This link is dead.

Any idea where it goes now?

I take it he has a different board to drive the motors.

--- In Sumovore@yahoogroups.com, "thedogmabot" <dave@s...> (by way of
Dave Hrynkiw <dave@s...>) wrote:
>  From ScottyDogma, an interesting mounting hack for the Sumovore:
> -Dave
>
> ---------------------------------------------------------------
>
>
> Hey Sumo-Beamers,
>
>
> If you want a different way to assemble your SumoVore,
> here is an option that is fairly simple, and very solid.
>
> http://scottydogma.solarbotics.net/H-BlockMod/
>
> I don't have a sumo section on my website so....
> There is no link on my Webpage yet, but you can get to it
> direct from the link above for now.
>
>
>
> Cheers
> --... ...--
> ScottyDogma
>
>
>
>
>
>
> Yahoo! Groups Links

Hello there.

I found this group almost by accident, and was very glad to see that
people were talking about the Sumovore, as I just built mine.

This is my first foray into amateur robotics, and it took me a good 3
months of researching microcontrollers and sumo kits before I pulled
the trigger and got the Sumovore.  I'm really looking forward to
programming the ATmega8 that I got as the brainboard.

As a little background, I'm an AI programmer for console video games,
and my last AI design was based around some robotics ideas.  I decided
to take the plunge into robotics itself to see what it's like having
sensors that don't work right all the time. ;-)

So anyways, here's my first question:

After assembling my Sumovore, I noticed that the power light doesn't
come on, and the two IR emitters on the sensor board don't come on.
All of the line sensors (all 5) work perfectly, and the SV works as a
great line following bot.  I taped a bunch of electrical tape in the
kitchen and my daughters have been having fun setting up a little town.

I did a couple diagnostic checks with my multimeter to verify
voltages, and it seems that there's no voltage going to any of the
LEDs involved (the power LED and the two IR LED's).  The diag pads on
the main board showed 9v, and the diag pads on the sensor board (for
the 555 presumably, I had no docs on this diag point) show a voltage
as well.

I'm hoping for some help/things to try to figure out why my IR
emitters don't work.  The detectors work fine, and my tivo remote
shows them "lighting up."

Thanks for your help,

Steve

Hi Steve and welcome to the group!

With your AI programming you may be one of the few that actually makes use
of the 8K of program memory on the Mega8 (I usually run out of pins first).

The Diag pads on the sensor board are for setting the 38kHz if you have a
scope or meter that can read frequency. The square pad goes to ground and
the round pad is connected directly to the output of the 555. I read a DC
voltage of 2.25V for your reference (it's close to a 50% duty cycle). The
voltage here should not read 0V or 5V if so then there is problems with the
oscillator part of the circuit.

The red power LED at the back is in series with the IR LED's in the front,
that way it doubles as a power indicator and that you are transmitting IR.

If any one of the LED's were in backwards none of the LED's would get any
power, double check the orientation of each LED. Also check that the pin
headers match up with the pins. That sounds like it is though if your line
sensors are working.

Cheers,

GrantM.


At 01:20 PM 9/20/2004, you wrote:
>Hello there.
>
>I found this group almost by accident, and was very glad to see that
>people were talking about the Sumovore, as I just built mine.
>
>This is my first foray into amateur robotics, and it took me a good 3
>months of researching microcontrollers and sumo kits before I pulled
>the trigger and got the Sumovore.  I'm really looking forward to
>programming the ATmega8 that I got as the brainboard.
>
>As a little background, I'm an AI programmer for console video games,
>and my last AI design was based around some robotics ideas.  I decided
>to take the plunge into robotics itself to see what it's like having
>sensors that don't work right all the time. ;-)
>
>So anyways, here's my first question:
>
>After assembling my Sumovore, I noticed that the power light doesn't
>come on, and the two IR emitters on the sensor board don't come on.
>All of the line sensors (all 5) work perfectly, and the SV works as a
>great line following bot.  I taped a bunch of electrical tape in the
>kitchen and my daughters have been having fun setting up a little town.
>
>I did a couple diagnostic checks with my multimeter to verify
>voltages, and it seems that there's no voltage going to any of the
>LEDs involved (the power LED and the two IR LED's).  The diag pads on
>the main board showed 9v, and the diag pads on the sensor board (for
>the 555 presumably, I had no docs on this diag point) show a voltage
>as well.
>
>I'm hoping for some help/things to try to figure out why my IR
>emitters don't work.  The detectors work fine, and my tivo remote
>shows them "lighting up."
>
>Thanks for your help,
>
>Steve

Hmmm I see a very possible source of confusion the diagram is actually
showing the *back* of the connector not the front. If your referencing from
the picture this is from the side of the connector that you solder to. If
you hacked an old parallel port cable then refer to the the numbers they
are right.

GrantM.

