----- Original Message ----- From: "kg4pid" <> To: <> Sent: Wednesday, November 17, 2004 11:13 PM Subject: [piclist] Re: Serial I/O on the 16F628, PDP's > > > In the mid 70's I was in the navy and worked on a flight simulator > that was mostly wire-wrapped and primarily 7400 series TTL. Had 40 > cockpits (real ones from salvaged aircraft) with simulated radar > covering the eastern half of the U.S. Computers were Data General > Nova 800 with 16K magnetic core memory (you could see the bits). When > it was built it was the only one like it so it was wire-wrapped. When > we ran diagnostics and had a failure we used a AM radio to listen to > the circuits. We would exercise portions of the circuit and use the > radio to listen for different patterns and compare good ones to bad > ones. It was amazing what could be done given the low amount of > memory. All programming was in assembly using punchcards. > The good ole days. Even better were the first transistor computers. When I was a student working for English Electric LEO computers at their Park Royal factory in the early 60s I designed an audible debugger for the new LEO IIIF (IIRC) which divided the pulses by a factor of 100 and output them to a loudspeaker. This was bolted onto the side of the CPU cabinet and allowed engineers to listen to the program running. The earlier LEO III was much slower and didn't need the divider. I also modified a register module (it was actually called a staticiser) so that it worked as a logic probe - driving little neon bulbs that were part of the module and indicated the output logic state. LEDs hadn't been invented then. The engineers plugged this unit into the rack and poked around with a probe I made from a Bic ballpoint pen. I think I cut a few tracks and wired the probe up to a gate input, rather than a flip-flop, and took the gate output to the neon. Computers were more interesting in those days. Leon

In the mid 70's I was in the navy and worked on a flight simulator that was mostly wire-wrapped and primarily 7400 series TTL. Had 40 cockpits (real ones from salvaged aircraft) with simulated radar covering the eastern half of the U.S. Computers were Data General Nova 800 with 16K magnetic core memory (you could see the bits). When it was built it was the only one like it so it was wire-wrapped. When we ran diagnostics and had a failure we used a AM radio to listen to the circuits. We would exercise portions of the circuit and use the radio to listen for different patterns and compare good ones to bad ones. It was amazing what could be done given the low amount of memory. All programming was in assembly using punchcards. The good ole days. --- In , Robert Rolf <robert.rolf@u...> wrote: > jrem wrote: > > wire wrap is fricken' cool. I had never seen it before, and an > > engineer suggested it when I told him I was making my own PCB's. > > It's is fast, efficient, and FAA approved. Great for one-offs, it's > > one step up from breadboard, comes apart easy, just awesome for > > prototyping, IMHO. The only downside is those pesky wire wrap > > sockets aren't cheap. > > Have you found any w-w sockets that can mount QFP chips like the > higher end PICs? <G> > And if you want to generate huge amounts of EMI, WW is the way to > go. > > The PDP5, PDP6, PDP 8, PDP 9, were all wirewraped collections > of 'standard' DTL card modules. With a clock speed of a few > 100's of kHz, EMI wasn't an issue. > > It's quite amazing what one can do with 4k words of core... > > R

jrem wrote: > wire wrap is fricken' cool. I had never seen it before, and an > engineer suggested it when I told him I was making my own PCB's. > It's is fast, efficient, and FAA approved. Great for one-offs, it's > one step up from breadboard, comes apart easy, just awesome for > prototyping, IMHO. The only downside is those pesky wire wrap > sockets aren't cheap. Have you found any w-w sockets that can mount QFP chips like the higher end PICs? <G> And if you want to generate huge amounts of EMI, WW is the way to go. The PDP5, PDP6, PDP 8, PDP 9, were all wirewraped collections of 'standard' DTL card modules. With a clock speed of a few 100's of kHz, EMI wasn't an issue. It's quite amazing what one can do with 4k words of core... R

