Tue, 28 Mar 2017, : > ...short traces with well-defined ground... How would you specifically define that kind of ground? :-) The circuit would likely be on a dual sided PCB, unless that well-defined g= round requires another layer. No RF circuitry will be on the board. It wo= uld be used in an automotive 12-volt environement (which is noisy). The ci= rcuit would run at 3.0V. Tue, 28 Mar 2017, : >...noise should not be an issue (except maybe if=20 > the switches are changed while power is on and=20 > the state is being monitored - then you might=20 > find your hand coupling in environmental noise=20 > resulting in random readings. Of course, the DIP switches would be used when the board is powered and the= PIC running. The PIC will read the settings and the possibly trigger an o= utput to a 6.5-ohm speaker (siren) connected to the PIC via H-BRIDGE. Tue, 28 Mar 2017, : > Are you designing for ultra-low power? I am exploring very low power for various designs, yes. But before I ditch= the use of pull-ups for something else, I wish to explore all the merits a= nd demerits first. - - - - - - - - -=20 Wed, 29 Mar 2017, "Brent Brown" brent.brown@clear.net.nz > ...is it that you are concerned > about power consumption when a DIP switch is on? Yes. And I am simply exploring "how low can you go" with respect to averag= e current consumption. There are 8 DIP switches, so reducing all of them w= ith higher pull-up values is noticeable. Some people install many devices in a car. Each consumes current. But whe= n all those devices combine get too far past 30mA average current draw, unl= ess you drive your car every few days, there will be noticeable battery dra= in. As such, it is best to design a system that is as low power as possibl= e, but of course, with a design that is not so sensitive to noise that it f= lakes out all the time and becomes unusable or causes user frustration. Thank you for the discussion on switch debouncing. Since I am using DIP sw= itches that will be touched infrequently and then left locked into position= , the situation is a bit different than a keypad. But so far using a bread= board bench test with 100k-ohm pull-ups and with a perf-board mockup being = used in an actual car, I've not had switch debounce issues at all. But I'v= e not conducted significant testing with 1M-ohm pull-ups. I am in the expl= oration and "research" stage with regard to those values, hence my posts on= this list of late. - - - - - - - - - Tue, 28 Mar 2017, "Harold Hallikainen" harold@mai.hallikainen.org": > One thing that caught me with reading switches=20 > immediately after power up is that you have to=20 > allow enough time for the weak pull-up to actually > pull the pin up. With a 1M-ohm resistance pulling a PIC16F1508 (16MHz operation) digital I/O= pin to Vcc=3D3.0V, I am seeing rise times no longer than roughly 3usec. - - - - - - - - - Wed, 29 Mar 2017, David C Brown dcb.home@gmail.com: > Are there any (non-financial problems) associated=20 > with using a change-over switch, so the pin is=20 > connected to either Gnd or Vcc? Can you safely > connect an input to Vcc or is a series resistor=20 > required? Aren't "change-over" switches big and used in industrial applications? I a= m talking about a small PCB to be easily installed inside any car, hence th= e need for physically small DIP switches. - - - - - - - - - Wed, 29 Mar 2017, "Van Horn, David" david.vanhorn@backcountryaccess.com > You don't have to power the pullups unless=20 > you are reading the switches. I have been speaking about use of EXTERNAL pull-ups. Never in this discuss= ion have I spoken about the PIC's INTERNAL pull-ups (mainly because I can d= ecrease the current more by using my larger external pull-ups). Now if you are talking about switching a PIC's I/O pin from being an INPUT = to being an OUTPUT, and then back again, I've not been doing that. My DIP = Switches are connected directly to PIC digital I/O pins that are exclusivel= y defined as INPUTS. They are used for no other purpose and are never chan= ged to OUTPUTs (not by my code). - - - - - - - - - Now that I have replied to everyone in today's digest, I can only add that = I've been playing with a circuit on a benchtop breadboard with 12-Volt benc= htop PSU and with overhead fluorescent lighting and the noise that enters t= he circuit through my body, and I am not seeing any random DIP Switch state= fluctuations that affect circuit performance. But I have written to this = list to tap the experience of you gentlemen to see if some of you might say= , "Well, most of the time, those 1M-ohm pull-ups may be fine, but I one had= this problem that induced noise in the circuit and caused all the states t= o flip." I am just exploring the theoretical prior to building another moc= kup and doing my own testing. I don't want a situation where I do extensiv= e testing and find no issues but then mass produce the device only to find = noise is flipping the state due to the pull-up being too high. Then again,= there is good reason to strive for the lowest current consumption possible= , and the most! noticeable current draw in this design is the DIP Switches when flipped i= nto the short-to-Ground state (current flowing through the pull-up). Fact is, when Googling for information, I am not seeing anyone using 1M-ohm= pullups. Perhaps people are, but they are all secret designs out there. = :-) Heck, I don't even see people bragging about use of 100k-ohm pull-ups.= I guess everyone is content with the great noise immunity of 4.7k pull-up= s at the great expense of battery drain. Thanks, James W. --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .