Hi A divider can do harm if poorly implemented. Knowing the source and load characteristics can be a big help.
Bob On Jul 29, 2010, at 5:10 PM, "Didier Juges" <[email protected]> wrote: > Attenuating the control voltage near the point of use is always a good idea > when you are concerned about common mode or ground noise and particularly > when you have excess dynamic range, but you have to consider the noise added > by the divider. Some filtering may be required, and use as low a set of > resistor values as practical. > > Didier > > ------------------------ > Sent from my BlackBerry Wireless thingy while I do other things... > > -----Original Message----- > From: "Bob Camp" <[email protected]> > Sender: [email protected] > Date: Thu, 29 Jul 2010 12:41:47 > To: 'Discussion of precise time and frequency measurement'<[email protected]> > Reply-To: Discussion of precise time and frequency measurement > <[email protected]> > Subject: [time-nuts] Rb Tuning > > Hi > > > > Something to think about as you go to tune your Rb: > > > > 1) If the tune range is 5 V for 5 x 10^-9 (+/- 2.5x10^-9) then it’s a ppb / > volt or a ppt / mV. (from http://www.ham-radio.com/sbms/LPRO-101.pdf) > > 2) If you are after a goal of 1x10^-13, then that’s 0.1 ppt and 0.1 mV > > 3) The Rb pulls a half amp to an amp running normally. > > 4) One foot of number 18 wire is 6.3 mOhms (from > http://www.thelearningpit.com/elec/tools/tables/Wire_table.htm) > > 5) One amp in that wire will give you 6.3 mV > > 6) Copper wire has a temperature coefficient of 0.00393 / C at room (1% > change in 2.5 C) (from > http://www.cirris.com/testing/temperature/copper.html) > > 7) The current in the Rb heater will move around a bit > > > > The current and resistance would have to be stable to 1/630 or 0.15% for it > to be negligible relative to your goal. At 1.5% it would be the same as the > goal. > > > > Your Rb may only be ½ as sensitive as the one in the example. It also may > pull ½ A compared to the one amp I used. That gets you to a stability that > still needs to be better than 1%. I suspect you also will find that the > connections to the wire have a *lot* more resistance than the foot of wire – > back to 0.1% land. There’s also the chance that you needed more than a foot > of wire or used something smaller than number 18. For 10 feet you would need > number 8 to get the same resistance. > > > > There are a couple of solutions: > > > > 1) Reference your tuning voltage directly to the Rb ground via sense > leads. > > 2) Float the controller and single point ground it at the Rb > > 3) Attenuate the control signal at the Rb by 600:1, your tune range is > now 8x10^-12 (or +/- 4x10-12). > > 4) Bolt everything to a 6” x 6” copper buss bar. (or is that to small > ….) > > > > Sense leads probably bring in another amp in the design. It’s stability > could be an issue. Floating the controller may or may not be possible > depending on the design. Attenuating the control voltage at the Rb by 10:1 > looks like a real good idea, regardless of what else you do. > > > > Of course that’s not the whole story. The connections also have thermo > electric effects. > > > > So many things to worry about. > > > > Bob > > > > > > > > > > > > > > _______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. > _______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. > _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
