05324 Re: MPNC was AUTOMATIC PNC

Alt-BEAM Archive

Message #05324

To: beam@sgiblab.sgi.com beam@sgiblab.sgi.com
From: James Wilson jameswilson1@home.com
Date: Monday, 19 July 1999 11:59
Subject: Re: MPNC was AUTOMATIC PNC

>Does ground need to be applied to each switch simultaneously or
>individually? I thought that the switches could be left floating and need
>only to be grounded when an action is required.
>
>James :-))
>----- Original Message -----
>From: David Perry
>To:
>Sent: Sunday, July 18, 1999 1:56 PM
>Subject: Re: MPNC was AUTOMATIC PNC
>
>
>> damn it why is it so quiet!!!!!!!
>> i can't see how that would work, you'd need to apply ground to each of
the
>> switches. oh come on, please help, are you trying to sobotage my robot
>> building or what? :-)
>>
>> David Perry
>>
>> -----Original Message-----
>> From: James Wilson
>> To: beam@sgiblab.sgi.com
>> Date: Monday, 19 July 1999 6:37
>> Subject: Re: MPNC was AUTOMATIC PNC
>>
>>
>> >Have you thought about an exclusive or gate? or an or gate? I think it
>is
>> >possible to make the gate you want by using inverters and a voltage
>> detector
>> >(things the beamer may already have in his bag of tricks)
>> >
>> >James :-))
>> >----- Original Message -----
>> >From: David Perry
>> >To:
>> >Sent: Saturday, July 17, 1999 1:11 PM
>> >Subject: Re: MPNC was AUTOMATIC PNC
>> >
>> >
>> >> holy crap! you have revolutionised the microcore!!!!!!!!!
>> >> It will be the end of the traditional microcore and PNC as we know
>it!!!!
>> >> damn, to late to implement into my caterpillerBot.
>> >>
>> >> One question while i'm at it, i am using 2x 74hc240 to reverse my new
>> bot,
>> >> but it has 4 reverse switches when i only want one. is it safe to hook
>> >them
>> >> all up ?
>> >>
>> >> David Perry
>> >>
>> >>
>> >> -----Original Message-----
>> >> From: Sean Rigter
>> >> To: BEAM
>> >> Date: Sunday, 18 July 1999 4:44
>> >> Subject: MPNC was AUTOMATIC PNC
>> >>
>> >>
>> >> >Hello group,
>> >> >
>> >> >Nudged by Bruce Robinson,
>> >> >
>> >> >>But, umm, tell me. How would your circuit look for a 6-Nv microcore?
>> >> >.
>> >> >
>> >> >here is an short article describing an improved PNC circuit with
>> >> >examples for 4 and 6 Nv cores. Also included is an introduction to
>MPIC
>> >> >pulse injection controlled gait changes. The latter is an untested
>> >> >experimental design and there is much detail to be worked out. Feel
>> >> >free to build it and report back to BEAM land.
>> >> >
>> >> >enjoy
>> >> >
>> >> >wilf
>> >> >
>> >> >
>> >> >
>> >> > MPNC Nvcores - a logical solution for saturation
>problems
>> >> > using the multiple pulse neutralizing circuit
>> >> > wilf rigter - 07/99
>> >> >
>> >> >In a conventional microcore design the dreaded twin process
saturation
>> >> >problem is addressed with a Nu based pulse neutralizing circuit (PNC)
>> >> >which absorbs all circulating processes and when it times out
>initiates
>> >> >a single process in the microcore. Since the PNC forces one Nv output
>> >> >active low any motor connected to that Nv would be rotating possibly
>> >> >mangling a pair of legs in the process. This is normally avoided by
>> >> >disabling the output drivers during the time the PNC is active.
>> >> >
>> >> >Despite the PNC's longevity in BEAM land, all things must change as
>old
>> >> >PNC based ucore designs are superseded by a new generation of Nv core
>> >> >designs which include a "multiple process neutralizing circuit"
>(MPNC).
>> >> >Unlike the PNC base ucore circuits, this new design instantly
>> >> >initializes the Nvcore and does not require any motor driver inhibit.
>> >> >After power up, the MPNC constantly monitors for and neutralizes
>> >> >multiple processes in the core. Since the MPNC does not require an Nu
>> >> >inverter, a single 74HC14 can be used for a 6Nv core.
>> >> >
>> >> >The MPNC design is added to a Nvcore with simple diode logic
requiring
>> >> >only 2 diodes for a 4Nv ucore and 4 diodes for a 6Nv hexcore.
>> >> >
>> >> >THE 4NvMPNC
>> >> >
>> >> >A 4Nv microcore uses 2 diodes to snuff out core saturation by
>> >> >neutralizing multiple (twin) processes with the anodes of the diodes
>> >> >connected to N1 and Nv2 outputs and the cathodes to Nv3 and Nv4 bias
>> >> >points respectively. 4NvMPNC.GIF shows the schematic for a MPNC ucore
>> >> >complete with a 74AC240 reverser circuit. The MPNC logic is simple:
>with
>> >> >two possible saturation states (NV1 AND Nv3 active or Nv2 AND Nv4
>> >> >active), the 1Nv output is active low and this level connected
>through
>> >> >the diode to the Nv3 bias point forces the Nv3 output high,
>neutralizing
>> >> >the saturation process. Similarly the Nv2 output neutralizes any
>> >> >potential saturation process in Nv4.
>> >> >
>> >> >THE 6NvMPNC
>> >> >
>> >> >A 6Nv MPNC hexcore requires 4 diodes compared to the 2 diode MPNC
>ucore
>> >> >to neutralize not only twin but also triplett multiple processes. To
>> >> >instantly start the 6Nv core with a simgle process in Nv1 the diodes
>are
>> >> >connected as follows: The anodes of 3 diodes are connected to Nv1
>output
>> >> >with cathodes connected to the bias points of Nv3, Nv4 and Nv5. The
>> >> >fourth diode anode is connected to Nv2 output and cathode to Nv6 bias
>> >> >point. The graphic 6NvMPNC.GIF is attached showing the schematic of a
>> >> >6Nv hexcore with MPNC.
>> >> >
>> >> >THE 6NvMPIC
>> >> >
>> >> >As an introduction to the more advanced concept of the Multiple
>Process
>> >> >Injection Circuit (MPIC) for gait control, I have included 6NVMPIC, a
>> >> >draft schematic which shows how one can inject (and regenerate) an
>> >> >arbitary number of processes in a 6 Nvcore. The MPIC includes the
same
>> >> >function as the MPNC since it must be able to inject as well as
>> >> >neutralize in order to control processes. It is possible to add the
>MPIC
>> >> >to a conventional PNC design to initiate the gait at power up. The
>MPIC
>> >> >control lines can be connected to a number of sources including DIP
>> >> >switch, "Collission Nu", "Turning Nu" etc but also (via a simple
>> >> >infrared or radio link) to an IBM PC printer port or onboard STAMP or
>> >> >PIC I/O line. The example shows a single process default HexCore with
>20
>> >> >second duration Nu controlled process injection circuits.
>> >> >
>> >> >
>> >>
>> >
>>
>

------------------------------------------------------------------------

eGroups.com home: http://www.egroups.com/group/alt-beam
http://www.egroups.com - Simplifying group communications

Home