At 01:24 PM 9/20/2004, you wrote:

>Ok, so after I got my Atmel mega8L brainboard going with my Sumovore,
>I decided to install the WinAVR package and build the sample
>sumoline.c code so I could get started with some AI programming.
>
>Unfortunately, I can't seem to communicate with my brainboard using
>the SP12-based programming cable that I built as per the atmel
>brainboard docs.  I build the 5 pin version, as I don't have an STK500
>or whatever.
>
>When I plug my programming cable into the brainboard, the green led
>next to the connector lights up, but I can't seem to contact the board
>in any meaningful way with either avrdude.exe or with avrmon.
>
>When it tries to connect, I see the green LED flash briefly, then it
>returns with an error (-2?) and tells me that no avr devices were
>found.  It returns an ID of 0xFFFFFFFF which maybe just be -1 and an
>error.
>
>I'm stumped, as I built the cable exactly like the diagram from
>Solarbotics, and the atmega8 does RUN just fine doing line following.

Ahhhh... ok.  I noticed the numbers were off (reversed) but stuck with
the pin positions as described, with the pins of the male connector
facing me.

I'll go resolder, going by numbers, and see how that goes.  I'll bet
that's the trick.  ;-)

It's cool that you have the parts for the SP12 on the board itself,
that's a nice freebie.  Makes interfacing much nicer.

--- In Sumovore@yahoogroups.com, "GrantM." <grant@s...> wrote:
> Hmmm I see a very possible source of confusion the diagram is actually
> showing the *back* of the connector not the front. If your
referencing from
> the picture this is from the side of the connector that you solder
to. If
> you hacked an old parallel port cable then refer to the the numbers
they
> are right.
>
> GrantM.
>
> At 01:24 PM 9/20/2004, you wrote:
>
> >Ok, so after I got my Atmel mega8L brainboard going with my Sumovore,
> >I decided to install the WinAVR package and build the sample
> >sumoline.c code so I could get started with some AI programming.
> >
> >Unfortunately, I can't seem to communicate with my brainboard using
> >the SP12-based programming cable that I built as per the atmel
> >brainboard docs.  I build the 5 pin version, as I don't have an STK500
> >or whatever.
> >
> >When I plug my programming cable into the brainboard, the green led
> >next to the connector lights up, but I can't seem to contact the board
> >in any meaningful way with either avrdude.exe or with avrmon.
> >
> >When it tries to connect, I see the green LED flash briefly, then it
> >returns with an error (-2?) and tells me that no avr devices were
> >found.  It returns an ID of 0xFFFFFFFF which maybe just be -1 and an
> >error.
> >
> >I'm stumped, as I built the cable exactly like the diagram from
> >Solarbotics, and the atmega8 does RUN just fine doing line following.

Hey Grant, thanks for your response!

All the LED's are in the right way, I quadruple-checked this.  Is
there a way to take one or more LED's "out of the loop" by making a
direct connection with a tinned wire between boards?

I'm trying to nail down what part of the circuit is having problems,
and if I could do a binary search somehow, that'd be really helpful.  ;-)

I'm going to go-home and resolder everything in that particular path
(the power led, ir leds, and the whole 555 circuit) and see if I had a
cold solder joint.  Everything LOOKS fine, and I'm not a novice
solderer, but I guess that's always possible.

Any more diagnostic or test suggestions would be welcomed with open arms.

--- In Sumovore@yahoogroups.com, "GrantM." <grant@s...> wrote:
> Hi Steve and welcome to the group!
>
> With your AI programming you may be one of the few that actually
makes use
> of the 8K of program memory on the Mega8 (I usually run out of pins
first).
>
> The Diag pads on the sensor board are for setting the 38kHz if you
have a
> scope or meter that can read frequency. The square pad goes to
ground and
> the round pad is connected directly to the output of the 555. I read
a DC
> voltage of 2.25V for your reference (it's close to a 50% duty
cycle). The
> voltage here should not read 0V or 5V if so then there is problems
with the
> oscillator part of the circuit.
>
> The red power LED at the back is in series with the IR LED's in the
front,
> that way it doubles as a power indicator and that you are
transmitting IR.
>
> If any one of the LED's were in backwards none of the LED's would
get any
> power, double check the orientation of each LED. Also check that the
pin
> headers match up with the pins. That sounds like it is though if
your line
> sensors are working.
>
> Cheers,
>
> GrantM.
>
>
> At 01:20 PM 9/20/2004, you wrote:
> >Hello there.
> >
> >I found this group almost by accident, and was very glad to see that
> >people were talking about the Sumovore, as I just built mine.
> >
> >This is my first foray into amateur robotics, and it took me a good 3
> >months of researching microcontrollers and sumo kits before I pulled
> >the trigger and got the Sumovore.  I'm really looking forward to
> >programming the ATmega8 that I got as the brainboard.
> >
> >As a little background, I'm an AI programmer for console video games,
> >and my last AI design was based around some robotics ideas.  I decided
> >to take the plunge into robotics itself to see what it's like having
> >sensors that don't work right all the time. ;-)
> >
> >So anyways, here's my first question:
> >
> >After assembling my Sumovore, I noticed that the power light doesn't
> >come on, and the two IR emitters on the sensor board don't come on.
> >All of the line sensors (all 5) work perfectly, and the SV works as a
> >great line following bot.  I taped a bunch of electrical tape in the
> >kitchen and my daughters have been having fun setting up a little town.
> >
> >I did a couple diagnostic checks with my multimeter to verify
> >voltages, and it seems that there's no voltage going to any of the
> >LEDs involved (the power LED and the two IR LED's).  The diag pads on
> >the main board showed 9v, and the diag pads on the sensor board (for
> >the 555 presumably, I had no docs on this diag point) show a voltage
> >as well.
> >
> >I'm hoping for some help/things to try to figure out why my IR
> >emitters don't work.  The detectors work fine, and my tivo remote
> >shows them "lighting up."
> >
> >Thanks for your help,
> >
> >Steve

