I have had a look at the CUMP project, and the Pocket project. I like very much the CUMP project, and it gaves me an idea for the progra= mmable voltage supply. Really, the Pocket is nice, but I just want a simple device conected to a= computer... What I have thought is having 2 analog sources, ground, and digital signa= l on each pin. For switching between them, I have thought on an analog multiplexer (4->1= ), and then a switching matrix is not required. I can control each pin if I have some shift registers connected, and each= output bit controls a pin of the programmer. By this way, you can simply= enter data on the=20 registers with two pins, one for data, and one for clock. A lot of hardwa= re is saved with them. And I think it's enough if the programmer has 48 pins... I will want to retrieve data from the device, so I think of having more s= hift registers, this time to latch the data from the device, connected in= a similar way as I told you=20 before... Each output pin of the register is connected to the digital entry of the = analog multiplexer, and the other inputs of the analog multiplexer are th= e same for all of them: the=20 two voltage supplies, and ground. This multiplexer is not required if you don't want to use this pin as a v= oltage output, so you can keep its place in the board, and plug it when y= ou really need it. I liked the idea of use 16F877 A/D converters to monitor output voltages,= and one pin to control the voltage, as shown in the CUMP project. It see= ms very nice. The interface to the PC is really simple, a serial port is enough. The co= mmands are simple:=20 - Get data: Gets all 48 pins data from the device. - Set data: Sets all 48 bits shift registers to output to the device. - Select source( pin, source): Select the source for each pin. - Set voltage (supply, voltage): Sets the voltage output on the supply. - Load program: Use as bootloader for uploading a new version of software= to the programmer control PIC. I think the design is simple, but the board will have a lot of connection= s, and maybe it will be difficult to route on a single layer board, but i= t doesn't seem too difficult to=20 do... and it will not required adapters unless the device you are using h= as other package than DIP. Regards, Carlos 26/05/01 14:42:25, Bob Ammerman escribi=F3: >Actually, for hobbiest purposes, you could build an 'everything' program= mer >in a much simpler manner. It would contain the following components: > >1: 2 or 3 programmable analog sources. > >2: A bunch of digital source/sense signals. > >All of these signals would then appear on a header. > >A scramble wired cable would shuffle them to the right pins. > >The other end of the scramble wired cable would go to the device socket. > >Now, for each unique pinout you'd just need to build up the correct scam= ble >wired cable (an hour or two's work). > >Bob Ammerman >RAm Systems >(contract development of high performance, high function, low-level >software) > > > > >----- Original Message ----- >From: "M. Adam Davis" >To: >Sent: Friday, May 25, 2001 10:46 PM >Subject: Re: [PIC]: Universal programmer > > >I've thought about something like this as well. I've since found that: >The majority of hobbyists stick to two or three parts (say, PIC, AVR, >EEPROM) and the cost of simple programmers for each total well under the >cost of building your own 'everything' programmer. >This may well be one of the reasons the CUMP did not do so well. > >To build such a programmer (as you are designing) you'd need each of the >50 pins to be switchable between at least 4 sources: >Digital (tri stateable, to simulate a no-connect) >Analog 1 >Analog 2 >Ground > >Though you really ought to have a dedicated connection to 'nothing', >which would make it 5 sources for each pin. The digital pin is tricky - >you either dedicate one digital pin for each socket pin and simple >switch it in (which means you can't use a simple matrix for this one, >use a cmos switch for each pin), or you use the larger 16x16 matrix cmos >switches and hope you never need to program with more than 10 digital >lines at any given time. > >There are software matrix chips you would end up using, say an 8x16. >You could use 3 of them for 48 pins (do you really need all fifty? Then >use 4...). They run around $25[US] each IIRC, then you'd need to add >the individual cmos switches (which are cheap - you could do the whole >thing with them but I wouldn't want to solder them, and it would be a >huge board...) The matrix/switch is shown below (the 8 vertical wires, >as suggested, would carry ground, 2 common analog signals, and 5 common >digital signals. The 8 wires on the left carry individual digital >signals (which could be matrixed as well, but you limit flexibility. >But then, few devices need more than 20 distinct signals to program, 50 >would be overkill.) >, as well as 8 which are left floating. The wires on the right signify >the socket.) > >Or, like many programmers, you add a simm socket, and then use wired >simm modules to hook everything up for you, but that isn't as cool. > >So for your time and energy, it is actually worth it to buy a good >"expensive' everything programmer. > >Otherwise you'd probably see home-built ones all over the web. > >-Adam > > G A A D D D D D > N N N I I I I I > D 1 2 1 2 3 4 5 >Digital6_______ __|_|_|_|_|_|_|_|________Socket Pin >Floating_______ / | | | | | | | | >Digital7_______ __|_|_|_|_|_|_|_|________Socket Pin >Floating_______ / | | | | | | | | >Digital8_______ __|_|_|_|_|_|_|_|________Socket Pin >Floating_______ / | | | | | | | | >Digital9_______ __|_|_|_|_|_|_|_|________Socket Pin >Floating_______ / | | | | | | | | >Digital10______ __|_|_|_|_|_|_|_|________Socket Pin >Floating_______ / | | | | | | | | >Digital11______ __|_|_|_|_|_|_|_|________Socket Pin >Floating_______ / | | | | | | | | >Digital12______ __|_|_|_|_|_|_|_|________Socket Pin >Floating_______ / | | | | | | | | >Digital13______ __|_|_|_|_|_|_|_|________Socket Pin >Floating_______ / | | | | | | | | > > >Carlos Nieves =D3nega wrote: > >>Hi all, >>I'm thinking in a design for an universal programmer, controlled by a P= IC. >>I'm a bit tired to search for a programmer for every device I plan to u= se, >so I would want to build one able to program all devices. >>My thought is that it will be "open hardware" (i.e. you will be allowed= to >use, build, and improve it at your own, like open-source software) with = an >open source >>software. >>I will make the software to program the devices I'm using, but other pe= ople >working with other devices can do the same... so in some time we could h= ave >a free >>universal programmer. >>I have one design in mind, with 50 pins to interface with the device. E= ach >pin is able to handle several sources: >> - Digital output. >> - Analog output (from two analog sources) (from 0 to 25V, digital >controlled). >> - Digital input (for reading data from the device). >>There will be two analog sources in the design, one will act as supply >voltage, and the other could be the programming voltage. >>I don't know if two only analog sources will be enough so I wanted to a= sk >you if you know any device needing more than one programming voltage (or >greater than >>25V). >> >>Please apologize me, because I know this mail is not about precisely ab= out >PICs, but I also know most of you have worked with many devices... >> >>Regards, >> >>Carlos >> >>-- >>http://www.piclist.com#nomail Going offline? Don't AutoReply us! >>email listserv@mitvma.mit.edu with SET PICList DIGEST in the body >> >> >> >> >> > >-- >http://www.piclist.com#nomail Going offline? Don't AutoReply us! >email listserv@mitvma.mit.edu with SET PICList DIGEST in the body > >-- >http://www.piclist.com hint: To leave the PICList >mailto:piclist-unsubscribe-request@mitvma.mit.edu > > > -- 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