Post by thread:[PIC:] Is it better to use Shift Registers or lots

I'm working on a project that will need a couple of inputs and around 16 outputs. The outputs will change at a few hertz at the fastest. Is it a better design to use an 8-pin PIC and send the outputs serially to a couple of shift registers or a 28 pin PIC and use the outputs directly. It seems like the shift register solution is better design because it just feels wrong to waste so many IO pins on something that simple, but the larger pic means less components, less soldering, and simpler software. Jason -- 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

I would go with the shift register so if something happens, you're not sinking all that current through the µController, but rather through a cheap chip. -Shawn Jason S wrote: {Quote hidden}>I'm working on a project that will need a couple of inputs and around 16 >outputs. The outputs will change at a few hertz at the fastest. Is it a >better design to use an 8-pin PIC and send the outputs serially to a couple >of shift registers or a 28 pin PIC and use the outputs directly. > >It seems like the shift register solution is better design because it just >feels wrong to waste so many IO pins on something that simple, but the >larger pic means less components, less soldering, and simpler software. > >Jason > >-- >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 > > -- 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

Jason S wrote : > I'm working on a project that will need a couple of inputs > and around 16 outputs. The outputs will change at a few > hertz at the fastest. Is it a better design to use an 8-pin > PIC and send the outputs serially to a couple of shift > registers or a 28 pin PIC and use the outputs directly. Small series (like 1->10->100) : use simplest programming. Large series (maybe from about 1.000 units ??) : focus more on component cost... It sounds as you are doing a single unit project ? If so, use a 28-pin PIC. > > It seems like the shift register solution is better design > because it just feels wrong to waste so many IO pins... And instead wasting time on more complicated programming ? But anyway, without knowing the number of units planned, it's close to impossible to answer I'd guess... Jan-Erik. -- 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

Jason S wrote: >I'm working on a project that will need a couple of inputs and around 16 >outputs. The outputs will change at a few hertz at the fastest. Is it a >better design to use an 8-pin PIC and send the outputs serially to a couple >of shift registers or a 28 pin PIC and use the outputs directly. > >It seems like the shift register solution is better design because it just >feels wrong to waste so many IO pins on something that simple, but the >larger pic means less components, less soldering, and simpler software. > > I often use a smaller PIC and a shift register. For driving lamps and relays I use a UCN5821. Also provides a degree of isolation from PIC <-> load. The code to drive the S/R is simple and well worth it. David... -- ___________________________________________ David Duffy Audio Visual Devices P/L U8, 9-11 Trade St, Cleveland 4163 Australia Ph: +61 7 38210362 Fax: +61 7 38210281 New Web: http://www.audiovisualdevices.com.au ___________________________________________ -- 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

OK - if the cost of the different PIC parts actually tracked the number of pins, there might be some reason to use the S-R. AS it is, the cost of a PIC chip doesn't change that much from 8 to 18 to 28 to 40 pins. And, as you said, lots fewer parts and an easier programming job. I usually use a 40-pin PIC for anything more complicated than a logic chip replacement - it is surprising (not!) how often those extra pins find uses on future projects, Roy J. Gromlich {Original Message removed}

On Monday, Apr 12, 2004, at 17:24 US/Pacific, Jan-Erik Soderholm wrote: >> It seems like the shift register solution is better design >> because it just feels wrong to waste so many IO pins... > > And instead wasting time on more complicated programming ? > Ah, but you should only have to do the 'complicated programming' ONCE, and then you can put as many shift registers as you want out there with little additional effort. And there are LOTS of types of shift registers, including high current drivers and fancy SPI chips. OTOH, more chips add a lot of complexity to a design when it comes to actually getting it built. And space issues, too. I'd lean toward using a bigger chip unless the rest of the circuit is already complex, or unless the next-bigger chip gets expensive (relatively speaking.) On the third hand, I have my "stock" of PICs and I'd rather use one of those (which tend to be SMALL pics) than have to order something new. BillW -- 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

I'll have to be a fence-straddler here. Oftentimes there's no substitute for more I/O pins in terms of hardware complexity, and often the larger chips are the only ones that offer the other peripherals I need. OTOH, I'm frequently doing stuff with remote displays, etc, where I want to minimize the number of wires running from one board to another. In this case, shift registers are invaluable. For instance, right the moment, I'm running an LCD, several LEDs, and several buttons at the end of ~4 feet of ribbon cable, using only 8 conductors. If I dedicated a wire to each I/O, I would end up with ~20 wires... yuk! As with most things in EE, it looks like the answer to this question depends on your individual problems and design goals. - Robert -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