At 04:13 PM 9/20/2004, you wrote:

>Ahhhh... ok.  I noticed the numbers were off (reversed) but stuck with
>the pin positions as described, with the pins of the male connector
>facing me.

Numbers == gooood,   Position == Baaaaad    :)


>I'll go resolder, going by numbers, and see how that goes.  I'll bet
>that's the trick.  ;-)
>
>It's cool that you have the parts for the SP12 on the board itself,
>that's a nice freebie.  Makes interfacing much nicer.

It makes building the cable a wee bit easier. Although if you start to make
allot of custom projects with the AVR it's a better idea to just put the
resistors in the cable itself.

GrantM.

You can try jumpering across the LED's to see if that makes a difference.
Try a peice of wire across the IR LED's first to see if the RED LED in from
turns on. If not then try jumpering across pins 26 and 27 of the top left
pin connector, this will shunt the RED LED at the back. The only way to
tell if you are transmitting IR then is with a digital camera or a video
camera. Or if you have it tuned to 38kHz the RED LED's by the IR detectors
will turn on if something is in front of it.

The only other thing to check is that the transistor is in properly. The
flat side should be facing to the side of the Sumovore.

Binary search? easy! if it's not working 0 if it is working 1...

GrantM.

At 04:10 PM 9/20/2004, you wrote:
>Hey Grant, thanks for your response!
>
>All the LED's are in the right way, I quadruple-checked this.  Is
>there a way to take one or more LED's "out of the loop" by making a
>direct connection with a tinned wire between boards?
>
>I'm trying to nail down what part of the circuit is having problems,
>and if I could do a binary search somehow, that'd be really helpful.  ;-)
>
>I'm going to go-home and resolder everything in that particular path
>(the power led, ir leds, and the whole 555 circuit) and see if I had a
>cold solder joint.  Everything LOOKS fine, and I'm not a novice
>solderer, but I guess that's always possible.
>
>Any more diagnostic or test suggestions would be welcomed with open arms.
>
>--- In Sumovore@yahoogroups.com, "GrantM." <grant@s...> wrote:
> > Hi Steve and welcome to the group!
> >
> > With your AI programming you may be one of the few that actually
>makes use
> > of the 8K of program memory on the Mega8 (I usually run out of pins
>first).
> >
> > The Diag pads on the sensor board are for setting the 38kHz if you
>have a
> > scope or meter that can read frequency. The square pad goes to
>ground and
> > the round pad is connected directly to the output of the 555. I read
>a DC
> > voltage of 2.25V for your reference (it's close to a 50% duty
>cycle). The
> > voltage here should not read 0V or 5V if so then there is problems
>with the
> > oscillator part of the circuit.
> >
> > The red power LED at the back is in series with the IR LED's in the
>front,
> > that way it doubles as a power indicator and that you are
>transmitting IR.
> >
> > If any one of the LED's were in backwards none of the LED's would
>get any
> > power, double check the orientation of each LED. Also check that the
>pin
> > headers match up with the pins. That sounds like it is though if
>your line
> > sensors are working.
> >
> > Cheers,
> >
> > GrantM.
> >
> >
> > At 01:20 PM 9/20/2004, you wrote:
> > >Hello there.
> > >
> > >I found this group almost by accident, and was very glad to see that
> > >people were talking about the Sumovore, as I just built mine.
> > >
> > >This is my first foray into amateur robotics, and it took me a good 3
> > >months of researching microcontrollers and sumo kits before I pulled
> > >the trigger and got the Sumovore.  I'm really looking forward to
> > >programming the ATmega8 that I got as the brainboard.
> > >
> > >As a little background, I'm an AI programmer for console video games,
> > >and my last AI design was based around some robotics ideas.  I decided
> > >to take the plunge into robotics itself to see what it's like having
> > >sensors that don't work right all the time. ;-)
> > >
> > >So anyways, here's my first question:
> > >
> > >After assembling my Sumovore, I noticed that the power light doesn't
> > >come on, and the two IR emitters on the sensor board don't come on.
> > >All of the line sensors (all 5) work perfectly, and the SV works as a
> > >great line following bot.  I taped a bunch of electrical tape in the
> > >kitchen and my daughters have been having fun setting up a little town.
> > >
> > >I did a couple diagnostic checks with my multimeter to verify
> > >voltages, and it seems that there's no voltage going to any of the
> > >LEDs involved (the power LED and the two IR LED's).  The diag pads on
> > >the main board showed 9v, and the diag pads on the sensor board (for
> > >the 555 presumably, I had no docs on this diag point) show a voltage
> > >as well.
> > >
> > >I'm hoping for some help/things to try to figure out why my IR
> > >emitters don't work.  The detectors work fine, and my tivo remote
> > >shows them "lighting up."
> > >
> > >Thanks for your help,
> > >
> > >Steve
>
>
>
>
>
>
>Yahoo! Groups Links
>
>
>
>

