After hearing or reading yet another comment about how the '84 is great because you don't have to wait for the UV erase cycle, I am finally going to add my 2 cents worth. Due to the method I use when designing with a windowed PIC, I don't find the '84 any faster to work with. First I use a Datarase II UV eraser, which will erase a 12CXXX or 16CXXX windowed PIC in 3-4 minutes (I haven't tried it with a 17CXXX part so I don't know how long that will take). Forget all those homebrew and long-winded commercial erasers, at about $50 US, the Datarase is the only way to go. How much is your time worth? If an erase time under 4 minutes is not fast enough for you, here's the other thing I do to eliminate the wait time. When I order the windowed version of the PIC I am working with, I buy TWO. With 2 PIC's instead of 1, while I am testing the results of the last compile with one of them, the other is in the eraser. By the time I have tested the last iteration, tweaked the code, and compiled it, the device in the eraser is blank. I simply pop the test PIC out of circuit, put it in the eraser, put the blank one in the programmer, then into the circuit to start the cycle over again. There simply is no wait time. The only benefit the '84 still holds is the ability to be programmed in-circuit, but I am willing to bet that most designers rarely do ISP during development of a circuit. ISP really comes into its own for field upgrades, but I doubt many people use it during development. Of course, I could be wrong. Add this to the fact that once the design is finished, the OTP parts are cheaper than an '84, and it becomes hard to justify using an '84 just to avoid the UV erase cycle. CIAO - Martin. ______________________________ Reply Separator _________________________________ Subject: Re: Baud timing- How close do I have to be? Author: pic microcontroller discussion list at Internet Date: 5/27/98 8:02 PM Lads - After some due consideration and scouring of the microchip databook I have decided that I am going to re-design a PIC into my battery powered product and the current processor is getting heaved. The wake-up on port b <7:4> and the super low current in sleep mode made up my mind - I'm looking at the 16C620 series as a likely candidate but the 16F84 could be used for much of the development without having to play the "wait to erase" game... The above having been proclaimed, I have a couple of questions that may sound trivial for those that have been designing with this processor already but I figure I'll pass them by in the interest of coming up the curve quicker - 1) Is it just my mis-reading, but does the brown out enabled, turn the pic into a current HOG in sleep mode? i.e. from 1uA to 300+uA? 2) I will be powering the system from 2 lithium cells and would like to detect a low battery without having to use a big$, low dropout, low Quiescent I regulator. I'm thinking of doing the following and wonder if it sounds reasonable; - Dedicate a port pin to the driving of a zener diode - Feed this reference into one of the analog comparators on the '620 chip. - Feed the raw Vdd battery voltage into the "on-chip" programmable reference and tie it to the other comparator input line - programming this reference for 2.4-2.5 volts. 3) If I protect the electronics from a reverse battery with a diode right across the battery terminal, I will protect the electronics at the sacrifice of the batteries - this is ok - but - can a lithium coin cell source enough current to be of harm - i.e. heat, meltdown etc.? Thanks a million for any assistance people can provide. The micros look to be just the ticket for many projects! Lewis cobb@zeus.ee.unb.ca