Searching \ for '[OT]: Pic-O-Botboard' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/microchip/devices.htm?key=pic
Search entire site for: 'Pic-O-Botboard'.

Exact match. Not showing close matches.
PICList Thread
'[OT]: Pic-O-Botboard'
2001\05\02@122020 by Alice Campbell

flavicon
face
Alice - could you please forward this to the piclist for me
-------------


To Roman, TomH, Jinx, RichO, and others:

After dawdling for months on this project, I finally got around to
designing my PIC botboard - which we talked about some time ago.
Basically, a PIC-based replacement for the 68HC11-based botboards.

I would appreciate it if some of you guys would take a look at the
preliminary info on this board, especially regards the features
provided, and shoot back some comments. I am about to order some
test boards to see how well the layout works, how susceptible the
PIC is to motor noise, etc.

This board actually contains many features not present on the HC11
boards, such as 4 opamp signal conditioning circuits, an additional
8-channel ULN2803 driver chip besides the 2 L293D chips, ability
to plug in up to 8 servos, etc.

I realize some of you would like to see 4A H-bridges, but not on
this go around - since that would require the pcb being about twice
as large. Big enough already at 2.5" x 4".

Preliminary info at:   http://www.oricomtech.com/bot40.htm

best regards,
- dan michaels
=================

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\05\02@130350 by jamesnewton

face picon face
Very nice! Amazingly low cost.

How about bootloader support for the F87x's? Is there a header for that?

---
James Newton (PICList Admin #3)
spam_OUTjamesnewtonTakeThisOuTspampiclist.com 1-619-652-0593
PIC/PICList FAQ: http://www.piclist.com or .org

{Original Message removed}

2001\05\02@141624 by Lawrence Lile

flavicon
face
I'm buyin' one!  I was just going to sit down and design just this very
thing, why reinvent the wheel?

I've been using PICPROTO boards from Microengineering labs - they have all
the support for one PIC, but no I/O other than a big pad-per-hole prototying
area.  Saves me a lot of time when building prototypes.

I can see using this board for other things than robot prototypes - it
duplicates the basic functions for almost any controller in a handy, cheap
package.  Cool!

-- Lawrence Lile

----- Original Message -----
From: "James Newton" <.....jamesnewtonKILLspamspam@spam@PICLIST.COM>
To: <PICLISTspamKILLspamMITVMA.MIT.EDU>
Sent: Wednesday, May 02, 2001 12:01 PM
Subject: Re: [OT]: Pic-O-Botboard


> Very nice! Amazingly low cost.
>
> How about bootloader support for the F87x's? Is there a header for that?
>
> ---
> James Newton (PICList Admin #3)
> .....jamesnewtonKILLspamspam.....piclist.com 1-619-652-0593
> PIC/PICList FAQ: http://www.piclist.com or .org
>
> {Original Message removed}

2001\05\03@095111 by Roman Black

flavicon
face
> To Roman, TomH, Jinx, RichO, and others:
>
> After dawdling for months on this project, I finally got around to
> designing my PIC botboard - which we talked about some time ago.
> Basically, a PIC-based replacement for the 68HC11-based botboards.
>
> I would appreciate it if some of you guys would take a look at the
> preliminary info on this board, especially regards the features
> provided, and shoot back some comments. I am about to order some
> test boards to see how well the layout works, how susceptible the
> PIC is to motor noise, etc.
>
> This board actually contains many features not present on the HC11
> boards, such as 4 opamp signal conditioning circuits, an additional
> 8-channel ULN2803 driver chip besides the 2 L293D chips, ability
> to plug in up to 8 servos, etc.

> Preliminary info at:   http://www.oricomtech.com/bot40.htm


Looks good Dan! Lots of hobby robot people will
like those. Hard to give you actual feedback re the
peripherals wihtout some basic circuits, like for
the opamps etc, like how many parts you could add.
It's nice to be able to add a RC network with cap
to ground and or diode, at the opamp input for
unusual signal types.

Extra 2803 is cool, that gives 8x darlington output
drivers.

How does LCD connect? Just use a dedicated digital
IO port??

I always find a dipsw or some 0.1 inch jumpers are
handy.

Anyway, it looks great! Is it size legal for
minisumo (10cm cube) or micromouse (150mm square)??
-Roman

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\05\03@160952 by Dan Michaels

flavicon
face
>> To Roman, TomH, Jinx, RichO, and others:
>>
>> After dawdling for months on this project, I finally got around to
>> designing my PIC botboard - which we talked about some time ago.
>> Basically, a PIC-based replacement for the 68HC11-based botboards.
>........

>Preliminary info at:   http://www.oricomtech.com/bot40.htm
>
Roman wrote:
>Looks good Dan! Lots of hobby robot people will
>like those. Hard to give you actual feedback re the
>peripherals wihtout some basic circuits, like for
>the opamps etc, like how many parts you could add.
>It's nice to be able to add a RC network with cap
>to ground and or diode, at the opamp input for
>unusual signal types.


Thanks for the feedback, guys, both onlist and offlist.

No room on the present board for ICSP. Also, the opamp ckts are
verstaile designs and allow various changes by leaving out & jumpering
components, etc. Also, separate pullups are provided, and the 2
inverter ckts have additional output parts:

        pullup
O----------R7---------| +5v


            +5v
             |       LM358 or
             R2      LMC6032
             |        |  \
             +--------|+  \             D1/R5
             |        |    >----+--------|>|-----+---+--->to A/D
O----C1---R1-------+---|-  /     |       diode    |   |
             |    |   |  /      |        or      R6  C4
O----C2-------+    |             |     resistor   |   |
             |    +-----R4------+                +-+-+
             R3   |             |                  |
             |    +-----C3------+                 gnd
             |
            gnd

This shows the basic ckt for all 4 amps. Two are inverters, where
input is to C1 and C2 is grounded. Two are non-inverters, where C1
is grounded, and the input is to C2 -[the two non-inv do not have
the output parts - D1/etc].

Note that you can play games with C1/R1 and C2 to make them dc-coupled,
and fiddle with D1/R5 on the output side to make different ckts, peak
detect, or low-pass, etc. Gains are adjustable via R1/R4, and BW via
C1/C3/C4.

The inverters are better for high ac-gain, where you want to limit the
possibility for oscillation, and the non-inverters can be configured for
low-gain dc-coupling of sensors. 2 pullups [R7] are available for biasing
external electret mikes/sensors/etc.
=================

>
>Extra 2803 is cool, that gives 8x darlington output
>drivers.
>

I did some more fiddling with this area of the pcb, so you can now
mix and match port B pins "simultaneously" for: direct I/O, ULN2803
driver control, direct servo control.

So you could could control say 2 servos with their own external power,
4 darlington drivers to LEDs/relays/motors/etc with hi-V power, and
also use 2 port B pins as direct I/O. OR any other combination of 8.
===========

>How does LCD connect? Just use a dedicated digital
>IO port??

Ummm - there aren't enough pins on a PIC40 to add an LCD and still
have all the other stuff that I have here. However, I "could"
leave out the ULN2803 and put in an LCD, but .........

Unfortunately, you can only push 40 pins and a small pcb and
not-smt and still cheap just so far. There's lots of other things
I wish I could put on there, but ...........

Actually, the idea is to have a basic standalone controller board,
and also the possibility to stack another SBC on top for additional
things, and talk between the two via RS-232 - using subsumption
architecture techniques:

http://www.oricomtech.com/emerge6.htm#Teli2
=================

>
>I always find a dipsw or some 0.1 inch jumpers are
>handy.
>

On-board dipsw was added in the latest go around. Can also add
a number of external dipsw and bumper sw - up to 9.
============

>Anyway, it looks great! Is it size legal for
>minisumo (10cm cube) or micromouse (150mm square)??
>-Roman

Ummmm, don't know. This thing has grown and grown as I have added
features - started as 2.5"x3.5", now 2.5"x4.1" [6.3cm x 10.4cm].

best regards,
- Dan Michaels
http://www.oricomtech.com
=======================

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\05\03@165622 by jamesnewton

face picon face
Amazing...

...so no one with an ICD will be able to program the chip on the bot board
and single step code while running the bot?

...and you can't use a bootloader?

...so, pretty much, you program the chip off the bot, insert it in the bot,
and hope it works?

...and that's NOT a ZIF socket, and it doesn't look like there's room for
one...

wow...

---
James Newton (PICList Admin #3)
EraseMEjamesnewtonspam_OUTspamTakeThisOuTpiclist.com 1-619-652-0593
PIC/PICList FAQ: http://www.piclist.com or .org

{Original Message removed}

2001\05\03@183647 by Dale Botkin

flavicon
face
Umm, yeah, it DOES sound like almost every other prototype or projct board
out there.  Other than all that extra hardware and all, that is...  8-)
ICSP would be nice, but personally I'd never expect to see a ZIF socket or
ICD support.  Not with that kind of density and cost.  It looks awfully
good to me for a few projects.

Dale

On Thu, 3 May 2001, James Newton wrote:

{Quote hidden}

> {Original Message removed}

2001\05\03@195247 by Bob Ammerman

picon face
Sorry,

ICD/ICSP takes just so little to do and is so valuable.

I for one would probably use this board for several projects, iff it has
ICD/ICSP.

There are already what seems like hundreds of .1x.1 header pins on this
thing. Would a few more be impossible?

How 'bout this:

2x5 header.

RB6    x
RB7    x
RB3    x
Vdd    z
Vss     z

The named pins on the left are connected directly to the corresponding PIC
pins. The 'x's on the right are connected to whatever those PIC pins are now
driving. The 'z's are N/C (or come up with some other function for them).

To disable ICDing and support all current functionality just install 3
jumpers between the RBn and x's.

To do ICD remove the jumpers and connect a cable to the left side of the
header.

Bob Ammerman
RAm Systems
(contract development of high performance, high function, low-level
software)


Bob Ammerman
RAm Systems
(contract development of high performance, high function, low-level
software)

----- Original Message -----
From: "Dale Botkin" <@spam@daleKILLspamspamMADMAX.BOTKIN.ORG>
To: <KILLspamPICLISTKILLspamspamMITVMA.MIT.EDU>
Sent: Thursday, May 03, 2001 6:25 PM
Subject: Re: [OT]: Pic-O-Botboard


{Quote hidden}

board
> > and single step code while running the bot?
> >
> > ...and you can't use a bootloader?
> >
> > ...so, pretty much, you program the chip off the bot, insert it in the
bot,
> > and hope it works?
> >
> > ...and that's NOT a ZIF socket, and it doesn't look like there's room
for
> > one...
> >
> > wow...
> >
> > ---
> > James Newton (PICList Admin #3)
> > RemoveMEjamesnewtonTakeThisOuTspampiclist.com 1-619-652-0593
> > PIC/PICList FAQ: http://www.piclist.com or .org
> >
> > {Original Message removed}

2001\05\03@210422 by Bob Ammerman

picon face
er....  That would be MCLR, not RB3 (sorry).

----- Original Message -----
From: "Bob Ammerman" <spamBeGoneRAMMERMANspamBeGonespamPRODIGY.NET>
To: <TakeThisOuTPICLISTEraseMEspamspam_OUTMITVMA.MIT.EDU>
Sent: Thursday, May 03, 2001 7:49 PM
Subject: Re: [OT]: Pic-O-Botboard


{Quote hidden}

now
{Quote hidden}

board
> > out there.  Other than all that extra hardware and all, that is...  8-)
> > ICSP would be nice, but personally I'd never expect to see a ZIF socket
or
{Quote hidden}

> > > {Original Message removed}

2001\05\04@005127 by rottosen

flavicon
face
Wow, Dan:

It looks like a lot of good stuff on one board. I agree with other's
comments that it sounds like it would make a good general purpose board.
I think I would buy one or more even though I have a similar, but
simpler, board myself.

I can't wait to hear more details! (I haven't had a chance to look
closely at what information you have already supplied, however...)


-- Rich


Dan Michaels wrote:
{Quote hidden}

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\05\04@005508 by rottosen

flavicon
face
Dan, maybe you can just change the pinouts on a couple of connectors to
do this without needing the jumpers? If the pins are chosen/used
carefully maybe they can do double duty.

Also, it wouldn't hurt my feelings if I had to give up a few pins
sometimes to get ICD.


-- Rich


Bob Ammerman wrote:
{Quote hidden}

> > > {Original Message removed}

2001\05\04@011726 by Tom Handley

picon face
  Dan, the board looks good but I think you will find that a 40-Pin PIC
really does not have enough I/O if you want to add several sensors. Right
now I'm using separate PICs for functions like the IR sensors, Compass,
Range finder, H-Bridge control, and speech I/O. I'm also looking at adding a
CPLD for extra I/O. One of the Lattice CPLDs I did provides 16 Inputs and
Outputs using an SPI-style interface. It's basically four shift registers in
a single package with some expanded control functions.

  While I've been focusing on testing sensors and have not finished the
overall architecture, the 16F877 core interface looks like this:

  PortA = 4 A/D channels, +Vref, and Right Tach Input (TMR0).
  PortB = Interrupt expansion. Uses an external 74HC30 and RB0/INT to
          provide 7 interrupts; 2 bumper switches, 3 IR, and 2 spares
          reserved for Rangefinder and Speech Input PICs.
  PortC = RS232, 5 SPI chip selects, Left Tach Input (TMR1).
  PortD = Dual Motor control for the H-Bridge PIC, 4 spares.
  PortE = SPI

  This makes it easy to add custom `smart modules' and frees the main CPU
to concentrate on higher functions. If I wired directly to the 16F877, I
wouldn't have enough pins and CPU resources.

  I'm looking at how much I can combine in a small CPLD. Things like motor
control, PWM, 40KHz source with three gated outputs for IR and Rangefinder,
interrupt expansion, additional I/O, etc. I'll let you know if it develops
into anything useful. Maybe we can look at another board downstream.

  - Tom

At 04:09 PM 5/3/01 -0400, Dan Michaels wrote:
{Quote hidden}

------------------------------------------------------------------------
Tom Handley
New Age Communications
Since '75 before "New Age" and no one around here is waiting for UFOs ;-)

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\05\04@050908 by Roman Black

flavicon
face
Dan Michaels wrote:

> >> After dawdling for months on this project, I finally got around to
> >> designing my PIC botboard - which we talked about some time ago.
> >> Basically, a PIC-based replacement for the 68HC11-based botboards.
> >........
>
> >Preliminary info at:   http://www.oricomtech.com/bot40.htm

>
> Thanks for the feedback, guys, both onlist and offlist.
>
> No room on the present board for ICSP.


Ouch!! That hurts... Really hurts. Especially for
most of the target markets like small bots which
need to be re-programmed all the time.

All you need is three wires, VPP,SDA,SCL on a tiny
3-pin header. Then just tie MCLR to +5v with a
47k resistor. Your board can be powered from it's own
+5v supply and the common GNDs can be connected.
All my small projects have this 3-pin ICSP header,
and we use the Picstart to program them. A 4-pin
header (including GND) is even better but you're
the one with real estate issues... ;o)

Are you saying you can't fit a 3-pin SIL header
(and one 47k resistor) anywhere on that big board??
So I would have to glue one on and run wires to
the PIC pins??
;o)

-Roman

PS. The op-amp circuit looks great!
PPS. Are you re-subscribed yet?? What happened?

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\05\04@094332 by Lawrence Lile

flavicon
face
Well, for my nickel, ICSP will be more important than the ULN2803 hooked to
port B.  I'll probably plan on hacking out the RB6 and RB7 pins somehow, and
adding an ICSP header *somewhere* .  I'm going to have another support board
nearby anyway because my battery charger is going to reside mostly on the
'Bot.  Having just recieved a big box of wierd little NiMH batteries
(Thanks, Tim!)  I'll need to build a custom charger.


I'd say ICSP is on my list of 2nd generation pic-o-botboard features.
Pretty much all my develpment now is done with a PIC16F877, serial LCD,  and
ICSP, and I port it to whatever micro I will use eventually.


My personal I/O list for a bot will be something like:

2 H bridges ( Got em on the botboard) for wheel motors
1 driver for a high current switch (could just be a single port pin) for
mowing head motor
RS232  for serial LCD
10-12 switch inputs  (mostly bump switches and a keyswitch, tiltswitch)
Possible  2 A/D for charger
1 D.O. for charger
ICSP

I'll not bother with the botboard's 12C and the ULN2803 diver chip, opening
up enough I/O for ICSP and all those inputs.   Other than that it looks like
it will meet my needs pretty well.


-- Lawrence Lile

{Original Message removed}

2001\05\04@173750 by Dan Michaels

flavicon
face
Rich Ottosen wrote:
>Wow, Dan:
>
>It looks like a lot of good stuff on one board. I agree with other's
>comments that it sounds like it would make a good general purpose board.
>I think I would buy one or more even though I have a similar, but
>simpler, board myself.
>
>I can't wait to hear more details! (I haven't had a chance to look
>closely at what information you have already supplied, however...)
>

Hi Rich,

Thanks. It's actually designed from scratch to be both botboard and
general controller. Having the ULN2803 really does add in a lot of
versatility. Plus as mentioned elsewhere, HDR3 was laid out to handle
direct port B channels, 2803 driver outputs, and servos simultaneously.

A lot of info is on the web:   http://www.oricomtech.com/bot40.htm

best regards,
- Dan Michaels
http://www.oricomtech.com
=======================

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\05\04@173755 by Dan Michaels

flavicon
face
Lawrence Lile wrote:
>Well, for my nickel, ICSP will be more important than the ULN2803 hooked to
>port B.  I'll probably plan on hacking out the RB6 and RB7 pins somehow, and
>adding an ICSP header *somewhere* .  I'm going to have another support board
>nearby anyway because my battery charger is going to reside mostly on the
>'Bot.
>

Hi Lawrence:

On the new go round, I have room for an ICSP header. Std pinout,
no RB3.

Regards the ULN2803, this chip really adds a lot of versatility to
the design. How about direct driving 8 LEDs at 500 mA, or a unipolar
stepper motor, or relays for hi-V switching, etc.

Also, all of the RB pins are already hackable - you can jumper them
straight to header HDR3, and cut the traces to the 2803 inputs
without disrupting the trace to the header.
=============

>
>I'd say ICSP is on my list of 2nd generation pic-o-botboard features.
>Pretty much all my develpment now is done with a PIC16F877, serial LCD,  and
>ICSP, and I port it to whatever micro I will use eventually.
>

This board has a MAX232, so it would probably be possible to directly
connect to a "serial" LCD if you like.
=============

{Quote hidden}

You basically have it all here - extra RS232 chan, 4 switches on R-network
to A/D RA3, 6 direct digital I/O, and 8 pins on HDR3 that can be used for
port B direct, ULN2803 driver, or servo connection. MORE than you asked for
:).

Ummm - what's a D.O. ??? [something obvious no doubt].

thanks,
- dan michaels
http://www.oricomtech.com
===================

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\05\04@173807 by Dan Michaels

flavicon
face
Dale Botkin wrote:
>Umm, yeah, it DOES sound like almost every other prototype or projct board
>out there.  Other than all that extra hardware and all, that is...  8-)
>ICSP would be nice, but personally I'd never expect to see a ZIF socket or
>ICD support.  Not with that kind of density and cost.  It looks awfully
>good to me for a few projects.
>

Hi Dale,

:)

I do have in mind a stripped down version, with no opamps and no driver
chip and no H-bridge chip. Just a regulator and a PIC and a MAX232. But
it one will have a ZIF and ICSP, and sell for peanuts too.

best regards,
- dan michaels
==================

{Quote hidden}

>> {Original Message removed}

2001\05\04@173816 by Dan Michaels

flavicon
face
Bob Ammerman wrote:
{Quote hidden}

Bob, do you really need RB3 to run the ICD, or can you get by without
it, and use the std setup?

Mclr-Vdd-gnd-RB7-RB6

I am already packing 10# into a 2# box. Twice the "normal" hole
density.

thanks,
- Dan Michaels
http://www.oricomtech.com
=======================

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\05\04@173827 by Dan Michaels

flavicon
face
Tom Handley wrote:
>   Dan, the board looks good but I think you will find that a 40-Pin PIC
>really does not have enough I/O if you want to add several sensors. Right
>now I'm using separate PICs for functions like the IR sensors, Compass,
>Range finder, H-Bridge control, and speech I/O. I'm also looking at adding a
>CPLD for extra I/O. One of the Lattice CPLDs I did provides 16 Inputs and
>Outputs using an SPI-style interface. It's basically four shift registers in
>a single package with some expanded control functions.
>


Hi Tom, thanks for the comments. As mine is a general use board, I wanted
to stay with a single PIC40 chip, like '74 and '77, since these are by
far more popular and much cheaper than larger pinout PICs. This also
argues against using gate arrays/etc.

With my approach, you can still have up to 14 general digital I/O lines
[counting the lines to the ULN2803], plus an additional 4 switch closures
-[using analog RA3 with an on-board resistor array]. I could always go
with a 68HC11 to get more pins, but that's what this project was supposed
to be getting away from :). Accessibility for picsters.

The Handyboard uses an expansion board to add many of the same things
you are doing - and I figure I can stack a 2nd [generic] SBC for
expanded features in my situation.

Basically a 2-tiered approach - provide enough capability at the
lower level to run your basic bot, and/or for general controller
apps - so 1 board suffices in many cases.

Then add another tier for fancy smanchy bot stuff, like
ultrasonics/speech recog/etc. This is where the subsumption
architecture stuff comes in that I mentioned earlier.
================


{Quote hidden}

Looks like we have slightly different approaches - [so what else is new? :)].
I'm going for a slightly more general approach with fewer specific
subsystems built in - to make the board more "general" like the Handyboard.
If the app really needs a lot of "higher functions", then another board can
piggy-back -[such as a high-powered Stamp :)].

Got any idea how "large" your board is gonna be? More than 2.6"x4.1"?
===============


>   I'm looking at how much I can combine in a small CPLD. Things like motor
>control, PWM, 40KHz source with three gated outputs for IR and Rangefinder,
>interrupt expansion, additional I/O, etc. I'll let you know if it develops
>into anything useful. Maybe we can look at another board downstream.
>

One reason I chose the ULN2803 is because it is spec'ed to switch at
less than 1 usec, so I can drive up to 8 hi--current LEDs/etc at
40 khz/whatever. I've found the chips will switch at 2 Mhz. I also
figure it should drive ultrasonic transducers directly and whatever
else.

I fiddled a lot with the layout to the 2803 to get max versatility.
A 3x10 pin header allows connections to port B direct or 2803 drivers,
plus pullups. There are also 2 different bus voltages on the header,
so you can have hi-V +DRV for the 2803, and +SVO [~5v] for driving
servos and other lo-V devices. You can mix and match RB pins, servos,
LEDs, and other devices on the header simultaneously.

best regards,
- dan michaels
http://www.oricomtech.com/bot40.htm
=======================

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\05\04@174837 by Dan Michaels

flavicon
face
At 01:54 PM 5/3/01 -0700, you wrote:
>Amazing...
>
>...so no one with an ICD will be able to program the chip on the bot board
>and single step code while running the bot?
>
>...and you can't use a bootloader?
>
>...so, pretty much, you program the chip off the bot, insert it in the bot,
>and hope it works?
>
>...and that's NOT a ZIF socket, and it doesn't look like there's room for
>one...
>
>wow...
>


Gosh, how soon we get locked into relying upon these new-fangled inventions.

But, as it turns out, I needed to add another 0.1" in height to the
PICoBotboard - now 2.6" x 4.1" - and is be room to add a std ICSP header,
which would normally be:  Mclr-Vdd-Gnd-RB7-RB6

1- Will this be adequate for using new-fnagled ICDs and bootloader?

2 - Do you need RB3?

3 - The RB lines normally go straight into the ULN2803, which is about 5K
   input resistance. Will the ICD handle this loading ok?

Also, as this is a targeted application, and not a dev board, no way
it's ever gonna have a ZIF.

thanks,
- dan michaels
http://www.oricomtech.com
======================

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\05\05@040859 by Roman Black

flavicon
face
Dan Michaels wrote:
{Quote hidden}

Dan, it might be good leave out the Vdd pin and just use a
4-pin ICSP header. Unless you want the poor programmer
to power the entire botboard?? I always just power the
board from it's own source, and use the 4-wire ICSP
approach. Works well. Imagine if the botboard is in
bot, it should have it's own supply anyway. Or otherwise
you have headaches with keeping two Vdd, one for the PIC
only and one for the rest of the botboard.
:o)
-Roman

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email RemoveMElistservKILLspamspammitvma.mit.edu with SET PICList DIGEST in the body


2001\05\05@041938 by Roman Black

flavicon
face
Dan Michaels wrote:

> I am already packing 10# into a 2# box. Twice the "normal" hole
> density.

Dan, a couple of suggestions. I know there would be
a good market with the small bot crowd. When the main
batch of boards go into production maybe you would
consider keeping them less than 100mm square, to comply
with minisumo rules, which is becoming a very popular
competition, especially with schools and tech colleges.
I think you said 4 1/4 " ?? That would be JUST too big.
Your botboard would not be considered. :o)

Also, I like the idea of maybe splitting it up into
two boards stackable. Maybe the PIC, eeprom and 2803
chip with all digital and analogue pin headers on
one board, this alone would be pretty impressive.
Then maybe an accessories board with the opamps
and power h-bridge chips, this could mount on top
and keep the heat away from the PIC and eeprom.

Just suggestions. :o)
-Roman

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservSTOPspamspamspam_OUTmitvma.mit.edu with SET PICList DIGEST in the body


2001\05\05@144831 by Dan Michaels

flavicon
face
Roman wrote:

>Dan, it might be good leave out the Vdd pin and just use a
>4-pin ICSP header. Unless you want the poor programmer
>to power the entire botboard?? I always just power the
>board from it's own source, and use the 4-wire ICSP
>approach. Works well. Imagine if the botboard is in
>bot, it should have it's own supply anyway. Or otherwise
>you have headaches with keeping two Vdd, one for the PIC
>only and one for the rest of the botboard.


Yeah, I had thought about that too - wondered if James' ICD
could drive 1.5A to power the board and output devices.

thanks,
- dan
================

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email spamBeGonelistservSTOPspamspamEraseMEmitvma.mit.edu with SET PICList DIGEST in the body


2001\05\05@144841 by Dan Michaels

flavicon
face
Roman wrote:

>
>Dan, a couple of suggestions. I know there would be
>a good market with the small bot crowd. When the main
>batch of boards go into production maybe you would
>consider keeping them less than 100mm square, to comply
>with minisumo rules, which is becoming a very popular
>competition, especially with schools and tech colleges.
>I think you said 4 1/4 " ?? That would be JUST too big.
>Your botboard would not be considered. :o)
>


Well shoot. Latest pcb incarnation is 2.6"x4.1", or 66x104 mm.
Just 0.2" too long. Thanks for the info - good to know - back
to head scratching.

I found some mini-sumo rules here:

http://www.sinerobotics.com/sumo/mini-sumo.html

and I suppose they are international in scope.
==============


{Quote hidden}

Thanks for the suggestions, Roman. Current plan is to have
enough basic capability in the main board to run a complete
bot, or general controller. It is just so nice to be able to
do it all with a single board, and not have to rely upon an
aux board from the get-go for half the basic reqs.

A piggy-back computer would be added for situations requiring
more juice or I/O, but I think the botboard as presently designed
could handle most small bot situations.

thanks,
- dan
============

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email KILLspamlistservspamBeGonespammitvma.mit.edu with SET PICList DIGEST in the body


2001\05\05@162528 by Bob Ammerman

picon face
Actually, the standard mChip ICD gets its power from the device under test,
not the other way around.

So, keep the Vdd pin on the ICD connector. You'll need:

RB6
RB7
Vdd
Gnd
MCLR

Bob Ammerman
RAm Systems
(contract development of high performance, high function, low-level
software)

{Original Message removed}

2001\05\06@044432 by Snail Instruments

flavicon
face
Personally I would preffer using RC3-5 for SPI and RE0-2 as chip selects, this way you can take advantage of the built-in 5MHz SPI. If you need to interface some odd non-standard chip, just switch the SPI off and do bit-banging.

