Alt-BEAM Archive
Message #12560
To: beam@sgiblab.sgi.com
From: BUDSCOTT@aol.com
Date: Sun, 26 Mar 2000 14:04:06 EST
Subject: [alt-beam] Re: Allmost complete walker.
In a message dated 3/25/00 8:12:12 AM Central Standard Time,
sebastiaan_van_v@hotmail.com writes:
<< I did, bu it started to act like a schizofreniac!
>>
Hmm...well, how did it go schizo? if it occilated at too low of a frequency
lower the value, if it just went all hibbity jibbity (screwed up) then i
dunno what to tell ya, it seemed to work on mine, however i did put in a
total of 3.33 uF of caps in parrelel (looks kinda ugly, but it works). Good
Luck!
-Spencer
12561 Sun, 26 Mar 2000 12:34:59 -0800 [alt-beam] TO MUX OR XOR ,THAT IS THE QUESTION was : Reverser circuit drawn "'beam@sgiblab.sgi.com'" Wilf Rigter
The "reversed" reverser is one of many little conundrums people run into
when learning about these circuits.
The reverser layout drawing on Ian's site, is of a 240 chip laying on its
back with pins mirror image compared to the data sheet. The layout drawing
contains TWO reverser circuits on one 240 chip. This is useful for "three or
more" motor walkers and other applications. In case of 2 motor walkers you
will only need one of those reversers and the remaining 4 inverters can be
used for other things like the master/slave bicore or motor drivers. So when
experimenting with this reverser on a breadboard, use just one half of the
240 chip wired up as one reverser. As always GROUND UNUSED INPUT PINS to
avoid unexplained problems.
The reverser circuit works by letting the control pin determine if input
signals are to be inverted or not at the ouputs and this function makes this
design a XOR reverser as distinct from a MUX reverser.
<>
The first figure shows a small part of Ian's XOR reverser with the
resistor connected across an 240 inverter between pin 2 (input) and pin 18
(output). This is one of four inverters controlled by tristate enable pin 1.
When pin 1 is 0V, the inverter "inverts" signals from input to output ie a
0V signal at the input pin produces a +V signal at the output pin and
conversely a 0V signal produces a +V output.
When pin 1 is positive (+V) , then the inverter output is an open circuit or
floating pin. A signal connected to the input pin 2 passes through the 47K
resistor unchanged to the output pin 18 when the inverter output is open
circuit ie V input = V output. Unlike the inverter output, in the
noninvering mode, the 47K resistor can only drive very small loads. This
restricts the use of this reverser to driving output loads like a slave
bicore or a HC139 type h-bridge but NOT a 4 or 6 transistor h-bridge input!
Ian's almost complete walker is an example of this XOR reverser used between
the master bicore outputs and the slave bicore inputs.
The reverser works by inverting or not inverting 2 inputs connected to
bicore or microcore outputs. For a bicore whose output signals are always
complementary polarities, this inverting or non-inverting of signals is
equal to swapping outputs. When used with a microcore the result is the same
although not strictly speaking by swapping microcore outputs but rather by
changing the polarity of the rest state at motor driver inputs (motor not
running) and using an inverted polarity for the active state motor state.
In fact, this type of XOR reverser can be much more easily made with a
74AC86 XOR gate, in which small motors can be driven directly from the
output pins without additional buffers or h-bridges and which I posted many
moons ago.
The other type of reverser is a MUX or multiplexer which consists of a set
of switches that are controlled to connect outputs to different inputs. The
4016/4066 and HC4066 are the simplest of this type. There are many beam
designs by Mark and others that used the 4016 or 4066 quad bilateral
(analog) switch which the equivalent to four switch or relay contacts that
can be individually switched on or off.
These older designs used a whole 4016 chip and an extra inverter for one MUX
reverser which made this design less popular than the XOR reverser. But a
new 4066 MUX WALKER design is much more efficient and in addition provides
left/right turning circuit as well as small motor drivers in this simple 2
chip circuit.
<>
Each of the 4016/4066 sections is not actually a mux with only 1 input to 1
output selection (1 to 1) but several sections can be wired in different
configurations to make up a N input to 1 output mux . The 4051, 4052 and
4053 chips are other versions offering single 8 to 1, quad 2 to 1 and
triple 3 to 1 mux decoding respectively. Just to show you how versatile
these chips are check out Steven Bolt's designs or my own 4053 voltage
doubler design.
A relay with Double Pole Double Throw (DPDT) contacts is a good example of
an electromechanical MUX reverser is capable of switching amps of motor
current and is very popular with other robotics groups. The same relay is
also used as an H-bridge. However there is a penalty for using relays: power
and size. Relays are bigger than equivalent semiconductor h-bridges and the
coil of a relay may require 50 ma of wasted power to turn on. Lastly rated
minimum coil voltage is 5V and many beam project run at 2-3V On the plus
side, the motor is efficiently connected, with very low losses through
metallic contacts, to the power source and you are less likely to smoke a
relay compared to an h-bridge.
enjoy
wilf
> -----Original Message-----
> From: Ken Hill [SMTP:hardlock@internetcds.com]
> Sent: Saturday, March 25, 2000 11:43 PM
> To: beam@corp.sgi.com
> Subject: 2nd post - Anyone please? - Reverser circuit drawn wrong??
>
> Regarding the 240 reverser circuit from the beam online site, If appears
> that from the data sheet of the chip that the number 2 inputs and outputs
> should actually be reversed on the drawing? Those two bottom buffers run
>
> the other direction across the chip don't they? Or does it matter? Or
> should they just be moved up one pin on the chip?
>
> I assume that the IN's are enable + and - from the core or whatever brain
> circuit, and the OUT's go to + and - motor bridge IN's. When it reverses,
>
> these switch polarities at the reversers output? Using 2 opposite facing
> LEDs across each of the outputs, I can't get it to work that way on the
> breadboard. (either as drawn, or reversed). I can verify that the buffers
>
> are turning on with a low to the enables, but no reverse. (And they said
> this was a "very easy" circuit..... :)
>
> Also, I found the circuit for the 245 H bridge and the 139 reverser / H
> bridge circuit. If I use the 139 I can eliminate both the 240 reverser
> and
> the 245 H bridge, correct? (I understand the current limits and the need
> to
> stack them).
>
> Thanks, Ken
Attachment: IANSREVERSER.gif
Attachment: MUXWALKER.gif
Home