Current mirroring issue in LC VCO

[ravet]

hello

in the lc vco, if i mirror a current 100u to 1m, by multiplying factor of 10, is there a possibility that current exceeds 1 mA cause of the presence of coil??, and how to overcome if this happens???

thanks

 

[jfyan]

Re: current in lc vco

hello

in the lc vco, if i mirror a current 100u to 1m, by multiplying factor of 10, is there a possibility that current exceeds 1 mA cause of the presence of coil??, and how to overcome if this happens???

thanks

hi,
it is possible, because the current source can't guarantee the output voltage(vds) is equal to vgs, so if vds > vgs, the output current is larger than you expected.
so you can use some cascode or some other complex current source ckt to get a relatively constant current.
good luck
jeff

 

[ravet]

Re: current in lc vco

hello

in the lc vco, if i mirror a current 100u to 1m, by multiplying factor of 10, is there a possibility that current exceeds 1 mA cause of the presence of coil??, and how to overcome if this happens???

thanks

hi,
it is possible, because the current source can't guarantee the output voltage(vds) is equal to vgs, so if vds > vgs, the output current is larger than you expected.
so you can use some cascode or some other complex current source ckt to get a relatively constant current.
good luck
jeff

it seems that you talk about the cross-coupled transistors, well , drain connected to gate of same mos by a coil which you consider short circuit , but there is small resistor in series with the coil and this has drop, and so vds not exactly equal vgs, also the vds of the current source varies according to the vds of the cross-coupled, ,,,,,,so, what make current exceeds the expected current ??as i mirror it to 1 mA , what make it larger???

 

[eeliuliu]

Re: current in lc vco

hello

in the lc vco, if i mirror a current 100u to 1m, by multiplying factor of 10, is there a possibility that current exceeds 1 mA cause of the presence of coil??, and how to overcome if this happens???

thanks

hi,
it is possible, because the current source can't guarantee the output voltage(vds) is equal to vgs, so if vds > vgs, the output current is larger than you expected.
so you can use some cascode or some other complex current source ckt to get a relatively constant current.
good luck
jeff

Yes, it's possible. there is a little difference between DC current and TRAN curent. It will be a little bigger than 1mA when the cuicirt works on.
Also, in the process corners, the current source might change too.

 

[ravet]

Re: current in lc vco

hello

in the lc vco, if i mirror a current 100u to 1m, by multiplying factor of 10, is there a possibility that current exceeds 1 mA cause of the presence of coil??, and how to overcome if this happens???

thanks

hi,
it is possible, because the current source can't guarantee the output voltage(vds) is equal to vgs, so if vds > vgs, the output current is larger than you expected.
so you can use some cascode or some other complex current source ckt to get a relatively constant current.
good luck
jeff

Yes, it's possible. there is a little difference between DC current and TRAN curent. It will be a little bigger than 1mA when the cuicirt works on.
Also, in the process corners, the current source might change too.

i talk about an increase of about 15% of the expected current , i.e. if i mirror a 1mA to 10 mA, by ratio 1:10 , the current drawn to the core is 11.5mA !! which it seems to me strange specially the mirroring devices are matched i.e both PMOS and same dimensions ,??????

 

[laglead]

Re: current in lc vco

it is possible

 

[benever]

Re: current in lc vco

the one of my design is lower ,why??

hello

in the lc vco, if i mirror a current 100u to 1m, by multiplying factor of 10, is there a possibility that current exceeds 1 mA cause of the presence of coil??, and how to overcome if this happens???

thanks

hi,
it is possible, because the current source can't guarantee the output voltage(vds) is equal to vgs, so if vds > vgs, the output current is larger than you expected.
so you can use some cascode or some other complex current source ckt to get a relatively constant current.
good luck
jeff