On Mon, Apr 12, 2004 at 04:57:55PM -0700, Jason S wrote: > I'm working on a project that will need a couple of inputs and around 16 > outputs. The outputs will change at a few hertz at the fastest. Is it a > better design to use an 8-pin PIC and send the outputs serially to a couple > of shift registers or a 28 pin PIC and use the outputs directly. > > It seems like the shift register solution is better design because it just > feels wrong to waste so many IO pins on something that simple, but the > larger pic means less components, less soldering, and simpler software. KISJ, Your last statement says it all: ... the larger pic means less components, less soldering, and simpler software. It would be the winner in my book. BAJ -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

On Tue, Apr 13, 2004 at 02:00:41PM -0400, Byron A Jeff wrote: > On Mon, Apr 12, 2004 at 04:57:55PM -0700, Jason S wrote: > > I'm working on a project that will need a couple of inputs and around 16 > > outputs. The outputs will change at a few hertz at the fastest. Is it a > > better design to use an 8-pin PIC and send the outputs serially to a couple > > of shift registers or a 28 pin PIC and use the outputs directly. > > > > It seems like the shift register solution is better design because it just > > feels wrong to waste so many IO pins on something that simple, but the > > larger pic means less components, less soldering, and simpler software. > > KISJ, > > Your last statement says it all: > > ... the larger pic means less components, less soldering, and simpler software. > > It would be the winner in my book. I need to thrown in the caveat of the small production run. I always tend to see questions like this from the hobbyist perspective. So to rephrase: If you're only going to build a handful, save yourself the hassle and use the larger chip. However if you designing for 100,000 widgets, then the cost contraints will determine the best route to take. BAJ -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

pic microcontroller discussion list wrote: > As with most things in EE, it looks like the answer to this question > depends on your individual problems and design goals. I think this about sums it up. It's always a tradeoff. Cost has been mentioned already, but I feel like jumping in on this thread Cost and complexity are almost always interrelated. Cost is more than just component cost. I personally think that component cost only becomes relevant when you start talking about 10's or 100's of thousands of boards. I look at it in terms of design time e.g. engineering salary. In this case, DigiKey price for 100 16F872-ISO is $2.81. 100 12F629-ISO plus 74HC164 is 1.32. Difference is $1.49 times 100 units = $149 dollars. How much more it costs to design in a shift register depends on many factors, but I doubt its more than 30 minutes to an hour even if you include PCB layout. As long as this isn't your first PIC project, clocking out serial data is trivial. So the cost difference shrinks to maybe $100. Its pretty much a negligible difference at low volumes in any case. But at higher volumes, the component costs start to make a difference. In general, the simpler the design, the less can go wrong, the less time it takes to design and develop, and the cheaper it becomes. Unless you are doing it as a hobbyist, time is not free. So, there is no right answer to the original question, just guidelines :) -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

>Thanks for all the replies. It seems like it was a better >question than I thought. Well that is one of the beauties of this list, there is a large amount of people experienced in a wide field of expertise who are only too willing to make their knowledge available. >This is a one-off hobby project, so component cost isn't much >of issue. I'll go with the bigger pic because it is easier. For this sort of use the advise given pretty well holds. The only thought that I had was you may still need to think about how big a chip you need in terms of pins available, as once you go beyond the 40 pin chips they are all surface mount, which can be a fright if your not used to them. I don't think you are using that many pins yet though. -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email spam_OUTlistservTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body

At 12:31 PM 4/13/2004, Eisermann, Phil [Ridg/CO] wrote: >Cost and complexity are almost always interrelated. Cost is more >than just component cost. I personally think that component cost >only becomes relevant when you start talking about 10's or 100's of >thousands of boards. I look at it in terms of design time e.g. >engineering salary. In this case, DigiKey price for 100 16F872-ISO >is $2.81. 100 12F629-ISO plus 74HC164 is 1.32. Difference is $1.49 >times 100 units = $149 dollars. Another consideration to add to this discussion: power handling. If you need to drive large loads, it can be easier to use a SR with built-in high current output stages. I'm thinking of the TPIC6595 family but there are others. Once you have SPI code working, it becomes just another tool in the arsenal. Use it if it makes sense to do so, use a larger package PIC if not. I use a LOT of 12c508 parts and using shift registers to provide the extra I/O that I need seems completely natural. But, then again, I've been using shift registers since long before I even heard of PICs. That said - for a one-off project being built from scratch (not able to use an existing board design), I'd probably use one of the larger PICs *if* all the I/O fits into one part. dwayne -- Dwayne Reid <.....dwaynerKILLspam@spam@planet.eon.net> Trinity Electronics Systems Ltd Edmonton, AB, CANADA (780) 489-3199 voice (780) 487-6397 fax Celebrating 20 years of Engineering Innovation (1984 - 2004) .-. .-. .-. .-. .-. .-. .-. .-. .-. .- `-' `-' `-' `-' `-' `-' `-' `-' `-' Do NOT send unsolicited commercial email to this email address. This message neither grants consent to receive unsolicited commercial email nor is intended to solicit commercial email. -- http://www.piclist.com hint: To leave the PICList piclist-unsubscribe-requestKILLspammitvma.mit.edu