| ??? 09/07/06 07:00 Read: times |
#123799 - not the cable alone Responding to: ???'s previous message |
Jez Smith said:
[...]the reason I don't believe this is that I have programmed many different devices in isp mode including other atmel devices such as AVR Does this personal experience include also programming of the AT89Sxx-s? If so, was it good or bad? Jez Smith said:
and I am of the opinion that half the problem lies in the AT89Sxx series devices because they are the only ones I have seen which cause any problems Look around better: there ARE problems with AVR's ISP, too; and some of it derives down to the "rusty nail" method, indeed. I don't follow PIC forums, but I suspect it will give a similar picture. However, I agree (and I said that in my initial post already) that it appears that these parts tend to have less problems when SPI-ISP. Jez said:
Anyway all the evidence leads me to susspect that its a problem with this particular part and not people using dodgy cables. Yes, partially; and I already tried to address the "cable is not enough" problem several times here. Nevertheless, I don't believe there are mysteries in electronics and I don't believe that the AT89Sxxx ISP is simply bad; I believe that there are circumstances under which these parts tend to have problems and these circumstances are not easy to get under control especially for newbies. I repeat myself: a professional might simply reject a part which is suspected to have problems of any kind and choose an another one, but that is not the position of the newbie in Nowhereland (Erik: what was that Danish name for this country?). On the other hand, the newbie, once mastering a simple beginner project, might learn better. This is exactly why I started this thread. One of the major concerns is the reset, powerdown, and proper sequencing of these plus the ISP signals, as addressed by Kai several times. Other might include an improper (different from AVR/PIC) "impedance" of the ISP signals, which in combination with improper cabling might cause troubles. There might be also problems with the MISO output level. There might be more of them, including sloppy oscillator design (we have threads here on the suspicious capacitors specified for the crystal oscillator on AT89S8253 and I read at least one report on problematic or no oscillator startup on this very part). I am trying to address all of these and also those issues I am not aware of - this is where and why I need the opinions and help of all of you. As I wrote, I have - positive only - experience with AT89Sxxx programming, but that was definitively NOT the LPT-cable-style programmer I was constructing/using. So, please, stop saying "it won't work", please be constructive. Thanks. Jan Waclawek |
