Suggested schematic BX-35 with Atmega8535 error?

Started by Ed August 5, 2005
Hi

I have been doing some circuits with the BX-35 chip and one day I
noticed that the onboard programmer I use (the suggested bx35 hookup
schematic) didn't work well. Most of the times I got a EEPROM Checksum
error, which puzzled me a lot since I have been using the same design
over and over. After a different test I noticed that the chip wasn't
anymore the AT90S but the Atmega8535. I read the forum and I found
nothing in therms of different hook up for it, but my tested circuit
didn't work anymore so I tested an old AT90S and it worked again at
first!!

I tryed then to program the Atmega on the BXDS and it worked fine so
something was wrong. Looking at the BXDS schematic I noticed that R47
is a 10k resistor instead of the 100k suggested on the Bx-35 schematic
(the one that goes from the ATN line to ground).

I did change the resistor on my circuit for a 10k and... it worked!!!
No more problems anymore with the onboard programmer.

I just though I should let everyone know (just in case someone has
notised the same)

Thanks

Ed



Thanks!.
I just posted a hack last week that allows you to use the Development
Station to run two processors at the same time.

Needless to say, the ATTN lines 100K resistor has been food for
thought. I went into the Amtel documents and the they explain
clearly that the processor Sinks 20ma, but they are a bit ellusive on
how much it will Source or internal clamping diodes or internal
pullup/pulldown resistors, so I just let it go for now.

Obviously 100K really seems to be very, very little current at 5V.
The 1OK makes everything much more sensible.

I just wish that NetMedia would coherently collect and publish
errata, as the documentation seems to never have been revised.

Also, the change over in processor name was due to substantial
improvements in design flaws in the older version. The new chip is
less prone to SPI interface problems and about 10 other problems.
You may want to take a look at Amtel's Errata Sheet on the changeover. --- In basicx@basi..., "Ed" <esteyezz@y...> wrote:
> Hi
>
> I have been doing some circuits with the BX-35 chip and one day I
> noticed that the onboard programmer I use (the suggested bx35 hookup
> schematic) didn't work well. Most of the times I got a EEPROM
Checksum
> error, which puzzled me a lot since I have been using the same
design
> over and over. After a different test I noticed that the chip wasn't
> anymore the AT90S but the Atmega8535. I read the forum and I found
> nothing in therms of different hook up for it, but my tested circuit
> didn't work anymore so I tested an old AT90S and it worked again at
> first!!
>
> I tryed then to program the Atmega on the BXDS and it worked fine so
> something was wrong. Looking at the BXDS schematic I noticed that
R47
> is a 10k resistor instead of the 100k suggested on the Bx-35
schematic
> (the one that goes from the ATN line to ground).
>
> I did change the resistor on my circuit for a 10k and... it
worked!!!
> No more problems anymore with the onboard programmer.
>
> I just though I should let everyone know (just in case someone has
> notised the same)
>
> Thanks
>
> Ed



G. Kramer Herzog wrote:

> I just wish that NetMedia would coherently collect and publish
> errata, as the documentation seems to never have been revised.

The FAQ page on my website was supposed to be a collection of links to
useful previous messages in the Yahoo group but I never quite got around
to finishing it. I have downloaded all the previous messages from Yahoo
into text files for easier offline processing. So if anyone wants to
take up the challenge I can supply the text and publish the results. I
did recently update the Links on my website categorized into 14
sections. I think everyone should find something useful.

>
> Also, the change over in processor name was due to substantial
> improvements in design flaws in the older version. The new chip is
> less prone to SPI interface problems and about 10 other problems.
> You may want to take a look at Amtel's Errata Sheet on the changeover.

The new BX-24p uses the ATMega8535. NetMedia claim a 26% performance
improvement. This has been verified by group members and we have
surmised that all of this gain is due to the 2x faster SPI bus on that
chip. NetMedia also make some other improvements in the download
procedure and made use of a few other chip improvements. The Atmel
application note explains the chip changes:
http://atmel.com/dyn/resources/prod_documents/doc2520.pdf

It seems that NetMedia have also chosen to upgrade the BX-35 to the same
chip, although this is the first I have heard about it - it is not at
all apparent on their website.

Mike
http://home.austin.rr.com/perks/basicx/

>
> --- In basicx@basi..., "Ed" <esteyezz@y...> wrote:
> > Hi
> >
> > I have been doing some circuits with the BX-35 chip and one day I
> > noticed that the onboard programmer I use (the suggested bx35 hookup
> > schematic) didn't work well. Most of the times I got a EEPROM
> Checksum
> > error, which puzzled me a lot since I have been using the same
> design
> > over and over. After a different test I noticed that the chip wasn't
> > anymore the AT90S but the Atmega8535. I read the forum and I found
> > nothing in therms of different hook up for it, but my tested circuit
> > didn't work anymore so I tested an old AT90S and it worked again at
> > first!!
> >
> > I tryed then to program the Atmega on the BXDS and it worked fine so
> > something was wrong. Looking at the BXDS schematic I noticed that
> R47
> > is a 10k resistor instead of the 100k suggested on the Bx-35
> schematic
> > (the one that goes from the ATN line to ground).
> >
> > I did change the resistor on my circuit for a 10k and... it
> worked!!!
> > No more problems anymore with the onboard programmer.
> >
> > I just though I should let everyone know (just in case someone has
> > notised the same)
> >
> > Thanks
> >
> > Ed
>




The BX-24, BX-35, and BX-01 are really fine hardware. The new chip
is excellent.

I am sorry, but I cannot carry the ball on all the documentation.

The dilema here is that good engineers often are not great writers.
So, it takes having a writer on board to present their vision. It
appears that almost all, if not all the BasicX documentation is
Verion 1.0! [It really should have the version and date listed on
every page for errata control]. A lot of it needs tweaking and the
BX-01 probably could use the removal and rewrite of the PROTO board
section. [I have I/O power specifications for the BX-01 that
list "TBD". A lot of engineers might shy away from such substantial
omissions.]

Sadly, I am not a writer either and most of my day is running around
on a Yamaha 100cc motorscooter as I teach English at two universities
and various other institutions in Kaohsiung, Taiwan. Many people
feel that i should be writing books on teaching English as a second
language, but I never sit down long enough to do so.

I love my job as language and linguistics are quite challenging. And
I love these little microcontrollers. But, I am usually writing on
line at 1am [if at all] due to hobbies come after the rest.

I would be happy to feed information about errata to someone that
really gets it into a useful form, but demands on my time make doing
much more impossible. I certainly can help with language and
communication problems.

But, I do feel that this is a marketing challange and that once the
right writer is found, Netmedia will benefit from more sales as the
easier it is to get started with your product, the more people will
want to use your product first. You need to have someone that you
pay to do all this and that you can demand results. Hobbyist and
volunteers will never finish the job.

There are three kinds of information - Need-to-know; nice-to-know;
and trash. I am struggling to get much of the Need-to-know and I
sense that others are too. We like Netmedia and BasicX and want
more. But, there are gaps.

For instance, I have been looking at the BX-01 networking
documentation and for the life of me, I can't figure out how to set
up the wiring. Does this work as parallel data and address buses, or
is it some hybrid of serial data with parallel address buses? Or what?

Your engineers seem to know what they have and seem to think that
everyone can infer what is happening. But for the hobbyist, it is a
blank wall that is impossible to overcome. I have just given up for
now and looked at using the RAMsandwich.

I have much more use for a datalogger than a large network of
microcontroller anyway. But, there are people out there that use
automated irrigation for golf courses and large scale farming that
would be interested if you had some good examples of how to get
started. They even have automated suppliments of fertilizers that
need to be monitored and used on a remote basis. There may be a
market for your network in lieu of more expensive PCs at each site.

I wish you well, but I hope you can see the customer's point of view.
We will spend more for more.

G. Herzog
-- In basicx@basi..., Mike Perks <basicx@a...> wrote:
> G. Kramer Herzog wrote:
>
> > I just wish that NetMedia would coherently collect and publish
> > errata, as the documentation seems to never have been revised.
>
> The FAQ page on my website was supposed to be a collection of links
to
> useful previous messages in the Yahoo group but I never quite got
around
> to finishing it. I have downloaded all the previous messages from
Yahoo
> into text files for easier offline processing. So if anyone wants
to
> take up the challenge I can supply the text and publish the
results. I
> did recently update the Links on my website categorized into 14
> sections. I think everyone should find something useful.
>
> >
> > Also, the change over in processor name was due to substantial
> > improvements in design flaws in the older version. The new chip
is
> > less prone to SPI interface problems and about 10 other problems.
> > You may want to take a look at Amtel's Errata Sheet on the
changeover.
>
> The new BX-24p uses the ATMega8535. NetMedia claim a 26%
performance
> improvement. This has been verified by group members and we have
> surmised that all of this gain is due to the 2x faster SPI bus on
that
> chip. NetMedia also make some other improvements in the download
> procedure and made use of a few other chip improvements. The Atmel
> application note explains the chip changes:
> http://atmel.com/dyn/resources/prod_documents/doc2520.pdf
>
> It seems that NetMedia have also chosen to upgrade the BX-35 to the
same
> chip, although this is the first I have heard about it - it is not
at
> all apparent on their website.
>
> Mike
> http://home.austin.rr.com/perks/basicx/
>
> >
> > --- In basicx@basi..., "Ed" <esteyezz@y...> wrote:
> > > Hi
> > >
> > > I have been doing some circuits with the BX-35 chip and one day
I
> > > noticed that the onboard programmer I use (the suggested bx35
hookup
> > > schematic) didn't work well. Most of the times I got a EEPROM
> > Checksum
> > > error, which puzzled me a lot since I have been using the same
> > design
> > > over and over. After a different test I noticed that the chip
wasn't
> > > anymore the AT90S but the Atmega8535. I read the forum and I
found
> > > nothing in therms of different hook up for it, but my tested
circuit
> > > didn't work anymore so I tested an old AT90S and it worked
again at
> > > first!!
> > >
> > > I tryed then to program the Atmega on the BXDS and it worked
fine so
> > > something was wrong. Looking at the BXDS schematic I noticed
that
> > R47
> > > is a 10k resistor instead of the 100k suggested on the Bx-35
> > schematic
> > > (the one that goes from the ATN line to ground).
> > >
> > > I did change the resistor on my circuit for a 10k and... it
> > worked!!!
> > > No more problems anymore with the onboard programmer.
> > >
> > > I just though I should let everyone know (just in case someone
has
> > > notised the same)
> > >
> > > Thanks
> > >
> > > Ed
> >



> ... Your engineers seem to know what they have and seem to think that
everyone can infer what is happening. But for the hobbyist, it is a
blank wall that is impossible to overcome...

Two observations, G.

- This group, while the "official" Basic-X group, was not created by Netmedia, is not maintained by Netmedia and is not a direct conduit to Netmedia. The engineers here are those of Basic-X users, not of Netmedia. There is no guarantee that anyone at Netmedia will see your messages here. If you have a comment that is specifically addressed to Netmedia, this is not the forum for it; mailto:Support@Supp... is.

- Some hobbyists have done the impossible. Tom


Good point and understood.

I have sent emails to support and answers have been slow or none.

I suppose that I am considered a 'high maintenance' user as I am just
learning and have a lot of enthusiasm. I had assumed that regularly
Netmedia reads this forum and is possibly the moderator.

In sum, I will put my soap box away and try to share more insight and
successes.

In passing....

Mike Perks has some recent material [2005] called BasicX Internals
and it is written in six parts. This seems a fresh approach and an
attempt to create a comprehensive overview of the programing
environment.

--- In basicx@basi..., Tom Becker <gtbecker@r...> wrote:
> > ... Your engineers seem to know what they have and seem to think
that
> everyone can infer what is happening. But for the hobbyist, it is a
> blank wall that is impossible to overcome...
>
> Two observations, G.
>
> - This group, while the "official" Basic-X group, was not created
by Netmedia, is not maintained by Netmedia and is not a direct
conduit to Netmedia. The engineers here are those of Basic-X users,
not of Netmedia. There is no guarantee that anyone at Netmedia will
see your messages here. If you have a comment that is specifically
addressed to Netmedia, this is not the forum for it;
mailto:Support@B... is.
>
> - Some hobbyists have done the impossible. > Tom


I have finally gotten to understand the two distinct configuratons of
the BX-01.

One is an multiplexed parrallel data/address bus design to interface
with the RAMsandwich. It ties up most of the I/0 lines that might be
of use for other things and still retains at least one RS-232 port
for connecting to the outside world.

The other is a RS-485 network that has IN SOFTWARE, NOT HARDWARE
64K address nodes. This does not require any parallel configuration
and has at least 22 I/O lines freely availible for other uses.

The only signifcant difference from the BX-24/35 in the second form
is that there are some limitations on I/O because the BX-01 processor
is a simpler chip in comparison to the BX-24/35.

Nonetheless, the RS-485 networking can really extend the usefulness
of your projects. You can run wire without a repeater to 4000 feet
and more with repeaters.

So, it you want to put a weather station on your roof, you are not
limited to RS-232 distances of 50 or less feet. --- In basicx@basi..., "G. Kramer Herzog" <hwanghetw@y...>
wrote:
> Good point and understood.
>
> I have sent emails to support and answers have been slow or none.
>
> I suppose that I am considered a 'high maintenance' user as I am
just
> learning and have a lot of enthusiasm. I had assumed that
regularly
> Netmedia reads this forum and is possibly the moderator.
>
> In sum, I will put my soap box away and try to share more insight
and
> successes.
>
> In passing....
>
> Mike Perks has some recent material [2005] called BasicX Internals
> and it is written in six parts. This seems a fresh approach and an
> attempt to create a comprehensive overview of the programing
> environment.
>
> --- In basicx@basi..., Tom Becker <gtbecker@r...> wrote:
> > > ... Your engineers seem to know what they have and seem to
think
> that
> > everyone can infer what is happening. But for the hobbyist, it is
a
> > blank wall that is impossible to overcome...
> >
> > Two observations, G.
> >
> > - This group, while the "official" Basic-X group, was not
created
> by Netmedia, is not maintained by Netmedia and is not a direct
> conduit to Netmedia. The engineers here are those of Basic-X
users,
> not of Netmedia. There is no guarantee that anyone at Netmedia
will
> see your messages here. If you have a comment that is specifically
> addressed to Netmedia, this is not the forum for it;
> mailto:Support@B... is.
> >
> > - Some hobbyists have done the impossible.
> >
> >
> > Tom