On Mon, Feb 06, 2006 at 09:48:40AM +0800, Chen Xiao Fan wrote: > > > -----Original Message----- > > From: piclist-bounces@mit.edu > > [mailto:piclist-bounces@mit.edu] On Behalf Of Byron A Jeff > > Sent: Monday, February 06, 2006 9:15 AM > > > > On Sun, Feb 05, 2006 at 01:59:15PM -0500, David VanHorn wrote: > >> So do it at a higher level. Send a hex file to a chip on > >> the other side of the USB interface, and let it do the pin > >> wiggling, reporting back success or failure as appropriate. > > I believe GTP-USB (with WinPIC800) is doing this since WinPIC800 > support simple serial Port based JDM as well using the same > host software. It is using the 18F2550 as the USB interface and > not those common USB-to-serial chips. I am not 100% sure though > since I have not tried it and both the firmware and the > host software are closed source. Well having closed source is not too helpful. > > > > As I said before the USB/serial (or possibly USB/parallel) > > cables represent compromises as they are easily obtainable > > and have more than one purpose. > > > > So the job becomes how to use such cables as interfaces. > > However, they do not have the same level of accessibility as > > the native ports they replace. > > > > I believe that it's a problem worth figuring out. > > > > I believe this is an interesting problem to figure out as > well. I think USB-to-serial converter will be more > common and more useful solution than USB-to-parallel > converter. As do I. > I agree with you it is worth to investigate > the possibility to produce a reasonably simple programmer > based on USB-to-serial converter and it needs only to support > popular PICs like 16F88/16F876A/18F2520/18F2550 to be > useful to hobbyists. Well once the hardware interface is in place it can theoretically program any flash based PIC. > And a bootloader solution is not bad > as well for 16F88/16F876A/18F2520/18F2550. Bingo. And that's the primary purpose I envision. As a bootloader user, I find the utility of a traditional programmer is greatly diminished. I am aware that it limits my chip selection and in particular locks me out of an ICD style debugging interface. But I've personally had good success using bootloaders with a serial debugging backchannel to complete projects. I also envision that when I delve into USB stuff, that having that debugging backchannel that is separate serial debugging interface separate from the USB, which the target is virtually guaranteed to use, will be important. So to me a programmer is little more than an ad hoc code dumper. However, that USB to serial interface will retain its value beyond simply serving as an interface for the ad hoc code dumper. > In fact I like the bootloader function of PICkit 2 very much. > > Still I think the chicken-and-egg problem is not as important > as you might think. Anyway, one needs to invest some time and > money to get into the PIC world. Programmer is only a > small part of the total investment. To me a multimeter and > an oscilloscope are much more essential than a programmer. > Guess I am a bit biased since I am an electronics engineer. I came in from the software side. Most of my hardware debugging is done with a logic probe and a multimeter. I rarely break out the Oscope. The chicken and egg problem is critical when one thinks that the programmer function is redundant. I received a PS+ when I attended a local Atlanta seminar several years ago. I used the PS+ for awhile then put it away. The next time I wanted to do PIC stuff, I first had a hard time locating the programmer, then when I finally located it, it didn't work. I ended up scratch building a simple Tait style programmer and loading WLoader into a 16F877. Development was a breeze. I also realized that the programmer was ad hoc. We all have different world views. What's important or not to each of us is colored by experiences such as the one I outlined above. Over the years I've had many users successfully utilize the Trivial programmer in its various forms. It served a need for each of my users. I'm simply extending the DIY development model that has already been established. It's certainly not for everyone. However, I feel that it's a niche that still needs to be filled even as we move to PC interfaces that are so complex that no hobby user has any hope of interfacing to it with simple hardware. BAJ -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist