??? 12/03/04 03:51 Read: times |
#82447 - RS485 Contention Responding to: ???'s previous message |
Yes this contention question is a BIG issue. I work on software for industrial / construction equipment joysticks for an ITT division here. These joysticks have from 1 -> 4 microcontrollers inside that are bussed together with an RS485 port that is used for a calibration and testing interface port. The microcontrollers each interface to the bus via the very nice ESD protected transceivers from Linear Technology under part number LT1785 in an SO8 package. The driver connects on the microcontroller side to the UART TxD and RxD lines. There are also a couple of port pins that control the DE (driver enable) and -RE (receiver enable) pins.
A problem was found with some of the joysticks in that when they power up they were drawing up to 7 or 8 times the normal current draw from the +12 volt input which typically uses 150 mA. The high current surge of an ampere or more was lasting for about 0.6 seconds and over current loading the power source from our customers engine control system and raising an fault emergency. The current surge was first noticed when the equipment was tested at -20C and -40C. It turns out that the current surge was lasting for as long as the width of the reset pulse from the reset supervisor plus the time it took for the microcontroller to go through its power up self checks and then initilize the hardware including the port pins that controlled the LT1785 chip. Further testing revealed that the LT1785 chips were going into current contention on the RS485 bus becasue all the drivers were enabled at once the bus...but do to variation of the start-up times of the distributed regulators on each of the microcontroller boards, one would become a low driver before another and massive currents would flow!! Linear Tech support told me the parts were safe from burning themselves up but that the short circuit current could be as high as 250 mA!! The fix for the problem was rather simple. We arranged for a pullup resistor on each -RE pin and a pulldown on each DE pin. After that correction no more current surges. Michael Karas |
Topic | Author | Date |
RS485 Contention. | 01/01/70 00:00 | |
485 ISP | 01/01/70 00:00 | |
RS485 'U' & ISP | 01/01/70 00:00 | |
the crux | 01/01/70 00:00 | |
Flashmagic !!!! | 01/01/70 00:00 | |
RS485 contention | 01/01/70 00:00 | |
Cygnal F12x | 01/01/70 00:00 | |
Reply to Erik | 01/01/70 00:00 | |
FM< half duplex | 01/01/70 00:00 | |
Reply to Erik | 01/01/70 00:00 | |
bus direction | 01/01/70 00:00 | |
A few comments | 01/01/70 00:00 | |
comnments on comments | 01/01/70 00:00 | |
comments^3 | 01/01/70 00:00 | |
boot NIT | 01/01/70 00:00 | |
An example worth 3E8h words :-)![]() | 01/01/70 00:00 | |
A fundamental query - Erik | 01/01/70 00:00 | |
testing | 01/01/70 00:00 | |
RS 485 Contention | 01/01/70 00:00 | |
RS485 Contention | 01/01/70 00:00 | |
Switch ? | 01/01/70 00:00 | |
RS485 contention | 01/01/70 00:00 | |
Possible solutions. | 01/01/70 00:00 | |
Thanks Erik. | 01/01/70 00:00 | |
SOLUTION | 01/01/70 00:00 | |
Thanks but thats not the solution | 01/01/70 00:00 |