Post by thread:Reset Curcuitry (Paul)

> > I've got a SIMILAR type of question..... The databook says that when > powering up a PIC, Vdd must start from 0V. I've found this, unfortunately > to be correct. Is there any way that you guys can think of to implement > an inexpensive circuit to bring the power to 0V QUICKLY after power is > disconnected so that you can QUICKLY power back up instead of waiting for > the PIC to drain your power supply caps? > Use a DS1231 to get a reliable Power-On-Reset. It's esecially useful when you use the 16C71 whose POR doesn't work... Anyway; this way is probably cheaper than convincing the power supply to go down to 0V before power-up. /Tomas spam_OUTtomasTakeThisOuTopq.se

> Use a DS1231 to get a reliable Power-On-Reset. It's esecially useful > when you use the 16C71 whose POR doesn't work... > Anyway; this way is probably cheaper than convincing the power supply > to go down to 0V before power-up. I can't find these in any of the catalogs I have with me - how much do they cost in low quantities and how may pins do they have? -- -- Paul Greenwood -- (.....pabloKILLspam@spam@austin.ibm.com) Any small object that is accidentally dropped will hide under a larger object.

The suggestion of a DS1231 is a good one, but...... I've had dreadful luck with Dallas Semiconductor parts. They've gotten a reputation for "cool stuff that blows up in production volumes". I might suggest the Maxim MAX705 reset chip, since it's available, cheap and as far as I know, reliable. Not to bash DS excessively, but I've just not had really great luck with their stuff. Multiple part types, over a period of years, at several different companies. Be forewarned.... Mark G. Forbes forbesmKILLspamcsos.orst.edu R & D Engineer

> > > Use a DS1231 to get a reliable Power-On-Reset. It's esecially useful > > when you use the 16C71 whose POR doesn't work... > > Anyway; this way is probably cheaper than convincing the power supply > > to go down to 0V before power-up. > > I can't find these in any of the catalogs I have with me - how much do they > cost in low quantities and how may pins do they have? 8 pins but I don't know the pricing. It's not very expensiv though. /Tomas .....tomasKILLspam.....opq.se

> The suggestion of a DS1231 is a good one, but...... OK - there has been a misunderstanding - the problem I'm having is with Vdd not starting at 0V. I have an RC circuit to delay the reset line and I think that my problem is with Vdd. >From the 1993 "Microchip Data Book", on page 2-25. In the bottom right-hand corner, the paragraph reads: "To summarize, the on chip power-on reset is guaranteed to work if the rate of rise of Vdd is no slower than 0.05V/ms. It is also necessary that the Vdd start from 0V." This DOES seem to hold true for what I am doing. If I reduce my Vdd to 3 volts, the uC stops functing. I can then increase to 5V and ramp up reset through an RC network and it does not reset correctly. -- -- Paul Greenwood -- (EraseMEpablospam_OUTTakeThisOuTaustin.ibm.com) The Crown is full of it! -- Nate Harris, 1775

>This DOES seem to hold true for what I am doing. If I reduce my Vdd to 3 >volts, the uC stops functing. I can then increase to 5V and ramp up reset >through an RC network and it does not reset correctly. > -- Paul Greenwood -- (pablospam_OUTaustin.ibm.com) Why, pray tell, do you reduce Vdd to 3 volts and then increase back to 5V? And by what means? - JohnR -- John R. Haggis @spam@haggisKILLspamnetcom.com Millennium Research (408) 269-1814 vox (408) 269-9323 fax

> > I've had dreadful luck with Dallas Semiconductor parts. > They've gotten a reputation for "cool stuff that blows > up in production volumes". I might suggest the Maxim concur. I love all of Dallas' stuff but I have had to repair too many boards that have had one of their chips on it. Except for thier clocks. later John ----------------------------------------------------------------------------- John Johnson Team OS/2 member | KILLspamjohnsonjKILLspambga.com | RemoveMEjohnsonjTakeThisOuTutdallas.edu C&H Technologies VXI Plug N' Play member ----------------------------------------------------------------------------- And the Seventh version of OS/2 raised into the air its bow of blue steel and cried," It. Is. Done." Around him lay Bill Gates and Microsoft apps. Their evil in this world at an end. Revelations of InfoWorld, Oct 11 1994

> Why, pray tell, do you reduce Vdd to 3 volts and then increase back to 5V? > And by what means? > > - JohnR I can't remember if I answered this already so here goes (again). I have a regulator on my board and a big capacitor (10uF) to keep out noise, spikes, etc. Well, I plug in my device, the cap. charges to 5V quickly. When I unplug the device, the cap. takes forever to discharge because the PICs use so little power. The problem is that the device seems to not work if - say, it's plugged from one computer to another or the cable is lengthened or for whatever reason I unplug and replug the device. I've been shorting out Vdd and Gnd every time so Vdd will start from 0V but I'd like a better solution. -- -- Paul Greenwood -- (spamBeGonepablospamBeGoneaustin.ibm.com) All true wisdom is found on T-shirts.

> Dallas makes a 3 pin TO-92 device that will hold reset until > power is stable. It will hold reset low when power drops below > usable levels. I have used it without any problems. My data > books are at home and I don,t remember the part number or the > threshold levels. Call your local Dallas rep. or dist. and > SOLVE this problem. > > Brian I checked this out but I really need something like a 5V regulator that will pull its output to 0V when the input falls below a threshold. -- -- Paul Greenwood -- (TakeThisOuTpabloEraseMEspam_OUTaustin.ibm.com) All true wisdom is found on T-shirts.

{Quote hidden}> Paul, sorry to ask you this, but how can you get your /MCLR input > down to a secure level of 0.15 Vdd (as stated as the max level for a > clear "0" on the /MCLR pin) by just having an RC network? As you > describe it, Vdd is reduced to 3V and that is what you get at the > /MCLR pin as well. It doesn't help if the 3V level is delayed an > hour, it won't go below 0.15 Vdd or ...? > With the DS1231 you will get a /MCLR tied down to GND as soon as the > input voltage drops to 4.5V or 4.0V (settable) and that level will be > held until the voltage is stable above the 4.0 or 4.5V limit. It also > has a small extra delay to allow any eventual crystals to start up > sufficiently. > Wouldn't this solve your problem? At any level, rise time and so on > you will be granted a secure power-on-reset. > As mentioned in other replies Maxim has lots of uP watchdogs. The > reason I use the DS1231 from Dallas is that I have quite the opposite > experinces with Dallas as the person who suggested Maxim. Maxim makes > a lot of good stuff but mainly they just make a LOT of stuff. In the > last three years they have produced 10.7 cm of just "New Releases > Data Book"s (measured, not estimated nor exaggerated). I have the > continous fear that they just will unsupport a device without prior > notice and then you have to change your construction to use another > circuit. > > /Tomas You know, you're right - when I have a "brown-out", my reset line does not actually reset since my Vdd line doesn't go to 0V. I, however, found the line in the databook that said Vdd must start from 0V for a proper bring-up and stopped looking at the problem. I guess that, I MAY be able to ignore the line in the databook and use the reset pin and the DS1231. I'll give it a try - testing - testing - testing..... Thanks...!!! -- -- Paul Greenwood -- (RemoveMEpabloTakeThisOuTaustin.ibm.com) All true wisdom is found on T-shirts.