On Fri, Aug 16, 2002 at 03:00:06PM +0300, Vasile Surducan wrote: > I must admit, this mail it's an excellent methode to learn from > other's ( or just from your's ) mistakes > My ideeas would be: > > On Thu, 15 Aug 2002, Byron A Jeff wrote: > > > design. After some discussion several points reached a state of consensus: > > > > * The product would be fully assembled, and not a kit. > > Error. A beginner must spend some time to familiarize himself with the > components. There are a lot of people which can't make difference between > a resistor and a capacitor. I bet there are someone on the list too... > A beginner will be much happy when something made with his own hands will > work. > Let his joy open ! The joy opens at the end when there is success. Unfortunately there is frustration during the journey. This is a compromise based on several factors: * Sean's factory can deliver fully assembled product. * Beginners will have more early sucess and less support issues with a fully assembled product. * This product isn't solely for beginner. Note that I changed the subject on this summary. More experienced users really don't need to be bothered with the hassle of assembling a kit. New users will have ample opportunity to work with components on the breadboard and eventually to start building their own boards. Kit building is an impediment to early success. That very fact is what started this whole thread. > > > * The PIC assembly would be the target language for development. > > as you wish... there are other people which are doing the same for jal, c > or basic And we would welcome any or all of them. However It's impossible to support all of them. The most available, and mode widely known is plain PIC assembly. > > > * The product would carry some type of breadboard prototyping. > > this "some tipe" could be the standard 100 mili_inch metalised wholes. > and must fill at least one half ( or more ) of eurocard ( 160x100mm ) Fine as long as we can meet cost constraints. > > > * The product would have a mechanism for programming PIC parts > > this is not really necessary. It's possible to create a less beatifuly > programmer than someone found already on the net, why to force the user to > cry ? > build a standard Tait lpt programmer, point to the user where are the best > available software and let him to choose ! The answer for this is simple: under your suggestion the user would have to have two tools instead of one. I belive that I already made clear that unless this device is the PICbase concept, that projects will need to be programmed separate from the unit. It just doesn't make any sense to force the user to buy or build another programmer when that functionality can so easily be integrated into the existing unit. > > > * The product would have some type of I/O expansion capability > > right ! I'm just curious what kind of "some type" you'll invent... Personally I'd keep it simple: a dual in line 40 pin connector. Why? Because then any old ordinary PC IDE cable can be used to connect the expansion port to an external board. They are cheap and plentiful. > > > * The product will carry a CD composed of software, tutorial documentation, and > > projects geared towards the novice to intermediate user. > > this is the most difficult part, I must wish you good luck because it's a > lot of work to do here... Yes. But I don't think we have to develop the world. One project per I/O device plus a few basic PIC concepts should be more than enough for starters. > > > > > However there has been considerable debate as to what specifications the > > product should be design toward. There are three basic forms proffered: > > > > * The traditional style inexpensive PIC programmer with one or more > > programming sockets (Roman, Olin, Jason, Sean). > > Without sockets on the standalone programmer ! ICSP for the beginners and > for the advanced. It's an important feature which someone forgot it. > The same ICSP > style socket ( female of course ) on the PCB. Beginners will use always > only flash PICs. Honestly I won't even attempt to address this. It seems to be the exact opposite of what all the programmer folks want. Could one of you folks take a crack at it please? > > > * A small assembled module, the PICbase, that consists of a PCB populated with > > a PIC, regulator, oscillator, and I/O expansion and programming interface. > > (Geert) > > I'm disagree. A plugin module may be inserted on the prototype board as I > already said, in 21 different modes + connectors which will grow up the > price. Personally I was less concerned about the price and more concerned about the lack of standardization... > > > * An all in one prototyping board, the PICLIST DESIGNER (PLD), that contains > > a PIC16F877A, a set of common I/O devices, and I/O expansion port, and an > > ICSP programming interface. (Byron) > > I like this, but needs an 16F628 too, No it doesn't. Primarily because they cannot be self programmed. > 877 is to difficult from the very begining. No it isn't. Especially not at the beginning. Each has the same basic peripherals and the same instruction set. The 16F877A has both comparators and A/D converters so it can pretty much behave exactly as a 16F628. Remember that the user isn't building the unit, it comes assembled. So there's no mechanical or wiring complexity to differentiate the parts. > I'll discard also the IO expansion port. Let only a numbers > of metalised wholes which will copy the PIC socket. What can do > a wrong designed expansion port you may see on Lascar's products. > Why to be angry on a socket when you can be the same looking at wholes ? Several reasons: breadboards are expensive. They are prone to mechanical breakdown. They have high capacitance due to their construction. They often do not have solid connections. They have their issues. > > > The prelude: > > I wish you success ! Thanks from all of us. BAJ -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics