avt
Member level 5
Hi,
for the start - I want to discuss CMFB in lvds drivers - as reference find attached the IEEE standard an the Boni LVDS driver paper :
My first question is - is the CMFB realy used to suppress disturbances cvoming into the output of the driver by the CMFB - even at frequencies of some MHZ or in the relevant of some 100 MHz.
Second question : In the IEEE-Standard (page 12, Figure 3-10) one finds a test for suppressing common signals coming into the output. Here specia care is towards transformation into differential signals - but this might be also of some concern, if you use a cmfb.
Obviuously there is no chance to reach with a current moder approach like the one used in Bonis paper single-ended termination resistances as described in the IEEE stadard - of 40 to 140 ohm (because one uses current sources - and my mosfets wit 1um length have some 10 kohm (gd) - so no adaption - and this means reflection. Further on the standard uses a quite huge capacitance (33nF) - this totaly disturbs in my case - when simulationg the circuit with "stb" in spectre the loop gain - and by this the suppression of high frequency common mode signals . So please be so kind - tell me what is realy the purpose of the cmfb - only to set up a "DC" voltage - and what can one do in order to fulfill then the requests of the IEEE standard in chapter 3.2.3 and 3.2.4. - last remark - as far as I fugured out my implementation of the circuit has a not so neat effects - in the range of 10 MHz and above it translates the common mode distrubance into a differential signal in the FIgure 3-10 test - I think this is realy not wanted ...
Ok - this was a lot of stuff - hope that someone even though is willing to start a discussion - at least about the use of cmfb in lvds drivers (only a DC-narrow-band setup circuit - but what about damping of common-mode-signals on the line then - without any good termination ???) - quite confused ...
for the start - I want to discuss CMFB in lvds drivers - as reference find attached the IEEE standard an the Boni LVDS driver paper :
My first question is - is the CMFB realy used to suppress disturbances cvoming into the output of the driver by the CMFB - even at frequencies of some MHZ or in the relevant of some 100 MHz.
Second question : In the IEEE-Standard (page 12, Figure 3-10) one finds a test for suppressing common signals coming into the output. Here specia care is towards transformation into differential signals - but this might be also of some concern, if you use a cmfb.
Obviuously there is no chance to reach with a current moder approach like the one used in Bonis paper single-ended termination resistances as described in the IEEE stadard - of 40 to 140 ohm (because one uses current sources - and my mosfets wit 1um length have some 10 kohm (gd) - so no adaption - and this means reflection. Further on the standard uses a quite huge capacitance (33nF) - this totaly disturbs in my case - when simulationg the circuit with "stb" in spectre the loop gain - and by this the suppression of high frequency common mode signals . So please be so kind - tell me what is realy the purpose of the cmfb - only to set up a "DC" voltage - and what can one do in order to fulfill then the requests of the IEEE standard in chapter 3.2.3 and 3.2.4. - last remark - as far as I fugured out my implementation of the circuit has a not so neat effects - in the range of 10 MHz and above it translates the common mode distrubance into a differential signal in the FIgure 3-10 test - I think this is realy not wanted ...
Ok - this was a lot of stuff - hope that someone even though is willing to start a discussion - at least about the use of cmfb in lvds drivers (only a DC-narrow-band setup circuit - but what about damping of common-mode-signals on the line then - without any good termination ???) - quite confused ...
