For PDN analysis purposes, current is one of the most important but least understood concepts. It is important to start from good definitions. Here are some of the definitions that we use. Static current: is present when the power supply is in spec but no clock is present. This includes leakage current and current sources associated with analog circuits (differential pair current sources, etc). Clock edge current: comes with each clock cycle and is only present when there is a clock. The time integral of clock edge current gives the charge consumed by die circuits during each clock cycle. The peak instantaneous current for a logic core may be in the 100's of amps immediately after the clock edge but it dwindles to zero at the end of the clock period (otherwise you would not meet timing). Clock edge current has extremely high di/dt but fortunately this is supplied by the on-die decoupling capacitance and greatly filtered so that very little of it makes it out of the die, through the package and onto the PCB. Dynamic current: the time averaged clock edge current, which is the charge consumed by the clock edge current divided by the clock period. This current is proportional to clock frequency; if you double the frequency, you double the dynamic current. Dynamic current is likely to be 10's of amps when the peak instantaneous clock edge current is in the 100's of amps. Maximum current: The maximum current consumed by the PDN as seen by the VRM or bench power supply when the die circuits are all functioning in their "most activity" mode. This is often the same as dynamic current under maximum use conditions. Note that the VRM or bench power supply time constant is in mSec to possibly uSec so this is the time average of many clock cycles. Minimum current: VRM or bench power supply current when the die is in minimum activity mode. The clock may be turned off or the supply may be opened by a pass gate. Transient current: changes in dynamic current due to power management, software code or data handling requirements. This occurs when die circuits begin or end the processing of packets, logic computations or memory activity. The transient current is the difference between maximum current and minimum current (di) and may be expressed as a percentage of maximum current. It often occurs in just a few clock cycles (dt). If the clock is never turned off, the transient current is likely to be near 50%. With clock gating there may be 90% transients and power gating can produce nearly 100% transients. Intimate knowledge of circuit operations in the system environment is required to fully evaluate transient current. Transient current is often a difficult and misunderstood concept. Robust PDN design begins with an estimation of the target impedance, a primary component of which is the transient current. Hopefully the above definitions will clarify what transient current is in the context of PDN design. More information can be found in our last two DesignCon papers under http://si-list.net/files/published/ldsmith/ldsmith_papers.html Best Regards, Larry -----Original Message----- From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On Behalf Of Rohit MISHRA Sent: Friday, June 24, 2011 5:24 AM To: Kirby Goulet Cc: si-list@xxxxxxxxxxxxx Subject: [SI-LIST] Re: difference between transition current & peak current !! Most of the chip manufacturers doesn't provide transition current information in specs instead they provide worst case peak current and quiescent current information But to calculate pdn voltage drop it's transition current which is all required so as a thumb rule, it's advised to use 1/2 X worst peak current for pdn voltage drop calculation. -----Original Message----- From: Kirby Goulet [mailto:kgoulet@xxxxxxxx] Sent: Friday, June 24, 2011 5:42 PM To: Rohit MISHRA Cc: si-list@xxxxxxxxxxxxx Subject: RE: [SI-LIST] Re: difference between transition current & peak current !! I have not seen the terminology before, but one could be a reference to surge currents on power-up. The limiting factor is to prevent damage from thermal effects. The other (transition?) current could be the current step when changing a CPU from hibernate to active - in this case maintaining the voltage is important. -----Original Message----- From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On Behalf Of Rohit MISHRA Sent: June-24-11 7:51 AM To: Rajan Hansa; si-list@xxxxxxxxxxxxx Subject: [SI-LIST] Re: difference between transition current & peak current !! Rajan, Here's my 2 cents : Peak current : It has dc as well as ac transition current. It could be mostly dc and very little transition current or it could be very low quiescent(dc) and mostly transition current. Actual ratio depends on functionality of device. Transition current : It is the switching current which device consumes during switching, it's basically above or below to the dc(quiescent) value. Example : If device is taking 5 ma in quiescent state and 2 ma transition current during switching. Then peak current = 7 ma and transition current is only 2 ma. Rgds, Rohit -----Original Message----- From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On Behalf Of Rajan Hansa Sent: Friday, June 24, 2011 4:45 PM To: si-list@xxxxxxxxxxxxx Subject: [SI-LIST] difference between transition current & peak current !! Guys, While calculating pdn network drop, it's generally advised to consider worst case transition current rather than worst case peak current. I have a problem understanding difference between worst case transition & peak current ? Is peak current a dc peak current while transition is ac switching current ?? Rgds, Rajan ------------------------------------------------------------------ To unsubscribe from si-list: si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field or to administer your membership from a web page, go to: //www.freelists.org/webpage/si-list For help: si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field List technical documents are available at: http://www.si-list.net List archives are viewable at: //www.freelists.org/archives/si-list Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu ------------------------------------------------------------------ To unsubscribe from si-list: si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field or to administer your membership from a web page, go to: //www.freelists.org/webpage/si-list For help: si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field List technical documents are available at: http://www.si-list.net List archives are viewable at: //www.freelists.org/archives/si-list Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu ------------------------------------------------------------------ To unsubscribe from si-list: si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field or to administer your membership from a web page, go to: //www.freelists.org/webpage/si-list For help: si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field List technical documents are available at: http://www.si-list.net List archives are viewable at: //www.freelists.org/archives/si-list Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu Confidentiality Notice. This message may contain information that is confidential or otherwise protected from disclosure. If you are not the intended recipient, you are hereby notified that any use, disclosure, dissemination, distribution, or copying of this message, or any attachments, is strictly prohibited. If you have received this message in error, please advise the sender by reply e-mail, and delete the message and any attachments. Thank you. ------------------------------------------------------------------ To unsubscribe from si-list: si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field or to administer your membership from a web page, go to: //www.freelists.org/webpage/si-list For help: si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field List technical documents are available at: http://www.si-list.net List archives are viewable at: //www.freelists.org/archives/si-list Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu