I am now designing a VCO with cross-coupled PMOS. After tran simulation, I found out that the drain current is
bidrectional with positive maximum about 2mA and negative maximum about -4mA. I think the drain current of
PMOS should always be nagative since drain voltage is always lower than  source voltage.  And in a VCO the drain current of a PMOS should be switched on and off with a value between 0 and some nagative value.

So why does this happen??  Does anybody encounter this problem before??

  Best regrads.

  Thank you.

                            Jerome

Vivkr:

The bulk and the sourcr are both conneted to Vdd and theie voltage are always higher than drain voltage.

The attatched file is the current waveform of each PMOS node and the drain voltage waveform. The gate voltage has  similar waveform as the drain voltage except 180 degree phase difference since this PMOS is cross-coupled with another PMOS.

Dear Jerome,

I see that the cross-couple PMOS have source-bulk shorted and this goes to the current mirror. So the
bulk of these devices is not really hard tied to VDD, neither is the source. Forgive me for harping on the same
theme but have you checked the transient voltages at the drain, soruce and bulk.

One way to get rid of the issue would be to tie the bulk of the cross-coupled PMOS to vdd directly.

However, the ultimate question is - What is the impact of the bidirectional drain current on your circuit?

Regards
Vivek

In my opinion, it is due to the parasitic components, specifically parasitic capacitors inside PMOS. If the frequency is high enough, the source current is not the inner source current of a transistor that you want. You can check it by reduce the oscillation frequency.

I suggest you investigate the model firstly.