Re: The Ring Oscillator Operates in a Range ...
posted on
Jun 16, 2013 03:05PM
Hopefully what I'm about to post will help you understand our position. How often is the RO controlled? Every revolution across the chain of inverters? The ring oscillator is controlled when enough error builds up in the phase detector, and then thr charge pump releases voltage to speed up or down the RO. The RO is not always controlled and is allowed to run off of the supply voltage - not the control voltage (charge pump) - a lot of the time. So since the RO is controlled within a range, a significant portion of the clock cycles are not being controlled and the RO (VCO) is running free and is running off of the familiar manufacturing process(how the silicon is "baked"), SUPPLY voltage, and temperature. While the ring oscillator is operating within the range of the phase detector, which it does a lot, it is not operating off of the PLL voltage, it is operating off of the supply voltage to the entire chip. The ring oscillators in the PLLs have two general voltage connections 1) is the SUPPLY voltage which is the voltage running across the entire chip, and the other 2) is the control voltage, which is part of the PLL via the phase detector, charge pump and filter. The control voltage is not always controlling. It controls within a range and while the RO is in the range of operation, it is running off of the supply voltage just like the 336. Since the 336 does not have to operate at its maximum possible speed at all times per the claim construction win, we have a good argument. The Texas Instruments expert wrote a paper essentially saying the same as the above. Will Judge Gildea buy our argument? While the RO is in the range where it is running off of the supply voltage (not PLL control voltage), from what I've read it can vary as much as 20% from where the control voltage regulates the speed. I would like to know the ratio of clock cycles that are within the range to clock cycles actually being controlled. Jim Otteson referenced the operational range and the Texas Instruments expert did also. I'm assuming Qualcomm and the other company's expert would say something similar. It is up to Otteson to extract information on how modern day PLLs operate in line with Dr. O's infringement contentions from the three chip manufacturing companies TPL called. I would like to see a few settlements before August 31st.