Grant
RadioShack once carried an infrared sensor card, 276-1099, that I've had
since long before there were digital cameras!  It was originally intended
for testing TV remotes.  It's an obsolete item but I've seen it in stores
(old, old stock!) for only a couple of bucks.  It's a little plastic card
with a strip of some chemical that fluoresces in IR light.  Mine still
works, and never needs batteries! :)
Keep BEAMing and dreaming (of days gone by, when RatShack and Heathkit sold
real nice electronics kits)
Tom


From: "GrantM." <grant@...>

> The only way to
> tell if you are transmitting IR then is with a digital camera or a video
> camera.

Hi Tom,

You can still pick it up at Radio Shack in the US (I've seen it in
several stores and picked one up myself).

Using the Digital camera works quite well too.

--
Dave Hylands
Vancouver, BC, Canada
http://www.DaveHylands.com/

> -----Original Message-----
> From: Tom Gray [mailto:grayed@...]
> Sent: Monday, September 20, 2004 7:51 PM
> To: Sumovore@yahoogroups.com
> Subject: Re: [Sumovore] Re: New Sumovore Owner with Questions
>
>
> Grant
> RadioShack once carried an infrared sensor card, 276-1099,
> that I've had
> since long before there were digital cameras!  It was
> originally intended
> for testing TV remotes.  It's an obsolete item but I've seen
> it in stores
> (old, old stock!) for only a couple of bucks.  It's a little
> plastic card
> with a strip of some chemical that fluoresces in IR light.
> Mine still
> works, and never needs batteries! :)
> Keep BEAMing and dreaming (of days gone by, when RatShack and
> Heathkit sold
> real nice electronics kits)
> Tom

I shorted across the left IR LED, and the power light lit up, though
the left IR LED wasn't in backwards.  None of them were in backwards.

I replaced the left IR LED with a standard green LED and the power
light went on, the green LED went on, and the right IR LED went on
(verified through my digital camcorder.)

Just for fun, I wired the left IR LED in backwards, but predictably,
it didn't work.

So my conclusion is that the left IR LED is a bad part.  I bought a
replacement IR LED at a local electronics store, but accidentally blew
it out with 9 volts the other night testing things in and around the
problem I was having.

So I guess I need a new IR LED.  Not sure how to go about getting that
specific one.

Thanks for your help in diagnosing it, btw, when the power LED came on
for the first time, I nearly jumped for joy. ;-)


--- In Sumovore@yahoogroups.com, "GrantM." <grant@s...> wrote:
> You can try jumpering across the LED's to see if that makes a
difference.
> Try a peice of wire across the IR LED's first to see if the RED LED
in from
> turns on. If not then try jumpering across pins 26 and 27 of the top
left
> pin connector, this will shunt the RED LED at the back. The only way to
> tell if you are transmitting IR then is with a digital camera or a
video
> camera. Or if you have it tuned to 38kHz the RED LED's by the IR
detectors
> will turn on if something is in front of it.
>
> The only other thing to check is that the transistor is in properly.
The
> flat side should be facing to the side of the Sumovore.
>
> Binary search? easy! if it's not working 0 if it is working 1...
>
> GrantM.
>
> At 04:10 PM 9/20/2004, you wrote:
> >Hey Grant, thanks for your response!
> >
> >All the LED's are in the right way, I quadruple-checked this.  Is
> >there a way to take one or more LED's "out of the loop" by making a
> >direct connection with a tinned wire between boards?
> >
> >I'm trying to nail down what part of the circuit is having problems,
> >and if I could do a binary search somehow, that'd be really
helpful.  ;-)
> >
> >I'm going to go-home and resolder everything in that particular path
> >(the power led, ir leds, and the whole 555 circuit) and see if I had a
> >cold solder joint.  Everything LOOKS fine, and I'm not a novice
> >solderer, but I guess that's always possible.
> >
> >Any more diagnostic or test suggestions would be welcomed with open
arms.
> >
> >--- In Sumovore@yahoogroups.com, "GrantM." <grant@s...> wrote:
> > > Hi Steve and welcome to the group!
> > >
> > > With your AI programming you may be one of the few that actually
> >makes use
> > > of the 8K of program memory on the Mega8 (I usually run out of pins
> >first).
> > >
> > > The Diag pads on the sensor board are for setting the 38kHz if you
> >have a
> > > scope or meter that can read frequency. The square pad goes to
> >ground and
> > > the round pad is connected directly to the output of the 555. I read
> >a DC
> > > voltage of 2.25V for your reference (it's close to a 50% duty
> >cycle). The
> > > voltage here should not read 0V or 5V if so then there is problems
> >with the
> > > oscillator part of the circuit.
> > >
> > > The red power LED at the back is in series with the IR LED's in the
> >front,
> > > that way it doubles as a power indicator and that you are
> >transmitting IR.
> > >
> > > If any one of the LED's were in backwards none of the LED's would
> >get any
> > > power, double check the orientation of each LED. Also check that the
> >pin
> > > headers match up with the pins. That sounds like it is though if
> >your line
> > > sensors are working.
> > >
> > > Cheers,
> > >
> > > GrantM.
> > >
> > >
> > > At 01:20 PM 9/20/2004, you wrote:
> > > >Hello there.
> > > >
> > > >I found this group almost by accident, and was very glad to see
that
> > > >people were talking about the Sumovore, as I just built mine.
> > > >
> > > >This is my first foray into amateur robotics, and it took me a
good 3
> > > >months of researching microcontrollers and sumo kits before I
pulled
> > > >the trigger and got the Sumovore.  I'm really looking forward to
> > > >programming the ATmega8 that I got as the brainboard.
> > > >
> > > >As a little background, I'm an AI programmer for console video
games,
> > > >and my last AI design was based around some robotics ideas.  I
decided
> > > >to take the plunge into robotics itself to see what it's like
having
> > > >sensors that don't work right all the time. ;-)
> > > >
> > > >So anyways, here's my first question:
> > > >
> > > >After assembling my Sumovore, I noticed that the power light
doesn't
> > > >come on, and the two IR emitters on the sensor board don't come on.
> > > >All of the line sensors (all 5) work perfectly, and the SV
works as a
> > > >great line following bot.  I taped a bunch of electrical tape
in the
> > > >kitchen and my daughters have been having fun setting up a
little town.
> > > >
> > > >I did a couple diagnostic checks with my multimeter to verify
> > > >voltages, and it seems that there's no voltage going to any of the
> > > >LEDs involved (the power LED and the two IR LED's).  The diag
pads on
> > > >the main board showed 9v, and the diag pads on the sensor board
(for
> > > >the 555 presumably, I had no docs on this diag point) show a
voltage
> > > >as well.
> > > >
> > > >I'm hoping for some help/things to try to figure out why my IR
> > > >emitters don't work.  The detectors work fine, and my tivo remote
> > > >shows them "lighting up."
> > > >
> > > >Thanks for your help,
> > > >
> > > >Steve
> >
> >
> >
> >
> >
> >
> >Yahoo! Groups Links
> >
> >
> >
> >

Has anyone interfaced to I2C devices with the Sumovore's Atmel
brainboard?

I'm tempted to try interfacing to a couple Devantech devices (the
digital compass, and the polaroid sonar device) and in theory the
Atmel Mega 8 supports the I2C interface, but since I'm new to
robotics, I thought I'd ask around about this combination of devices
before plunking down the considerable cash for those devices.

I'm also tempted to try adding wheel encoders as well.  I want to get
position and direction at any point in time so I can play with
steering behaviors.

Any thoughts?

I have heard about these but have yet to see one. Sounds useful though.

I will look for it when I am in Seattle this weekend.

GrantM.

At 10:29 PM 9/20/2004, you wrote:

>Hi Tom,
>
>You can still pick it up at Radio Shack in the US (I've seen it in
>several stores and picked one up myself).
>
>Using the Digital camera works quite well too.
>
>--
>Dave Hylands
>Vancouver, BC, Canada
>http://www.DaveHylands.com/
>
> > -----Original Message-----
> > From: Tom Gray [mailto:grayed@...]
> > Sent: Monday, September 20, 2004 7:51 PM
> > To: Sumovore@yahoogroups.com
> > Subject: Re: [Sumovore] Re: New Sumovore Owner with Questions
> >
> >
> > Grant
> > RadioShack once carried an infrared sensor card, 276-1099,
> > that I've had
> > since long before there were digital cameras!  It was
> > originally intended
> > for testing TV remotes.  It's an obsolete item but I've seen
> > it in stores
> > (old, old stock!) for only a couple of bucks.  It's a little
> > plastic card
> > with a strip of some chemical that fluoresces in IR light.
> > Mine still
> > works, and never needs batteries! :)
> > Keep BEAMing and dreaming (of days gone by, when RatShack and
> > Heathkit sold
> > real nice electronics kits)
> > Tom
>
>
>
>
>
>Yahoo! Groups Links
>
>
>
>

We can ship a replacement part out to you just send me a private message
with your name, address, phone number.

Without the IR LED's the Sumovore would still function even though the
power LED does not light. This would make a "passive" IR seeker. It's very
common that IR sensors use close to 38kHz. So if your opponent is
broadcasting IR you can still see them.

GrantM.


