Re: TI-H: TI-Modem


[Prev][Next][Index][Thread]

Re: TI-H: TI-Modem




Yeah, Atmel has had all sorts of unexpected problems with the
AVR line. The AT90S8515 was supposed to clock at 20 MHz but they
had to cut it back to 8 MHz, and for some reason one of my beta AVR
chips just croaked. All of its I/O lines are floating at some
psuedo-low impedence or something and it appears to be totally dead.
My heart isn't broken over a $3 chip failing after probably more than
1,000 programmings but, still, things like that shouldn't happen.

Bryan

Grant Stockly wrote:
> 
> >heheh, but I/O is cheap, why bother... I noticed that the AVRs
> >(at least the early ones I got) sometimes puke when you change from
> >output to input and this can mess up TI routines in rare cases.
> 
> whats the number on the bottom of yoru chip?  i'd order a few samples from
> marshall and get the new ones.  there were problems the atmel mailing list
> was discussing about the EEPROM and I/O failing to work properly on hte
> newer ones...
> 
> >It can certainly be done, but you probably still need the pull-ups.
> >You may also want (just to be safe) current limiting resisters so
> >that the TI isn't outputing a high while the AVR is outputting a low.
> >This won't hurt the TI link port (because of the way it works,
> >wired-AND) but it may cause problems for the AVR's relatively weak
> >output drivers.
> >
> >I'd love some more info tho, I may need two more I/O lines in the
> >future. Although I think the module port is robust enough as-is.
> >
> >Bryan
> >
> >Grant Stockly wrote:
> >>
> >> No...  I just said taht because I can...  :)  Anyway, you can rewrite your
> >> tiwork routines to only use 2 I/O wires instead of 4...  Just cannge the
> >> data direction on the fly.  It works almost too well for me...
> >>
> >> What I am going to release is a botboard type thing (like the 68HC version)
> >> but for the calc.  THat would be cool...
> >>
> >> >Nevertheless, I wish you would stop calling your designs "EII"s.
> >> >You're my original subcontractor, but that doesn't mean you can slab
> >> >an AVR and a mem chip together and have an authentic EII. An EII
> >> >clone, maybe, but I doubt you have the latest firmware source.
> >> >
> >> >That said, I'd like to get on with my life and hope you aren't
> >> >still mad at me. As for providing info on my website, I'm not
> >> >trying to milk this project for thousands of dollars. I want
> >> >to recover my costs and hundreds of hours of effort, definitely,
> >> >but I see no point in keeping information secret.
> >> >
> >> >Sorry,
> >> >
> >> >Bryan
> >> >
> >> >Grant Stockly wrote:
> >> >>
> >> >> Bryan, I don't have any of the source you gave for the EuP or the EuP92
> >> >> driver.  What you posted on your web site and what I can find on the
> >> >> internet is overly sufficient to build a "clone".  Remember that e-mail
> >> >> about 3 months ago that you never responded to in which I asked you
> >>for the
> >> >> source because my HD died?
> >> >>
> >> >> >The hell you can... remember that thing called an NDA? Also...
> >> >> >I wish you'd stop using the phrase 'EII' when naming your
> >> >> >projects. Any modification you make to the 'EII' firmware
> >> >> >that I sent you makes it into your code. Mel handed the EII
> >> >> >over to me, like it or not, and its going to get pretty damn
> >> >> >confusing if we're both working on a device by the same name.
> >> >
> >> >--
> >> >Bryan Rittmeyer
> >> >mailto:bryanr@flash.net
> >> >http://www.bridges.edu/horizon/
> >
> >--
> >Bryan Rittmeyer
> >mailto:bryanr@flash.net
> >http://www.bridges.edu/horizon/

-- 
Bryan Rittmeyer
mailto:bryanr@flash.net
http://www.bridges.edu/horizon/


Follow-Ups: References: