Re: [SI-LIST] : Variability of Supply voltages

About this list Date view Thread view Subject view Author view

From: Michael Nudelman (mnudelman@tellium.com)
Date: Thu Oct 12 2000 - 13:15:20 PDT


One more thing about these power supply. SOme chips require proper sequencing.
Please don't do it, if you possibly can. It's such a royal pain in the...

Mike.

Lynne Green wrote:

> Please, please run all corners if the end users have control
> of the power supplies. If one I/O can run at all three voltages
> at the same time, you get 9 corners. If only two voltages at
> the same time, you get fewer. And include temperature
> & process corners, of course.
>
> Where I used to work, we received a dual-voltage I/O cell
> design from a contractor that worked ONLY if the power
> supplies tracked. And other I/O cells could fall out of spec
> at any Vcc1/Vcc2/Temp/Process corner, depending on the
> circuit design.
>
> It's always much better to do it right the first time. More
> work up front, but it is worth it. The ounce of prevention
> vs. the pound of cure, sigh....
>
> - Lynne
>
> -----Original Message-----
> From: owner-si-list@silab.eng.sun.com
> [mailto:owner-si-list@silab.eng.sun.com]On Behalf Of Michael Nudelman
> Sent: Tuesday, October 10, 2000 12:57 PM
> To: Bill Cohen
> Cc: si-list@silab.eng.sun.com
> Subject: Re: [SI-LIST] : Variability of Supply voltages
>
> Well, I would (meaning all 9 combinations).
>
> Suppose you use independent DC-DCs to produce all the voltages (we do, for
> 3.3 and
> 2.5). It means, that each one of them is totally independent of another. And
> the
> min-max initial setup point does not really depend on the input voltage, so
> you
> cannot assume, that if one goes to max, so will another. The initial setup,
> due to
> very high precision of regulation in todays DC-Dcs is almost solely the
> function of
> component precision.
>
> Now, suppose, you derive 2.5 from 3.3 by using a linear regulator. Again,
> for the
> linear, the output is not solely a function of the input; it is rather the
> function
> of the internal gain and the reference accuracy. So, your output may be at
> maximum,
> when the input is at minimum and vice versa.
>
> The only case I can think of when everything is at max/min at the same time
> is when
> you use dual/triple voltage DC-DC; then, since the regulation inside the
> DCDC is
> done using one voltage, and others are just simply rectified from different
> winding
> taps of the output transformer, they will follow the voltage, that is
> regulated and
> will follow it to max or min.
>
> Mike.
>
> Bill Cohen wrote:
>
> > I am a CMOS integrated circuit designer. I need some board level
> expertise.
> >
> > In our I/O circuits we usually have various transistors that run at unique
> > maximum voltages. So usually we specify three power supplies in our
> circuits.
> > The most common are: 3.3V, 2.5V, and 1.8V.
> > We then try to perform a worst case analysis using the spice simulator.
> > At the present time we don't assume that the three power supplies track.
> > So to do a worst case simulation we might run a low 3V supply and a
> > high 2V supply. The power supplies would supply a single IC and be
> > resident on the same PC board. My question is as follows:
> >
> > Can we assume that when one power supply is at a maximum the other two
> supplies
> > will be at a maximum? Should we assume that all three supplies can be
> > independent and we need to perform a maxtrix of 9 combinations to do
> > a proper worst case analysis?
> >
> > Bill Cohen
> > Circuit Design
> > Technolgoy Development Group Toshiba
> > Marlborough, MA
> > wcohen@taec.com
> >
> > **** To unsubscribe from si-list or si-list-digest: send e-mail to
> > majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
> > si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
> > si-list archives are accessible at http://www.qsl.net/wb6tpu
> > ****
>
> **** To unsubscribe from si-list or si-list-digest: send e-mail to
> majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
> si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
> si-list archives are accessible at http://www.qsl.net/wb6tpu
> ****

**** To unsubscribe from si-list or si-list-digest: send e-mail to
majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
si-list archives are accessible at http://www.qsl.net/wb6tpu
****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Tue May 08 2001 - 14:29:44 PDT