The plot thickens!
in [Design]
|
Prev: They let that loser Polanski off in Sweden. The UN ought to jump all over them.
Next: In search of MTP16N25E mosfets
From: oparr on 13 Jul 2010 20:30 Decided to monitor wiper voltage using a storage oscilloscope. With 10VDC connected to the 10K element and the scope's probe connected to the wiper, I'm able to see the sudden drops in voltage. See picture; http://www.pbase.com/eldata/image/126505306 Drop in voltage is around 4V and occurs at around 6V output. Worst case wiper bounce duration is around 2ms. Here's the puzzle, why is the waveform so square as though a solid state device is involved?
From: Grant on 14 Jul 2010 01:43 On Tue, 13 Jul 2010 17:30:45 -0700 (PDT), "oparr(a)hotmail.com" <oparr(a)hotmail.com> wrote: >Decided to monitor wiper voltage using a storage oscilloscope. With >10VDC connected to the 10K element and the scope's probe connected to >the wiper, I'm able to see the sudden drops in voltage. See picture; > >http://www.pbase.com/eldata/image/126505306 > >Drop in voltage is around 4V and occurs at around 6V output. Worst >case wiper bounce duration is around 2ms. Here's the puzzle, why is >the waveform so square as though a solid state device is involved? What's loading the wiper? Just the scope probe?
From: Tim Williams on 14 Jul 2010 05:53 <oparr(a)hotmail.com> wrote in message news:cec93a29-c0b8-47ef-91ca-a79c8e6a8ece(a)w30g2000yqw.googlegroups.com... > Decided to monitor wiper voltage using a storage oscilloscope. With > 10VDC connected to the 10K element and the scope's probe connected to > the wiper, I'm able to see the sudden drops in voltage. See picture; > > http://www.pbase.com/eldata/image/126505306 > > Drop in voltage is around 4V and occurs at around 6V output. Worst > case wiper bounce duration is around 2ms. Here's the puzzle, why is > the waveform so square as though a solid state device is involved? With just a probe, the fall will be a couple pF || 10M (or 1M), rather slow. The rise is actually faster than most SS, limited by lower impedances (charging that couple pF probe with a Thevenin source of around 2.5kohms). Back in the day, Tektronix had a pulse generator using mercury-wetted reed relays. Easily 1ns edges. It's kind of funny that mechanical devices are so slow and sloppy, yet generate harmonics out to insane frequencies. The switch doesn't make full contact for at least a few microseconds, so the edge is caused by point contact alone. At low voltages, quantum tunneling will carry current across the gap a few nanometers before it touches. Over 50V, in gas, arcing will occur at some distance. In high vacuum, this is replaced by field emission at high electric fields (usually leading to sputtering and vaporization of the contacts, so there is wear even on a vacuum relay). Needless to say, any of these contact methods has high resistance and high current density until things settle down and the contact area widens (which takes a few bounces!). Tim -- Deep Friar: a very philosophical monk. Website: http://webpages.charter.net/dawill/tmoranwms
From: oparr on 14 Jul 2010 08:12 > What's loading the wiper? Just the scope probe? Yes. Guess that's normal for no-load scenarios then. BTW, the bouncing has cleared up nicely after wiping back and forth in the bounce areas as you and others suggested in the components group. However, as stated eariler, my experience with wirewound pots is that the bouncing will return especially after lack of use. Others have suggested a film of dirt/dust/oxidation/stale lubricant on the element/ wiper may be the culprit. Another fellow seems to indirectly suggest that current levels play a role in that a high pot resistance combined with a relatively low voltage is not a good idea. Again, I have had no issues using conductive plastic or cermet pots of identical resistance in this application. On Jul 14, 1:43 am, Grant <o...(a)grrr.id.au> wrote: >
From: Robert Baer on 15 Jul 2010 04:49
Tim Williams wrote: > <oparr(a)hotmail.com> wrote in message news:cec93a29-c0b8-47ef-91ca-a79c8e6a8ece(a)w30g2000yqw.googlegroups.com... >> Decided to monitor wiper voltage using a storage oscilloscope. With >> 10VDC connected to the 10K element and the scope's probe connected to >> the wiper, I'm able to see the sudden drops in voltage. See picture; >> >> http://www.pbase.com/eldata/image/126505306 >> >> Drop in voltage is around 4V and occurs at around 6V output. Worst >> case wiper bounce duration is around 2ms. Here's the puzzle, why is >> the waveform so square as though a solid state device is involved? > > With just a probe, the fall will be a couple pF || 10M (or 1M), rather slow. The rise is actually faster than most SS, limited by lower impedances (charging that couple pF probe with a Thevenin source of around 2.5kohms). > > Back in the day, Tektronix had a pulse generator using mercury-wetted reed relays. Easily 1ns edges. * As far as i could tell (in those days) with a "S" plugin and a few tricks, the rise/fall was better than 500pSec and most likely better than 100pSec. > > It's kind of funny that mechanical devices are so slow and sloppy, yet generate harmonics out to insane frequencies. The switch doesn't make full contact for at least a few microseconds, so the edge is caused by point contact alone. At low voltages, quantum tunneling will carry current across the gap a few nanometers before it touches. Over 50V, in gas, arcing will occur at some distance. In high vacuum, this is replaced by field emission at high electric fields (usually leading to sputtering and vaporization of the contacts, so there is wear even on a vacuum relay). Needless to say, any of these contact methods has high resistance and high current density until things settle down and the contact area widens (which takes a few bounces!). > > Tim > |