At 09:47 PM 9/20/2004, you wrote:

>I shorted across the left IR LED, and the power light lit up, though
>the left IR LED wasn't in backwards.  None of them were in backwards.
>
>I replaced the left IR LED with a standard green LED and the power
>light went on, the green LED went on, and the right IR LED went on
>(verified through my digital camcorder.)
>
>Just for fun, I wired the left IR LED in backwards, but predictably,
>it didn't work.
>
>So my conclusion is that the left IR LED is a bad part.  I bought a
>replacement IR LED at a local electronics store, but accidentally blew
>it out with 9 volts the other night testing things in and around the
>problem I was having.
>
>So I guess I need a new IR LED.  Not sure how to go about getting that
>specific one.
>
>Thanks for your help in diagnosing it, btw, when the power LED came on
>for the first time, I nearly jumped for joy. ;-)

Hi smariotti,

> Has anyone interfaced to I2C devices with the Sumovore's Atmel
> brainboard?
>
> I'm tempted to try interfacing to a couple Devantech devices (the
> digital compass, and the polaroid sonar device) and in theory the
> Atmel Mega 8 supports the I2C interface, but since I'm new to
> robotics, I thought I'd ask around about this combination of devices
> before plunking down the considerable cash for those devices.

The TWI (two wire interface - aka I2C) requires the SDA and SCL pins,
which are pins 27 and 28 on the Mega-8 used on the Sumovore. On the
Sumovore, pin 28 is available, but pin 27 is being used for the left
edge detector. So you'll need to do some hacking in order to make things
work.

> I'm also tempted to try adding wheel encoders as well.  I want to get
> position and direction at any point in time so I can play with
> steering behaviors.

You'll need 4 input pins (each encoder requires two pins), so if you're
willing to sacrifice some LEDs you could connect to these pins (you'll
need to remove the LEDs).

You could also use the Xtal 1 & 2 as inputs provided you don't use an
external crystal (the Mega 8 has an internal oscillator, although it
isn't accurate for serial communications).

--
Dave Hylands
Vancouver, BC, Canada
http://www.DaveHylands.com/

--- In Sumovore@yahoogroups.com, "Dave Hylands" <dhylands@b...> wrote:
> Hi smariotti,
>
> > Has anyone interfaced to I2C devices with the Sumovore's Atmel
> > brainboard?
> >
> > I'm tempted to try interfacing to a couple Devantech devices (the
> > digital compass, and the polaroid sonar device) and in theory the
> > Atmel Mega 8 supports the I2C interface, but since I'm new to
> > robotics, I thought I'd ask around about this combination of devices
> > before plunking down the considerable cash for those devices.
>
> The TWI (two wire interface - aka I2C) requires the SDA and SCL pins,
> which are pins 27 and 28 on the Mega-8 used on the Sumovore. On the
> Sumovore, pin 28 is available, but pin 27 is being used for the left
> edge detector. So you'll need to do some hacking in order to make things
> work.

I'm not afraid of a little hacking.  I could try to find something
else for the left edge detector, I guess.

> > I'm also tempted to try adding wheel encoders as well.  I want to get
> > position and direction at any point in time so I can play with
> > steering behaviors.
>
> You'll need 4 input pins (each encoder requires two pins), so if you're
> willing to sacrifice some LEDs you could connect to these pins (you'll
> need to remove the LEDs).
>
> You could also use the Xtal 1 & 2 as inputs provided you don't use an
> external crystal (the Mega 8 has an internal oscillator, although it
> isn't accurate for serial communications).

If the internal oscillator isn't accurate enough for serial comm, is
it good enough for I2C?  That's basically a simple serial interface,
right?

When you say that each encoder requires two pins, is this assuming
it's a quadrature encoder?  Could I save pins (maybe for the left edge
sensor) by using the motor drive settings (forward/reverse) to extract
direction info?

WOOT.

Ok, going by the numbers, the cable works great.  I'm now sp12 enabled
and ready to flash.  Now I just have to learn about the mega8.
Programming it in C is surprisingly similar to programming it in
assembly due to the very specific ways of addressing the chip.  My
first microcontroller, wish me luck!  ;-)

--- In Sumovore@yahoogroups.com, "GrantM." <grant@s...> wrote:
> At 04:13 PM 9/20/2004, you wrote:
>
> >Ahhhh... ok.  I noticed the numbers were off (reversed) but stuck with
> >the pin positions as described, with the pins of the male connector
> >facing me.
>
> Numbers == gooood,   Position == Baaaaad    :)
>
>
> >I'll go resolder, going by numbers, and see how that goes.  I'll bet
> >that's the trick.  ;-)
> >
> >It's cool that you have the parts for the SP12 on the board itself,
> >that's a nice freebie.  Makes interfacing much nicer.
>
> It makes building the cable a wee bit easier. Although if you start
to make
> allot of custom projects with the AVR it's a better idea to just put
the
> resistors in the cable itself.
>
> GrantM.

Hi smariotti,

...snip...
> > The TWI (two wire interface - aka I2C) requires the SDA and
> > SCL pins, which are pins 27 and 28 on the Mega-8 used on the
> > Sumovore. On the Sumovore, pin 28 is available, but pin 27
> > is being used for the left edge detector. So you'll need to
> > do some hacking in order to make things work.
>
> I'm not afraid of a little hacking.  I could try to find something
> else for the left edge detector, I guess.