Josef

>   PortC = RS232, 5 SPI chip selects, Left Tach Input (TMR1).
>   PortD = Dual Motor control for the H-Bridge PIC, 4 spares.
>   PortE = SPI


=======================================================================
Electronical devices for chemical laboratory, custom electronics design
-----------------------------------------------------------------------
Snail Instruments                      Josef Hanzal, M. S.
Vojanova 615                           phone/fax: +420-311-624433
266 01 Beroun                          e-mail: EraseMEsnailspamEraseMEiol.cz
Czech Republic                         URL: http://www.vitrum.cz/snail/
GPS: 49deg58'28" North, 14deg 4'35" East
=======================================================================

--
http://www.piclist.com hint: To leave the PICList
@spam@piclist-unsubscribe-request@spam@spamspam_OUTmitvma.mit.edu


2001\05\06@153602 by Dan Michaels

flavicon
face
Roman wrote:
>Dan Michaels wrote:
>
>> I am already packing 10# into a 2# box. Twice the "normal" hole
>> density.
>
>Dan, a couple of suggestions. I know there would be
>a good market with the small bot crowd. When the main
>batch of boards go into production maybe you would
>consider keeping them less than 100mm square, to comply
>with minisumo rules, which is becoming a very popular
>competition, especially with schools and tech colleges.
>I think you said 4 1/4 " ?? That would be JUST too big.
>Your botboard would not be considered. :o)
>


Shoot, Roman, trick question. It finally occurred to me
[slow, slow, slow]:

+-------------------+
| .                 |
|   .               |
|     .             |
|       .           |
|         .         |
|           .       |
|             .     |
|               .   |
+-------------------+

The diagonal of a 100mm square is ______ [how long]?

;)

- dan

--
http://www.piclist.com hint: To leave the PICList
spamBeGonepiclist-unsubscribe-requestspamKILLspammitvma.mit.edu


2001\05\06@191806 by Tony Nixon

flavicon
picon face
Dan Michaels wrote:
>
> I do have in mind a stripped down version, with no opamps and no driver
> chip and no H-bridge chip. Just a regulator and a PIC and a MAX232. But
> it one will have a ZIF and ICSP, and sell for peanuts too.


Sounds like ROMzap without the ZIF :-)

--
Best regards

Tony

mICros
http://www.bubblesoftonline.com
.....salesspam_OUTspampicnpoke.com

--
http://www.piclist.com hint: To leave the PICList
TakeThisOuTpiclist-unsubscribe-request.....spamTakeThisOuTmitvma.mit.edu


2001\05\06@213907 by Dan Michaels

flavicon
face
Tony Nixon wrote:
>Dan Michaels wrote:
>>
>> I do have in mind a stripped down version, with no opamps and no driver
>> chip and no H-bridge chip. Just a regulator and a PIC and a MAX232. But
>> it one will have a ZIF and ICSP, and sell for peanuts too.
>
>
>Sounds like ROMzap without the ZIF :-)
>


Not quite - doesn't include a PIC downloader either.

Hope you caught the msg this vectored off of, Tony - else you may
have missed the entire point - a Pic-o-Botboard without the stuff
that makes it a Pic-o-Botboard is a _______ [what?] - plain pcb
;-).

--
http://www.piclist.com hint: To leave the PICList
TakeThisOuTpiclist-unsubscribe-requestKILLspamspamspammitvma.mit.edu


2001\05\07@040957 by Roman Black

flavicon
face
Dan Michaels wrote:
{Quote hidden}

Actually this won't work either. The minisumo
rules state 10cm CUBE, and from what i've seen
the bottom 90% is jam packed full of motors,
batteries and wheels. I have actually seen
comments on hobby robot pages like "that board
is too big, get the brand-x board that will fit".
:o)
-Roman

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\05\07@113424 by Lawrence Lile

flavicon
face
D.O. = digital output
D.I. = Digital Input

This used to be common terminology in HVAC Energy Management controllers, I
still use it when I am counting up I/O at the beginning stages of a project.
I guess it dates me at about 20 years ago.  ;-)

-- Lawrence Lile


{Original Message removed}

2001\05\07@113829 by Dan Michaels

flavicon
face
Roman wrote:

{Quote hidden}

Hmmm, the following guys say it can be "any" height, but I
appreciate the heads-up on this entire matter. I'll explore
further.

thanks,
- dan
===============

http://www.sinerobotics.com/sumo/mini-sumo.html

 Rules based on Official rules of the All Japan Robot Sumo Tournament,
 ......
 Specifications For Robot

 A robot must be in such a size that it can be contained in a square tube
 of 10 cm (width) x 10 cm (depth) x any (height).  The robot can be any
 shape within this tube.
 .......

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\05\07@115135 by Lawrence Lile

flavicon
face
> You basically have it all here - extra RS232 chan, 4 switches on R-network
> to A/D RA3, 6 direct digital I/O, and 8 pins on HDR3 that can be used for
> port B direct, ULN2803 driver, or servo connection. MORE than you asked
for
> :).
>

Wow, Dan!

The only time I usually get more than I asked for is when I ask for Spam!
</sarcasm off>

-- Lawrence Lile

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\05\07@130043 by Dan Michaels

flavicon
face
Lawrence Lile wrote:
>> You basically have it all here - extra RS232 chan, 4 switches on R-network
>> to A/D RA3, 6 direct digital I/O, and 8 pins on HDR3 that can be used for
>> port B direct, ULN2803 driver, or servo connection. MORE than you asked
>for
>> :).
>>
>
>Wow, Dan!
>
>The only time I usually get more than I asked for is when I ask for Spam!
></sarcasm off>
>

Well, we "do" try to be all things to all people - even though some
always want more and more and more ;-).

I've been fiddling up schemes related to how many sensors and effectors
you can put on this sucker, and it's more than comes on the $600 Rug
Warrior and tons more than the Parallax "Grow"-bot [only 16 I/O pins
total] - plus because of the built-in driver/servo channels, it
would appear that the PIC-o-Botboard will require less "external"
circuitry to gain the same functionality.

BTW, I fiddled some more with the input layout, dropped out 1 of the
opamps and made a 8x11 hole proto area, which allows more versatility
in what you connect to RA2/5 and RE0/1/2. The original scheme was a
bit rigid. The proto area can take a 2nd opamp, if you want it.

