The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> High-Speed I/O Design >> what decides the value of off-chip ac coupling cap
https://designers-guide.org/forum/YaBB.pl?num=1316594086

Message started by bharat on Sep 21^st, 2011, 1:34am

Title: what decides the value of off-chip ac coupling cap
Post by bharat on Sep 21^st, 2011, 1:34am

I want to understand; how do we really decide the value of ac coupling cap ? I think, it has to do with encoding scheme ( 8b/10b, 64b/66b) so that max consecutive 1s or 0s but still not having clarity on it.

I am seeing the spec and it states this was
CTX 200 75 nF

AC coupling capacitor on each Lane placed in close proximity to transmitter.

Thanks

Title: Re: what decides the value of off-chip ac coupling cap
Post by raja.cedt on Sep 21^st, 2011, 5:07am

hello be clear what you are asking. Regarding decoupling caps, generally they will attached to bias nodes and supply..otherwise simply for DC points. The lower limit is defined by amount of ripple you can tolerate max depends on area.

Are you working serial links, because you are jumping into encoding techniques. If I understand your question you are talking about AC coupling capacitors in receiver side, if so study about DC wanderwall effect.

Thanks.

Title: Re: what decides the value of off-chip ac coupling cap
Post by bharat on Sep 21^st, 2011, 7:14am

My bad, I thought the topic is High Speed I/O so one will look at the question in that context.
While sending the data from TX to RX, we block the DC common mode of TX data and give the local common mode when the data is received in RX. This is done by the off-chip cap and called ac-coupling cap. This value is given from the range of 75nF -200nF.
I want to understand, how one should get to know the value of this cap. Obviously encoding should be one of the factor which ensures that there would be minimum nos of transition. I am looking for more clarity on it.

Regards

Title: Re: what decides the value of off-chip ac coupling cap
Post by Ken Kundert on Sep 21^st, 2011, 7:19am

Just to clarify, Bharat's question is about AC coupling capacitors (also known as DC blocking capacitors), not decoupling capacitors. (This is another example of how the term decoupling capacitor is confusing and should abandoned in preference for the more appropriate bypass capacitor).

-Ken

Title: Re: what decides the value of off-chip ac coupling cap
Post by raja.cedt on Sep 21^st, 2011, 7:59am

hello barat,
sorry for the mistake. So for any RX you have a CDR, which works with a specific CID(consecutive identical digitas). Let us say CID for your CDR is 7 means you get 7 continuous one's, and due to this at RX input common mode will drop (this has to be with in limit's of the first block common mode).

Now decide time constant.

Thanks.

Title: Re: what decides the value of off-chip ac coupling cap
Post by harpoon on Sep 22^nd, 2011, 2:45am

The function of the ac coupling cap is to pass the signal and filter the DC component, and to a certain extent what the lowest signal you want to pass.

I am guessing at 75-200nF, your wanted data is in excess of 5MHz ?

You place it close to the TX as you do not want trace resistance on the PCB to factor in to the filtering/attenuation.

Title: Re: what decides the value of off-chip ac coupling cap
Post by loose-electron on Sep 26^th, 2011, 6:35pm

Capacitor needs to be big enough to pass the lowest frequency signal of interest, into the chip without significant attenuation associated with reactive voltage division between the capacitor and the inpout impedance of the chip.

Capacitor needs to be small enough such that at the highest frequency of interest going into the chip, the capacitors non ideal parameters associated with series inductance do not effect the signal integrity.

Any PCB mounted capacitor has a series inductance in it that causes the capacitor to go into self resonance at a defined frequency (series LC) and above that frequency the device looks more and more like an inductor.

Title: Re: what decides the value of off-chip ac coupling cap
Post by bharat on Oct 24^th, 2011, 12:30am

Thanks loose-electron, for the reply.
"Capacitor needs to be small enough such that at the highest frequency of interest going into the chip, the capacitors non ideal parameters associated with series inductance do not effect the signal integrity.

Any PCB mounted capacitor has a series inductance in it that causes the capacitor to go into self resonance at a defined frequency (series LC) and above that frequency the device looks more and more like an inductor. "

You mean, the above mentioned fact will push for lower value of the dc-blocking capacitor, while the ability to pass the least frequency will push for the higher value of capacitor or value of capacitor should lie between these two considerations.

Title: Re: what decides the value of off-chip ac coupling cap
Post by bharat on Oct 24^th, 2011, 12:43am

When I first saw the loose-electron's reply
"Capacitor needs to be big enough to pass the lowest frequency signal of interest, into the chip without significant attenuation associated with reactive voltage division between the capacitor and the inpout impedance of the chip."

I wasn't very sure about it. Therefore I am elaborating his point in my words for sake of others:
High Pass RC filter (C coupling and 50 ohms termination resistor), will have pole freq = 1/(2*pi*Rterm*Ccoup).
The response of RC filter would be, from that pole freq to infinity -- the gain would be 0 dB and below pole freq there will be attenuation of <0dB.
Therefore one should pick the value of Ccoup such that, the pole freq should be less than or equal to the minimum freq of of the data.
After 8b/10b encoding, the input data can be represented as band pass filter ( because minimum and maximum number of transitions are fixed), the minimum freq of that
band pass filter should coincide with the pole freq of the HP filter.
We would never want that data frequency should see any attenuation.
Obviously, if one goes for 64b/66b, the band pass filter will have much lower min freq as compared to 8b/10b encoding; therefore to accommodate 64b/66b encoding, one needs to use larger Ccoup to bring down pole frequency to lower values.

Regards,
-Bharat

Title: Re: what decides the value of off-chip ac coupling cap
Post by loose-electron on Oct 24^th, 2011, 3:32pm

Yes, there is no such thing as a perfect capacitor.

See:
http://en.wikipedia.org/wiki/Self-resonant_frequency

Also, go to a vendor site and look up the detailed specifications on capacitors. They all have some series inductance inherent to their design.

www.stanford.edu/class/ee122/.../Capacitors_demystified_ee133.pdf

http://scitation.aip.org/journals/doc/PHTEAH-ft/vol_44/iss_2/106_1.html?bypassSSO=1

THat should start you in the right direction.