Thanks for the reply,

This makes things more clear.

I'm actually using a chip with datecode AB0004AB. Is this a
production revision with all final changes?

I'm using the SXKey52.exe ver 1.16.

I can debug and run the chip with the SXKey and SXkey52.exe, ver
V1.16 so my development can continue. I am converting a program from
SX28 to SX52 and adding new features. Still I can't get the
oscillator to start on its own with V1.16. I tried with V1.12 and
then the oscillator started and runs fine on its own. Does this mean
that my chip is a older chip?

I have been informed that V1.19 of SXKey52.exe will be released by
Parallax today (still nothing though). Hope this fixes everything!

Is the resistor between OSC1 and ground still recomended in the
production release? Mine works fine without a resistor. I have a
10MHz crystal, two 15pF caps and a 5.6Mohm parallell resistor. I am
told that a 1Mohm resistor should be used, will check this later. I
also tried a 12 MHz ceramic resonator, but didn't get a good result
(didn't spend a lot of time on this).


> Ruben,
>
> To program the SX48/52 with the Parallax SX-Key requires different
> software revisions, depending on whether you are using engineering
> samples or production release variants. The datecodes up to and
> including 9931 are early engineering samples that require SXKey52.exe
> rev 1.12. All datecode devices after that, including mass production
> release, are supported by SXKey52.exe rev 1.19+.
>
> To answer some of the problems mentioned in earlier emails:
> -The software version designed for the SX18/28, SXKey28L.exe, will
> definitely not work on the SX48/52. -The reset issues you mention are
> fixed in the production release SX48/52 devices and their tools.
> -Also, Parallax did release some versions of SXKey52.exe that
> incorrectly disabled the OSC crystal buffer by default, thus turning
> off the oscillator.
>
> The reason for the different SX silicon revisions are that the SX48/52
> FUSE/FUSEX register bit assignments have been modified to resemble the
> FUSE/FUSEX register bit assignments for the currently shipping version
> of SX18/28. Along the way, new features have also been added, such as
> an RTCC overflow flag, PORTRD setting to ensure the I/O rather than
> the latch is used for read-modify-write instructions, better use of
> BANK instruction, crystal buffer disable capability, among other
> improvements. The differences, as well as information regarding
> upgrading from SX18/28 has been made availablein the application note
> AN15, available on the Scenix website.
>
> Regarding an eledged Parallax/Scenix 'hissy spat'... This could not be
> further truth. Parallax continues to work very hard towards supporting
> the SX devices, as also Advanced-Transdata. Parallax has been
> reluctant to support all SX variants from a single IDE as does
> Advanced-Transdata, but is currently working a complete overhaul of
> their IDE to provide all the features necessary. Advanced-Transdata on
> the other hand, already supports alot of these features, including
> (but not limited to) multiple include files, external clock
> synchronization, external asynchronous breakpoints, 8K trace (non
> real-time), 2.5-5.5V in system programming and debugging, and support
> for all variants (new and old) of SX. Especially for SX48/52, Parallax
> has chosen not to support the ability to use BANK 0, or make the MODE
> addressed readable registers available while debugging. As the
> Advanced-Transdata tools do support the full capability of the SX48/52
> in debug, they have been given a qualified status by Scenix. Parallax
> is still working on there software to support these features.
>
> Clearly, the Parallax and ATD tools cannot be compared equally, not by
> features or by price. The Parallax tool is mature, we love it and we
> use it every day. At it's price point, there is no other competitor,
> but when more features are required, ATD has the tool, albeit at a
> higher price. Other tools will also be released by other vendors, with
> even more features, and higher prices - you pay for what you get.
>
> In addition, Scenix has recently posted the software and schematics
> necessary to build your own in-field programmer on the website.
>
> Best Regards,
> Stephen Holland
> Sr. Applications Engineer
> Scenix Semiconductor Inc.
> 1330 Charleston Road,
> Mountain View, CA 94043
> (650) 210-1500
> Email: mailto:stephen.holland@scenix.com
> Web: http://www.scenix.com
>
>

==============================
Ruben Jvnsson
AB Liros Elektronik
Box 9124, 200 39 Malmv, Sweden
TEL INT +46 40142078
FAX INT +46 40947388
ruben@pp.sbbs.se
==============================