--- In , "Leon Heller" <leon.heller@d...> wrote: > ----- Original Message ----- > From: "jrem" <jrem123@y...> > To: <> > Sent: Wednesday, November 17, 2004 1:53 PM > Subject: [piclist] Re: Serial I/O on the 16F628 > > > > > > > > --- In , "Leon Heller" <leon.heller@d...> > > wrote: > >> ----- Original Message ----- > >> From: "jrem" <jrem123@y...> > >> To: <> > >> Sent: Wednesday, November 17, 2004 12:32 AM > >> Subject: [piclist] Re: Serial I/O on the 16F628 > >> > >> > >> > > >> > > >> > doh! I didn't tie mclr high on my new wire wrap board. > >> > >> I didn't think anyone still did wire-wrapping. I last wire- wrapped > > a board > >> about eight years ago, when I was feeling nostalgic. It was an old > > design of > >> mine using an MC68008, updated to use a CPLD for the glue logic, > > instead of > >> the six or so TTL chips I used originally. > >> > >> Leon > > > > wire wrap is fricken' cool. I had never seen it before, and an > > engineer suggested it when I told him I was making my own PCB's. > > It's is fast, efficient, and FAA approved. Great for one-offs, it's > > one step up from breadboard, comes apart easy, just awesome for > > prototyping, IMHO. The only downside is those pesky wire wrap > > sockets aren't cheap. > > I used to do it sitting in an armchair whilst watching TV. 8-) > > I just design and make a PCB these days. Single-sided boards are easy to > make with rudimentary facilities - my equipment cost me about 70 GBP: a > second-hand laser printer for the artwork, home-made UV exposure unit, a few > plastic containers and a small drill and stand. > > Leon yish, I use Eagle, a second hand laser printer ($25 on ebay, $25 to deliver, $40 for a refurb cartridge), print onto magazine paper, iron on to clean boards, center punch, drill with a dremel, etch with FeCl, and clean up with automotive brake cleaner (aerosol). They look damn good. Traces to fifteen thou. Wire wrap is faster for prototyping/one-outs, though, IMO, and more secure/bug-proof than breadboard. Breadboard is great for trying stuff out, but not good for in-field stuff.

----- Original Message ----- From: "jrem" <> To: <> Sent: Wednesday, November 17, 2004 1:53 PM Subject: [piclist] Re: Serial I/O on the 16F628 > --- In , "Leon Heller" <leon.heller@d...> > wrote: >> ----- Original Message ----- >> From: "jrem" <jrem123@y...> >> To: <> >> Sent: Wednesday, November 17, 2004 12:32 AM >> Subject: [piclist] Re: Serial I/O on the 16F628 >> >> >> > >> > >> > doh! I didn't tie mclr high on my new wire wrap board. >> >> I didn't think anyone still did wire-wrapping. I last wire-wrapped > a board >> about eight years ago, when I was feeling nostalgic. It was an old > design of >> mine using an MC68008, updated to use a CPLD for the glue logic, > instead of >> the six or so TTL chips I used originally. >> >> Leon > > wire wrap is fricken' cool. I had never seen it before, and an > engineer suggested it when I told him I was making my own PCB's. > It's is fast, efficient, and FAA approved. Great for one-offs, it's > one step up from breadboard, comes apart easy, just awesome for > prototyping, IMHO. The only downside is those pesky wire wrap > sockets aren't cheap. I used to do it sitting in an armchair whilst watching TV. 8-) I just design and make a PCB these days. Single-sided boards are easy to make with rudimentary facilities - my equipment cost me about 70 GBP: a second-hand laser printer for the artwork, home-made UV exposure unit, a few plastic containers and a small drill and stand. Leon

--- In , "Leon Heller" <leon.heller@d...> wrote: > ----- Original Message ----- > From: "jrem" <jrem123@y...> > To: <> > Sent: Wednesday, November 17, 2004 12:32 AM > Subject: [piclist] Re: Serial I/O on the 16F628 > > > > > > doh! I didn't tie mclr high on my new wire wrap board. > > I didn't think anyone still did wire-wrapping. I last wire-wrapped a board > about eight years ago, when I was feeling nostalgic. It was an old design of > mine using an MC68008, updated to use a CPLD for the glue logic, instead of > the six or so TTL chips I used originally. > > Leon wire wrap is fricken' cool. I had never seen it before, and an engineer suggested it when I told him I was making my own PCB's. It's is fast, efficient, and FAA approved. Great for one-offs, it's one step up from breadboard, comes apart easy, just awesome for prototyping, IMHO. The only downside is those pesky wire wrap sockets aren't cheap.