New info/pics:  http://www.oricomtech.com/bot40.htm

thanks,
- dan michaels
====================

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\05\07@132540 by Roman Black

flavicon
face
Dan Michaels wrote:

> Well, we "do" try to be all things to all people - even though some
> always want more and more and more ;-).
>
> I've been fiddling up schemes related to how many sensors and effectors
> you can put on this sucker, and it's more than comes on the $600 Rug
> Warrior and tons more than the Parallax "Grow"-bot [only 16 I/O pins
> total] - plus because of the built-in driver/servo channels, it
> would appear that the PIC-o-Botboard will require less "external"
> circuitry to gain the same functionality.
>
> BTW, I fiddled some more with the input layout, dropped out 1 of the
> opamps and made a 8x11 hole proto area, which allows more versatility
> in what you connect to RA2/5 and RE0/1/2. The original scheme was a
> bit rigid. The proto area can take a 2nd opamp, if you want it.
>
> New info/pics:  http://www.oricomtech.com/bot40.htm


Dan, I think many people overlooked praising the thing
in their rush to critique it for you. It really is a nice
powerful board, and there are many times I have built
stuff the hard way using veroboard when I could have
used your botboard. It's cool. :o)
-Roman

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\05\07@202623 by Tony Nixon

flavicon
picon face
Dan Michaels wrote:
>
> Tony Nixon wrote:
> >Dan Michaels wrote:
> >>
> >> I do have in mind a stripped down version, with no opamps and no driver
> >> chip and no H-bridge chip. Just a regulator and a PIC and a MAX232. But
> >> it one will have a ZIF and ICSP, and sell for peanuts too.
> >
> >
> >Sounds like ROMzap without the ZIF :-)
> >
>
> Not quite - doesn't include a PIC downloader either.
>
> Hope you caught the msg this vectored off of, Tony - else you may
> have missed the entire point - a Pic-o-Botboard without the stuff
> that makes it a Pic-o-Botboard is a _______ [what?] - plain pcb
> ;-).


Yep, you're right :-)

Hey, nice stuff. I only wish I had the time to play with projects like
that.


--
Best regards

Tony

mICros
http://www.bubblesoftonline.com
.....salesspamRemoveMEbubblesoftonline.com

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\05\08@143716 by Dan Michaels

flavicon
face
Roman wrote:

>
>Dan, I think many people overlooked praising the thing
>in their rush to critique it for you. It really is a nice
>powerful board, and there are many times I have built
>stuff the hard way using veroboard when I could have
>used your botboard. It's cool. :o)


Thanks Roman, always good to get a little praise from one of the
experts. Also helps to get some critical feedback too - keep things
in perspective.

About have it wrapped now - at the point of thinking how to put 1
or 2 extra things on the pcb that will simplify external harnesses
and circuitry, plus squeezed it again - down to 100mm. Push in here,
stick out there - mini-sumo land.

best regards,
- dan
=====================

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\05\08@143723 by Dan Michaels

flavicon
face
Tony Nixon wrote:

>> Hope you caught the msg this vectored off of, Tony - else you may
>> have missed the entire point - a Pic-o-Botboard without the stuff
>> that makes it a Pic-o-Botboard is a _______ [what?] - plain pcb
>> ;-).
>
>
>Yep, you're right :-)
>
>Hey, nice stuff. I only wish I had the time to play with projects like
>that.
>


Once you've finished with the Engine, Romzap, and PIC-a-Pocket,
you might try something easy for a change :). Doesn't take near
as much time.

best regards,
- dan
================

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\05\08@201754 by Tom Handley

picon face
  Dan, I use MicroEngineering proto boards for prototying. Using a PICproto
64 board for 40-Pin devices, I managed to squeese a 512KByte NV SRAM (Dallas
with built-in battery and RTCC), 74C922 keypad encoder, LCD, MAX232AE, a
Lattice CPLD in a 44-Pin PLCC socket, headers for RS-232, LCD, keypad, SPI,
Buzzer, and 3 external chip-selects. Also, 13 Jumpers to
enable/disable/configure various functions. The CPLD provides address
generation for the SRAM and chip-selects for onboard and offboard devices.
It's a bus-based system using Ports D and E. This was hand-soldered to the
prototype area which is 2" x 3" and the board is 3" x 4". The protoboard
provides layout for the PIC, crystal, power supply, DB-25, and a 40-pin
header that brings out all the PIC signals. There is still room to use a
40-Pin Zif socket. This is the board I'm currently using as the core CPU
for testing the robot though most of the features of the protoboard are not
used and are disabled.

  The dual H-Bridge board is also hand-soldered to a blank protoboard
measuring 2" x 4". This board has 8 power HEXFETs, 4 NPN transistors, two
switching power supplies (12V charge pump for the HEXFETs and 5V for logic),
dual Tach interface using an LMC662 as a dual comparator, and a PIC16F84.
The IR is on a 0.5" x 1.5" protoboard and the interrupt expansion is on a
1" x 1.5" protoboard. The is another tiny board for a 3.3V/4A low-dropout
regulator that powers the motors and and a 5V switching supply for the core
CPU board and peripherals.

  The H-Bridge, IR, Rangefinder, Voice I/O, Compass, and power supplies
will remain on their separate boards but a lot can be integrated with the
core CPU board assuming I stick with the PICproto64

  Well, I'm starting to `babble' off-topic but I wanted to give you an idea
how things are shaping up here. I'll post an update before long. For one
thing, I've been disappointed with this Vector 2X compass...

  - Tom

At 05:38 PM 5/4/01 -0400, Dan Michaels wrote:
{Quote hidden}

------------------------------------------------------------------------
Tom Handley
New Age Communications
Since '75 before "New Age" and no one around here is waiting for UFOs ;-)

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\05\08@201810 by Tom Handley

picon face
  Josef, I normally bit-bang SPI because I run across so many devices that
have different output formats that don't fall on convenient 8-Bit
boundaries. I also normally use the related Port C pins for SPI but this
robot project is a special case. I prefer the modular approach but I may
still use a bus-based design with Ports D and E as the data bus and control
lines. As far as using Port E for SPI chip-selects, I need/want more than 3
selects right now but this could change...

  - Tom

At 04:44 AM 5/6/01 -0400, Josef wrote:
>Personally I would preffer using RC3-5 for SPI and RE0-2 as chip selects,
this way you can take advantage of the built-in 5MHz SPI. If you need to
interface some odd non-standard chip, just switch the SPI off and do
bit-banging.
>
>Josef
>
I wrote:
>>   PortC = RS232, 5 SPI chip selects, Left Tach Input (TMR1).
>>   PortD = Dual Motor control for the H-Bridge PIC, 4 spares.
>>   PortE = SPI


------------------------------------------------------------------------
Tom Handley
New Age Communications
Since '75 before "New Age" and no one around here is waiting for UFOs ;-)

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\05\21@231030 by Dan Michaels

flavicon
face
.....
>Preliminary info at:   http://www.oricomtech.com/bot40.htm
>

Here is a little update on my botboard project. I received the pcbs
and have been testing them. Everything looks fine so far.

Classically, the 68HC11 Handyboards, which also use the L293D H-bridge
chips, are said to not work well with little 3vdc type motors - too
much current and too noisy - so of course, these are precisely what
I have been concentrating on in my testing - get it to work for the
worst case :).

I have also tested a beautiful 12vdc, 100mA high-torque gear motor,
and it is quiet as can be - both sound-wise and noise-wise. No problem.

However, regarding the little 3v motors, I have been PWM'ing them up
to 600 mA, and the faster they go the more electrical noise they make,
of course. The noise that is bothersome is slow - large pulses in the
0.1-0.2 msec range, which I assume are due to brush bouncing/jitter,
and is not correlated with the PWM. The much faster noise, related to
PWM'ing, is well-filtered by the bypass and reservoir caps on the board.

The question is ---> is there any good way to get rid of the slow motor
[brush bounce?] noise?

Things done so far:

- the L293's have internal catch diodes, of course.
- also have a 20v transzorb and .047uF/250V bypass cap on the motor
 power, and have tried reservoir caps up to 1000uF.
- also tried a snubber cap on the motor directly.
- also tried a series R in the motor power lead.
- the pcb is laidout with motor drivers on one side, cpu and analog
 ckts on the other side, and a single point gnd connection between
 logic and motor gnds, made thru a ferrite bead.
- motor and logic power are separate.

Given the nature of this noise, I don't think shielding the wires
would help much.

What else can I try?

Interestingly, the noise generated by the 3v motors is no problem for
microcontroller operation, even when PWM'ing 2 motors simultaneously
- maybe PICs are less sensitive than HC11's. ??????!!

Also, it looks like the pcb is fairly well laidout, because I have
an op amp ckt with 100X gain on the board amplifying an electret
mike, and the motor electrical noise is not apparent in the output
of the amp.

Any advice about dealing with these motors is appreciated.

best regards,
- Dan Michaels
http://www.oricomtech.com
=======================

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestspamspamBeGonemitvma.mit.edu


2001\05\22@154550 by Peter L. Peres

picon face
DC brush motors generate noise in the HF-VHF range (they are small spark
gap transmitters ;-) and it is useless to add caps, you will only tune the
emissions to another band. Treat the two motor pins as a symmetrical RF
source and the casing as a shield. NEVER connect the casing to a pin. Make
a small PI filter for balanced line (2 or 4 coils, 2 or 3 caps across),
using ready made inductors (or wind your own on dispersive ferrite beads -
ready made inductors are pretty dispersive - in case of doubt add damping
resistors in parallel with the motor and the caps). Caps should be 0.01 to
0.1uF ceramic for <20W motors. This is a boilerplate recipe that needs to
be adapted to reality. The filter should be shielded and mounted directly
at the motor. It also works without a shield. Sometimes it is good to lead
the motor casing wire to some other ground than the circuit's and you
might add ferrite beads to it too if necessary. Some motors will not quiet
down no matter what.

If you ever have doubtes about a DC brush motor making noise in your
circuit, disconnect it and reconnect it through a good quality mains line
filter (Corcom etc). If the noise goes away or is reduced it was the
motor.

Hope this helps,

Peter

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\05\22@165208 by Dan Michaels

flavicon
face
Peter Peres wrote:
>DC brush motors generate noise in the HF-VHF range (they are small spark
>gap transmitters ;-) and it is useless to add caps, you will only tune the
>emissions to another band. Treat the two motor pins as a symmetrical RF
>source and the casing as a shield. NEVER connect the casing to a pin. Make
>a small PI filter for balanced line (2 or 4 coils, 2 or 3 caps across),
>using ready made inductors (or wind your own on dispersive ferrite beads -
>ready made inductors are pretty dispersive - in case of doubt add damping
>resistors in parallel with the motor and the caps). Caps should be 0.01 to
>0.1uF ceramic for <20W motors.

Peter, thanks for the suggestions. Sounds like the following type
of filter is to try:

      v   chokes(2)
     -------MMMM---+---+----+
      .            |   |    |   +-----+
      .            C   R    +---|     |
      .            |   |        |motor|
     -------MMMM---+---+--------|     |
      ^                         +--+--+
      |shield                      |case
      +----------------------------+

Probably best to connect the shield at the motor case end only, but
can try variations.

I'll have to find some chokes that can handle .5-1A, and probably play
with the R value.

The one thing I am not too sure about is how this will affect the
PWM control, as the L293D chips don't actually put dc on the motors
but rather dc interrupted at 19.3 khz.
==============


>If you ever have doubtes about a DC brush motor making noise in your
>circuit, disconnect it and reconnect it through a good quality mains line
>filter (Corcom etc). If the noise goes away or is reduced it was the
>motor.
>

There is no doubt that the motor is the source of the noise I
was referring to. Motor on, noise; motor off, no noise. More speed,
more noise.

And trying a mains filter as 1st step to determining whether I can
filter out the noise sounds like an excellent idea too.

thanks a heap,
- dan michaels
http://www.oricomtech.com
========================

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\05\22@224622 by Dan Michaels

flavicon
face
Peter,

I went out to the local surplus electronics shop and found a nice
little EMI line filter rated at 1A - std issue, 2x1.86 mH in-line
inductors, 0.01 uF across the lines, and 2x0.01 uF from lines to
gnd. This didn't do anything to squelch the noise - I think the
noise is just too slow for this type of filter.

I did play around some more with snubbers, and turns out a plain
old 15 to 120 ohm resistor across the motor drops the noise by
half, and also rounds the edges nicely. 15 ohms increases the
power drain a "bit" though.

Also, 0.01-0.47uF caps used alone increase the overshoot during PWM
switching, so they are not a good route, however, a cap + small series
R seems to be the best - knocks down the noise by 50%, to about 1v
riding on a 5v signal out of the L293D. Possibly the best I can hope
for with these particular motors.

Can't think of much else to try ????????/

tonkas,
- dan michaels
http://www.oricomtech.com
======================



thanks for the suggestions. Sounds like the following type
of filter is to try:
{Quote hidden}

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\05\23@060443 by Roman Black

flavicon
face
Hi Dan! Sorry I was too busy to comment before on
your noise problem. In all the new VCRs I see a
small toroid, with the motor wires doing 2 or 3 turns
around it right at the motor end.

Also, using a capacitor only across an inductive
load is not a good spike filter. You get all sorts
of complex tank effects where the inductor and cap
feed each other energy.
You need to *waste* the energy, try a cap and series
resistor, called a snubber network. I am familiar with
the little motors you are using and suggest about 0.47uF
poly cap and 15 ohm R.

Also you could try my 5% rule, drop about 5% of the
motor voltage with a series resistor, then put a cap
across the supply end from the resistor to ground.

Also remember these small motors take a few hours to
"seat" the brushes, and will make much worse sparks
at the comm until the square brushes wear around the
comm. In the motor shop we used to sand the brushes
on large motors until they fit the comms exact, you don't
get proper motor performance until the brushes are
bedded.

Lastly, check the L293 to make sure it has flyback
diodes on both high and low bridge sides, I think it
does, but this is important if you want decent PWM
without major spike issues. :o)
-Roman


Dan Michaels wrote:
{Quote hidden}

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\05\23@121124 by rottosen

flavicon
face
Dan:

Good to here that your board is mostly working.

Roman mentions checking if there are back EMF diodes on both the supply
and ground sides. Even more important is to check if there are diodes at
all. I seem to remember that not all versions of the L293 have the
diodes.


-- Rich


Roman Black wrote:
{Quote hidden}

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\05\23@124850 by Dan Michaels

flavicon
face
Roman wrote:
>Hi Dan! Sorry I was too busy to comment before on
>your noise problem. In all the new VCRs I see a
>small toroid, with the motor wires doing 2 or 3 turns
>around it right at the motor end.
>

Hi Roman, thanks for the response. I picked up some pre-wound
toroids but haven't got to trying them yet. I was going to rewind
the coils to have both motor windings on the same toriod with
bucking fields.

However, they are spec'ed at 60 uH and the EMI filter I tried
yesterday is 1.86 mH, so I thought the toroid chokes wound not
do much. The brush noise is just too slow, I think. One spike
alone traverses 4 to 6 19khz PWM cycles.

BTW, what is the voltage of the VCR motors you referred to?
============

>Also, using a capacitor only across an inductive
>load is not a good spike filter. You get all sorts
>of complex tank effects where the inductor and cap
>feed each other energy.
>You need to *waste* the energy, try a cap and series
>resistor, called a snubber network. I am familiar with
>the little motors you are using and suggest about 0.47uF
>poly cap and 15 ohm R.
>

Yes, this is exactly what I discovered. Small resistor, 100 ohms
or less, in series with cap works best. Cap alone makes things
worse - superimposes glitches on the PWM switching phases. The
R value doesn't seem to be critical, as long as it is small
enough. However, small 15 ohms doesn't drop the brush noise
any more than 100 ohms.

Maybe precise R-C value selection will prove optimal, have
to experiment some more. Resonance, et al [but see below].
=========

>Also you could try my 5% rule, drop about 5% of the
>motor voltage with a series resistor, then put a cap
>across the supply end from the resistor to ground.
>

I already tried a small series R in the power lead to the
motor, with little effect, but again, maybe precise selection
of series R value in conjunction with snubber R-C will prove
optimal.

Unfortunately, I have a gut feeling that if I tune this
thing too very much, it will not prove to be very robust in
the long term [ie, aging] and/or with different motors/etc.
===============

>Also remember these small motors take a few hours to
>"seat" the brushes, and will make much worse sparks
>at the comm until the square brushes wear around the
>comm.

Thanks, did not know this.
===========


>Lastly, check the L293 to make sure it has flyback
>diodes on both high and low bridge sides, I think it
>does, but this is important if you want decent PWM
>without major spike issues. :o)

Yes, the L293D chips have both pullup and pulldown diodes on
all outputs. As you know, I decided to go with this chip on this
board partly so I wouldn't have to use discrete diodes - 16 of
those big 1N54xx jobbers are as big as my entire board :).

Thanks for the suggestions. I guess playing with the snubbers
some more is the best alternative.

tonkas,
- dan michaels
http://www.oricomtech.com
======================

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\05\23@133433 by Dan Michaels

flavicon
face
Rich Ottosen wrote:
>
>Good to here that your board is mostly working.
>
>Roman mentions checking if there are back EMF diodes on both the supply
>and ground sides. Even more important is to check if there are diodes at
>all. I seem to remember that not all versions of the L293 have the
>diodes.


Hi Rich,

The pic-o-board is working fine - I am just trying to nail down
operation for the worst case situations. The Handyboard, which also
uses the L293D chips, swears people off using it with the little 3v
motors, but I think it's a layout and ckt design problem in their
case. One msg I ran across indicates that cutting some HB gnd trace
and rewiring it elsewhere cures the problem.

With a nice [expensive $20] 12vdc 100mA gearmotor, there is no noise
problem at all with my board - must be good brushes. The little 3vdc
motors cost $1, and I guess it shows.

However, with everything I have on the board, the 3v motor noise is
manageable I think - diodes in L293D chips, transzorb + bypass/reservoir
caps on motor buss, isolated gnds/busses between logic and motors,
snubbers on the motors, etc.

I also have small series Rs in all control lines leading from the PIC
to the motor ckts - as past experience had convinced me these Rs are
the single best way to protect PIC pins from many hazards.

best regards,
- dan michaels
http://www.oricomtech.com
======================

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\05\23@154232 by Peter L. Peres

picon face
Dan, you did not say that you're chopping the DC motors directly ! In that
case you'll have to 'transplant' the controller to be as close as possible
to the motor. You see, the driver tries hard to switch fast, for
efficiency, and the wires to the motor become antennas. They need to be
short! (<30 cm for sure).

The mains filter and the PI filter will have a limited effect on a
switched motor. Maybe go with a PI filter designed for >10MHz. The motor
brush noise is very high frequency. Perhaps build a ladder filter designed
for separating lower TV bands from 27MHz ham receivers. The coils can be
wound on resistors but you need more turns than with ferrite and the Q and
Cp are going to be higher (you do not want either here).

Peter

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\05\23@154243 by Peter L. Peres

picon face
> Can't think of much else to try ????????/

Oh 1V ripple on 5V (on the motor wires I presume) is low enough for a
direct switched motor.

Those switcher chips were never meant to be far from their loads afaik. I
have never used a L293 for this role so I don't know what is 'good' for
this chip.

Peter

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\05\23@161011 by Dan Michaels

flavicon
face
Peter Peres wrote:

>Dan, you did not say that you're chopping the DC motors directly ! In that
>case you'll have to 'transplant' the controller to be as close as possible
>to the motor. You see, the driver tries hard to switch fast, for
>efficiency, and the wires to the motor become antennas. They need to be
>short! (<30 cm for sure).
>

Sorry, I thought I had mentioned the motors were being PWM'ed a couple
of times. With the L293 chips, you have to run PWM straight into the
chips, and this is also what they put out. There is no smoothing of the
PWM into graded voltage anywhere in the chain. I suspect most H-bridges
work this same way. So of course you end up with very choppy voltages
on the motors, which of course results in even more noise.

However, what I found was that the noise coincident with the PWM
was effectively removed by the bypass and reservoir caps on the
motor power buss.

The noise that is left, which I suspect is due to brush bounce, is
kind of jaggedly, but is much longer in duration (0.1-0.2 msec)
than the PWM. It obviously does have some hi-freq stuff in it, as
evidenced by the risetimes and little nibs on the waves. Adding
the snubbers I mentioned last time pretty much removes the nibs
but you still have the longer stuff.

What you see looks something like the following - 5vpk, vertical
lines are PWM pulses - this represents PWM set to about 0F0h.
So you have nice clean PWM pulses, with occasional longer term
slower stuff superimposed. The snubbers effectively smooth the
superimposed stuff without significantly affecting the PWM
stuff.

5v  |
   |+-----
   |      +---------
   |----------------+--------
   |    |    |    |    |    |
   |    |    |    |    |    |
   |    |    |    |    |    |
   |    |    |    |    |    |
   |    |    |    |    |    |
   |    |    |    |    |    |
0v  |    |    |    |    |    |
   0   50   100  150  200  250 usec


>The mains filter and the PI filter will have a limited effect on a
>switched motor. Maybe go with a PI filter designed for >10MHz. The motor
>brush noise is very high frequency. Perhaps build a ladder filter designed
>for separating lower TV bands from 27MHz ham receivers. The coils can be
>wound on resistors but you need more turns than with ferrite and the Q and
>Cp are going to be higher (you do not want either here).
>

From what it looks like, I don't think the really high freq stuff
is a problem - my current circuitry pretty much eliminates it. So maybe
this is an adequate fix - since mainly what is left is that slow stuff
and it may not be a problem to the uC. However, I have a couple of
ferrite rings and plan to try winding my own chokes too.

thanks,
- dan michaels
http://www.oricomtech.com
======================

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


More... (looser matching)
- Last day of these posts
- In 2001 , 2002 only
- Today
- New search...