at the temporal coordinates: 1/23/02 11:56 AM, the entity known as Richard
G. DAVIS at msys1_at_charter.net conveyed the following:
>
> on 1/23/02 10:47 AM, Zachery Bir at zbir_at_urbanape.com wrote:
>
>>
>> On Wednesday, January 23, 2002, at 10:34 , Ed Kummel wrote:
>>
>>> For instance, forget binary computing...how about
>>> "trinary?" three states, on, off and maybe...
>>
>> Uh, 'cause electricity is on or off?
>>
>> Zachery Bir <zbir_at_urbanape.com>
>> <URL:http://www.urbanape.com/>
>>
>
> Trinary state devices are available in the market. Try Motorola for a good
> line of devices that have been in the market for a number of years.
>
> They are preferred for several reasons.
>
> One, the number of 'switches' needed to set up some device is reduced,
> and/or the number of logical alternatives specified by a trinary number is
> larger for a given number of 'digits'.
>
> Two, the Motorola devices, and similar devices by other vendors, are
> inherently more resistant to noise appearing in data that come in trinary
> format instead of binary.
>
> Remember in a serial bit stream, if it is binary coded data, then in any
> instant in time the probability that the observed data is wrong is 50%. For
> trinary coded data, the probability of an instantaneous error is reduced to
> 33%.
>
> That may not seem like much, but in an intrinsically highly noisy
> application, going from binary to trinary data coding can significantly
> improve the noise immunity of the overall system. So, don't dismiss the
> idea of trinary coded data without considering that alternative.
>
So, would that be used in rad-hard devices? The recent Galileo upset at a
most inopportune time (30 min before Io flyby) brings this to mind.
- Eric.
--Eric Strobel (fyzycyst_at_NOSPAM^mailaps.org)
===================================================================== Remember... Moderation in all things, INCLUDING MODERATION!! =====================================================================
-- This is the Newtontalk mailinglist - http://www.newtontalk.net To unsubscribe or manage: visit the above link or mailto:newtontalk-request_at_newtontalk.net?Subject=unsubscribe
This archive was generated by hypermail 2.1.2 : Fri Feb 01 2002 - 16:03:03 EST