You could drop one of the other line sensors (one of the middle three).
There are only six A/D converters.

...snip...
> If the internal oscillator isn't accurate enough for serial comm, is
> it good enough for I2C?  That's basically a simple serial interface,
> right?

One of the neat things about I2C is that it adapts itself to the slowest
device, so absolute clock speeds are irrelevant. It requires open
collector outputs, and all devices can drive both SCL and SDA. If one of
the devices is slow, then the SCL line will stay low. The master has to
detect this and wait for the change.

> When you say that each encoder requires two pins, is this assuming
> it's a quadrature encoder?  Could I save pins (maybe for the left edge
> sensor) by using the motor drive settings (forward/reverse) to extract
> direction info?

You could. You'll have some fun though when you're not moving but the
encoder falls right on the line. Minute changes (maybe due to vibration,
etc) could cause black-white-black-white transitions because the wheel
is moving ever so slightly back and forth.

This problem goes away with quadrature encoding, but using a single line
per wheel may work just fine if the application is right.

With a bit of extra circuitry, you could also make change the line
sensors from analog to digital. Then you could move one of the sensors
to a digital I/O pin.

--
Dave Hylands
Vancouver, BC, Canada
http://www.DaveHylands.com/

Hey all

It's time to register your bots for this years Eastern Canadian Robot
games.  Save some cash register soon. You can pre-register but pay
only when you show at the games. This is a bonus over a lot of
games.

So even if your still working on a bot register it so we can get an
idea of how many competitors will be at the games.

For more info go to www.robotgames.ca

See you at the games.

Gord

http://www.bigreviews.com/

just a few snapshots this morning.

I don't advise exploring the site unless you can tolerate bad language
and bad attitude ranting.

Steve

LOL nice page.

Just got back from Seattle where one of the SumoVores took first place in
the novice class minisumo event. Now im waiting to see one in the robomagellan!

I see that the IR LED is still missing in the picture have you received the
replacement yet?

Cheers,

GrantM.

At 10:20 PM 9/22/2004, you wrote:
>http://www.bigreviews.com/
>
>just a few snapshots this morning.
>
>I don't advise exploring the site unless you can tolerate bad language
>and bad attitude ranting.
>
>Steve

Yes, got the replacement.  Thank you.


--- In Sumovore@yahoogroups.com, "GrantM." <grant@s...> wrote:
> LOL nice page.
>
> Just got back from Seattle where one of the SumoVores took first
place in
> the novice class minisumo event. Now im waiting to see one in the
robomagellan!
>
> I see that the IR LED is still missing in the picture have you
received the
> replacement yet?
>
> Cheers,
>
> GrantM.
>
> At 10:20 PM 9/22/2004, you wrote:
> >http://www.bigreviews.com/
> >
> >just a few snapshots this morning.
> >
> >I don't advise exploring the site unless you can tolerate bad language
> >and bad attitude ranting.
> >
> >Steve

At 02:54 PM 9/20/2004, smariotti wrote:
>This link is dead.

The direct link works now.
http://scottydogma.solarbotics.net/H-BlockMod/


---------------------------------------------------------------
"Um, no - that's H,R,Y,N,K,I,W. No, not K,I,U,U, K,I,_W_.  Yes,
that's right. Yes, I know it looks like "HOCKYRINK." Yup, only
             2 vowels. Pronounciation? _SMITH_".
                  http://www.solarbotics.com

I recently finished assembly of my Sumovore.  First, I turned on
just the sensors to get them tuned.  When I was satisfied with the way
they were responding, I turned on the main power.
   As soon as the main power was switched on, the red LED's right
behind the IR detectors came on as if they had detected something.
Even in a dark room, with no objects within 10 feet, the detectors are
still detecting.  I have tried tuning the detectors with both power
sensor power and main power on, but that does not seem to work.  They
just seem to be constantly detecting.  There is no change if I place
an object directly in front of the bot.
   I have also tried to completely cover the IR detectors with black
electrical tape, but they still act as if they were receiving light.
   Any suggestions?  How do I get the IR detection sensors to respond
when both power supplies are turned on?
   The edge detetion works great at all times.
Thx

Randy

At 06:43 PM 10/17/2004, Randy wrote:
>   As soon as the main power was switched on, the red LED's right
>behind the IR detectors came on as if they had detected something.
>Even in a dark room, with no objects within 10 feet

Hmmm. In brighter light they become less sensitive.


>, the detectors are
>still detecting.  I have tried tuning the detectors with both power
>sensor power and main power on, but that does not seem to work.  They
>just seem to be constantly detecting.  There is no change if I place
>an object directly in front of the bot.

Out of curiosity, what happens when you yank the front sensor board
entirely? And it's BOTH sensors activating?

>   I have also tried to completely cover the IR detectors with black
>electrical tape, but they still act as if they were receiving light.
>   Any suggestions?  How do I get the IR detection sensors to respond
>when both power supplies are turned on?


BOTH power supplies? You must be referring to the "motor enable" switch on
the left rear corner, right?

