussion is related to NMOS.

Back to top

8.zip

IP Logged

aaron_do

Senior Fellow

Offline

Posts: 1398

Re: NMOS Bulk connections - Source -Bulk voltage
Reply #1 - Nov 21^st, 2013, 5:33pm

Hi again,

I'm assuming you are talking about the OFF branch. Your body diode is turning on. If you have deep n-well, use a big resistor at the body connection to create a floating body, but make sure you simulate the behavior. Even better, if you have access to a negative supply voltage, bias the body at a negative voltage (-VDD for example). If you can't do that and you are still having trouble, you could use a big RC to bias your source and drain at +VDD while keeping the body at 0V. That will guarantee that the diode doesn't turn on.

Aaron

Back to top

there is no energy in matter other than that received from the environment - Nikola Tesla

IP Logged

shaikss_rf Junior Member Offline Posts: 22	Re: NMOS Bulk connections - Source -Bulk voltage Reply #2 - Nov 21^st, 2013, 9:36pm I tried all the above suggested methods. But I ended up with similar plots for the last 2 suggestions. Pls check the plots and help me.
Back to top	11.zip IP Logged

shaikss_rf

Junior Member

Offline

Posts: 22

Re: NMOS Bulk connections - Source -Bulk voltage
Reply #3 - Nov 21^st, 2013, 9:50pm

aaron_do wrote on Nov 21^st, 2013, 5:33pm:

Hi again,

I'm assuming you are talking about the OFF branch. Your body diode is turning on. If you have deep n-well, use a big resistor at the body connection to create a floating body, but make sure you simulate the behavior. Even better, if you have access to a negative supply voltage, bias the body at a negative voltage (-VDD for example). If you can't do that and you are still having trouble, you could use a big RC to bias your source and drain at +VDD while keeping the body at 0V. That will guarantee that the diode doesn't turn on.

Aaron

Pls check the below plot. Bulk current of M1 and M3 is going to few mA. I have connected the bulk to -Vdd(-2 V). But still I see bulk current.

Back to top

12.png

IP Logged

aaron_do

Senior Fellow

Offline

Posts: 1398

Re: NMOS Bulk connections - Source -Bulk voltage
Reply #4 - Nov 21^st, 2013, 10:01pm

Hi,

Is the body source or body drain voltage going above 0 V? If not, then its unlikely the diode is turning on. Most likely you are just looking at the current through the junction capacitances. You can check by lowering your input frequency to a very small number. If the body current is still there, then its not the caps.

If it is the caps, then you need to realize that the switch has an inherent trade-off between isolation and insertion loss. If you want better isolation, you need to reduce your switch width. This will of course degrade your insertion loss. You just need to decide if the isolation is good enough for you. One more thing which I mentioned before, if you really want good isolation, you need a shunt branch too (see below).

INPUT------OFF Transistor--------OUTPUT
|
|
|
ON Transistor
|
|
|
GND

Back to top

there is no energy in matter other than that received from the environment - Nikola Tesla

IP Logged