----- Original Message ----- From: "jrem" <> To: <> Sent: Wednesday, November 17, 2004 12:32 AM Subject: [piclist] Re: Serial I/O on the 16F628 > > > doh! I didn't tie mclr high on my new wire wrap board. I didn't think anyone still did wire-wrapping. I last wire-wrapped a board about eight years ago, when I was feeling nostalgic. It was an old design of mine using an MC68008, updated to use a CPLD for the glue logic, instead of the six or so TTL chips I used originally. Leon

doh! I didn't tie mclr high on my new wire wrap board. Thanks to all for the suggestions, though, I know the mclr thing wasn't the only error. Don't know what I'd do with out you folks. The bits be flippin' on da scope! --- In , "rtstofer" <rstofer@p...> wrote: > > John, > > I went ahead and rewrote the code. It works on my demo board with a > 20 MHz resonator at 38400 baud: > > list pF628A,r,nH > #include <p16f628a.inc> > > errorlevel -302 ;suppress "not in bank 0" message > > FALSE EQU 0 > TRUE EQU ! FALSE > > ifdef __16F877 > BootLoader EQU TRUE > else > BootLoader EQU FALSE > endif > > if ! BootLoader > __CONFIG _CP_OFF & _WDT_OFF & _BODEN_OFF & _PWRTE_ON & > _HS_OSC & _LVP_OFF > endif > > byte equ 1 > word equ 2 > > cblock 0x20 > count1 :byte > endc ; end c block > > ;This code executes when a reset occurs. > ResetCode: > if ! BootLoader > org 0x0000 > clrf STATUS > clrf PCLATH > else > org 0x0003 > endif > > goto setup ;go to beginning of program > > ;------------------------------- - > -------- > ;This code executes when an interrupt occurs. > > ORG 0x0004 ;place code at interrupt > vector > > InterruptCode: > > ;============================== > ; setup code > ;============================== > setup > movlw b'00000111' > movwf CMCON > movlw b'00000000' > movwf INTCON > > banksel PORTA > clrf PORTA > clrf PORTB > > banksel TRISA > movlw 0xFF > movwf TRISA > movlw b'00000110' > movwf TRISB > > bsf RCSTA,SPEN ; don't think this is necessary > > banksel TXSTA > bsf TXSTA,BRGH > movlw d'31' > movwf SPBRG > bcf TXSTA,SYNC ; don't think this is necessary > bsf TXSTA,TXEN > > banksel RCSTA > bsf RCSTA,SPEN > > ;============================== > ; main loop > ;============================== > main > banksel TXREG > incf count1,f > movf count1,w > movwf TXREG > > banksel TXSTA > > main1 > btfss TXSTA,TRMT > goto main1 > goto main > > end > Richard

yes, table 3-1 and 3-2 are indispensable too bad adobe won't let you split a page, I'm forever switching between indexes and pages, etc., etc., not just for pic's but mysql and others. the php manuals are way cool, totally interactive, very easy to work with online. The pic style manuals are easiest to work with when printed and bound, IMO. , --- In , "rtstofer" <rstofer@p...> wrote: > > > Don't know how I missed the bit about SYNC - it is required. > However, it is cleared at RESET and unless set somewhere else, it > shouldn't need to be touched. > > I find tables 3-1,2,3 & 4 quite helpful in determining initial > conditions as well as naming and locating register and bits within > registers. > > Yes, banksel, pagesel, bankisel are quite useful in that they help > document exactly which bank/page is being selected and, to some > extent, why. > > All of the various bits are given names in the include file. I > personally prefer using the bit name to the bit number.

Don't know how I missed the bit about SYNC - it is required. However, it is cleared at RESET and unless set somewhere else, it shouldn't need to be touched. I find tables 3-1,2,3 & 4 quite helpful in determining initial conditions as well as naming and locating register and bits within registers. Yes, banksel, pagesel, bankisel are quite useful in that they help document exactly which bank/page is being selected and, to some extent, why. All of the various bits are given names in the include file. I personally prefer using the bit name to the bit number.