Gus Calabrese wrote: > Is there any news about Scenix? I recently received a batch of SX18AC-DP and SX28AC-DP parts with 9810 and 9812 date codes (vs. the earlier 9747, 9749, and 9750). I had hoped that the new parts would fix some of the errata, but I called Scenix today and was told that the 9812 parts are still revision 2.1 of the silicon. Not that any of the errata is particularly terrible, except #5, which says that the Vdd must be limited to the range of 4.5V to 6.5V. We wanted to run the part on 3.3V. Apparently the part was designed for operation at 3.3V, and they aren't aware of any reason why it won't run at 3.3V, but they haven't finished characterization. I also asked whether the maximum operating frequency had to be derated for a lower supply voltage, and was told that it should work to the maximum specified frequency (50 MHz) even at 3.3V. I don't know whether the SX-Key is supposed to work with 3.3V target systems. A few weeks ago Parallax told me that they thought they'd have the debugging features of the SX-Key working in a few weeks, but didn't make a firm commitment. As everyone is probably aware, they plan a free upgrade for owners of the pre-release SX-Key. Parallax told one of my associates that the debugging features wouldn't work on the "current" SX parts. However Scenix claims that debugging features *will* work with the rev 2.1 parts. Someone doesn't have their story straight. I discussed with Scenix several observations that Richard Ottosen and I have made regarding the in-circuit programming: 1) The ISP header should be keyed to prevent connecting the programmer backwards or off-by-one. 2) The SX chip doesn't automatically reset reliably after programming. I don't know if this is a fault of the chip or the SX-Key, but Richard and I suggest that the MCLR pin should be added to the standard ISP header. Scenix was receptive to the first suggestion (keying), and said that they have been considering that themselves. They had observed the reset problem but didn't think it was important; they claim that usually people have a manual reset button or jumper on a target system. I maintain that I don't want a manual reset button; the target chip may be buried deeply inside a box, and I want the programmer to reset the chip automatically and reliably. EPROM programmer vendors provide this, and "real" ICE products provide this, so it seems reasonable to require that programmer/debugger products for the SX should provide it. Anyhow, what's the harm in adding the pin to the spec for the connector? Parallax wouldn't have to actually support it if they didn't want to. Anyhow, Richard wrote up a proposal for the enhanced ISP header, which I will put on my web site soon. The gist of it is to take the existing four-position header and add two positions, of which the fifth is a key (no physical pin), and the sixth would be MCLR. > Is anyone using them? Yes. Cheers, Eric http://www.brouhaha.com/~eric/scenix/