On Thu, 18 Sep 1997 06:21:55 -0500 "Bryson, William G (Bill)" writes: >I have a project that requires 4 16C84 pics to receive data from one >source >serially. That is to say the Port B lines would have to be either >connected >directly together or via a OR series of gates. The premise is that I >want the >4 pics to be listening for commands sent by one common radio and then You can connect them together directly and use the tris registers to turn off the outputs on all chips except the one that should be transmitting. It sounds like you have a reasonably airtight method to be sure only one chip will attempt to transmit at a time. > >I have one main controller Pic #1 that will be controlling the power >of each of the other 3 Pics. . If you're going to be turning the power off then the tris register method won't work because the protection diodes in the unpowered PICs will draw power from the line. Maybe the A4 pin could be used this way because it doesn't have a protection diode. But it seems a bad idea to apply voltage to an I/O pin without powering the chip. Not so obvious is that the *input* line to the PICs also needs to be isolated. This can be done with either diodes or high-value resistors. If diodes are used the PORTB pull-ups could be used instead of pull-up resistors at each PIC. Isolation of the outputs could be OR gates or diodes. The data-out bus would have a pulldown resistor. PICs which are not transmitting would either have no power, put the output pin in tri-state input mode, or hold it low. The complexity of this may outweigh the tiny savings of putting the PICS in sleep, WDT off. It may be good to control the power PICs 2, 3, and 4 as a group though, for those long times when the entire box is idle. This will also restart them with every use so they won't be able to "hang up". Isolation would then be requred only between the group of power- controlled PICs, PIC#1, and the radio. It could be done with no extra hardware by using PIC#1 to relay the radio data out to another port that is connected to the bus for the other PICs. >off the power of the other 3 Pics to conserve power. Pic #1 is the >only >Pic that >will always have power and be ether sleeping or awake. Thanks for your >help... If the radio is half-duplex, presumably PIC #1 is always controlling whether it transmits or receives. If not, then the transmit/receive line needs to be similarly isolated so it can be controlled by any PIC. Systems like these generally use much more power in the radio receiver than anywhere else (usually, more than even the radio transmitter since it operates only very intermittently). If the protocol allows it, have PIC #1 turn the radio receiver off most of the time and only wake it up occasionally to see if a message is about to come in. The simple implementations of this require more power for the transmitter of the message, since it needs to send a "preamble" long enough to ensure that the receiver is awake and listening.