Regards,
Dave
---------------------------------------------------------------
"Um, no - that's H,R,Y,N,K,I,W. No, not K,I,U,U, K,I,_W_.  Yes,
that's right. Yes, I know it looks like "HOCKYRINK." Yup, only
             2 vowels. Pronounciation? _SMITH_".
                  http://www.solarbotics.com

I've built up the Sumovore with Atmel brainboard and the pre-
installed code seems to run properly.

I've installed WinAVR (on a laptop running Windows ME) and loaded
sumoline.c into Programmers Notepad.  I plugged a blank, new, Mega8L
chip into the Brainboard.  I built a SPI2 programming cable
according to the pinouts in the Brainboard instructions, step 18.

When I connect the laptop to the Sumovore with this cable, and turn
on Sumovore power, the green Brainboard LED comes on and stays on.

When I click on Tools then [WinAVR] Program, the last few lines in
the new window include "AVR device not responding" and
": initialization failed, rc =-1".

I tried to install give_io, but this doesn't seem to be needed with
Windows ME.

The SPI2 cable wiring is correct according to ohmmeter check (ie,
programming header GND connects to DB-25 pin 25, etc.).  Initially I
had DB-25 pins 2 and 3 switched, but fixing this hasn't changed the
results even with another copy of Atmel Mega8L chip.

Can somebody suggest where I might go from here?  Do I need to do
some other initialization to a fresh Mega8L chip before it will
communicate?

Thanks for your help.  The clock is ticking down towards ECRG,
and I still have lots of c to learn *s*.

Nicholas Barker

This sounds like a problem that was run into recently. Turns out that the
cable diagram is wrong but the numbers on the diagram are correct. Just go
by the numbers and give that a try.

GrantM.

At 10:08 AM 10/24/2004, you wrote:
>I've built up the Sumovore with Atmel brainboard and the pre-
>installed code seems to run properly.
>
>I've installed WinAVR (on a laptop running Windows ME) and loaded
>sumoline.c into Programmers Notepad.  I plugged a blank, new, Mega8L
>chip into the Brainboard.  I built a SPI2 programming cable
>according to the pinouts in the Brainboard instructions, step 18.
>
>When I connect the laptop to the Sumovore with this cable, and turn
>on Sumovore power, the green Brainboard LED comes on and stays on.
>
>When I click on Tools then [WinAVR] Program, the last few lines in
>the new window include "AVR device not responding" and
>": initialization failed, rc =-1".
>
>I tried to install give_io, but this doesn't seem to be needed with
>Windows ME.
>
>The SPI2 cable wiring is correct according to ohmmeter check (ie,
>programming header GND connects to DB-25 pin 25, etc.).  Initially I
>had DB-25 pins 2 and 3 switched, but fixing this hasn't changed the
>results even with another copy of Atmel Mega8L chip.
>
>Can somebody suggest where I might go from here?  Do I need to do
>some other initialization to a fresh Mega8L chip before it will
>communicate?
>
>Thanks for your help.  The clock is ticking down towards ECRG,
>and I still have lots of c to learn *s*.
>
>Nicholas Barker

Hi, Grant.  Thanks for your reply.

I've just checked the cable.  The connections are:

Header pin 1   GND  DB-25 pin 25
Header pin 2  RESET DB-25 pin 3
Header pin 3   SCK  DB-25 pin 2
Header pin 4  MISO  DB-25 pin 11
Header pin 5  MOSI  DB-25 pin 8

Each header pin connects to only one DB-25 pin.

What does the green LED on the brainboard signify?

This LED is on whenever the SPI2 cable is plugged into the laptop and
the brainboard, and the Sumovore power is switched on.  Should this
LED ever be off, or should it ever be flashing?

With the SPI2 cable plugged into the laptop and the brainboard, I
measure 5.1 V on each of the pins (RESET, CLK, MISO, and MOSI) wrt
GND at P5.  Is this as it should be?

All the solder joints look good at the brainboard.

With the SPI2 cable connected only to the laptop (ie the header end
is not connected), I also measure 5.1 V at each pin wrt GND.

Can you tell if the problem is at the laptop end or at the Sumovore
end?

Many thanks,

Nicholas






--- In Sumovore@yahoogroups.com, "GrantM." <grant@s...> wrote:
> This sounds like a problem that was run into recently. Turns out
that the
> cable diagram is wrong but the numbers on the diagram are correct.
Just go
> by the numbers and give that a try.
>
> GrantM.

After posting my last message I looked for clues at the site shown in
Step 18 of the ATmega8 Brainboard instructions:

           www.xs4all.nl/~sbolt/e-spider_prog.html

I learned two things - this programmer is called SP12 not SPI2, and
MOSI should be connected to DB-25 pin 9 not pin 8 on the cable.

Once I switched the wiring at my DB-25 connector, it works!

I've downloaded sumoline.c to a blank Mega8L chip.

Turns out the error came from my trying to conserve paper.  I had
printed the .pdf version of the brainboard manual as 4 sheets to the
page, so I misread the tiny "9" to be pin 8 on the SP12 Physical
Wiring Diagram.

Now, let's see if I can make those LEDs blink in C.

Nicholas