Equip your Genius
Welcome to the Parallax Discussion Forums, sign-up to participate.
Learn with BlocklyProp
2020-10-10 - 10:22:23
edited 2020-10-10 - 13:52:20
I need to do something like this...opinions or better solutions? Prolly need schottky diodes for 3.3v(?)
720 x 419
2020-10-10 - 10:41:15
What's the intended use ? How's the master to know which slave ack's or is it just one way master to slave without the ack ?
2020-10-10 - 11:15:17
It's a primitive arrangement but even with silicon diodes it will work fine although the pullup will probably have to be much lower unless you were running a low baud rate and a short distance. Alternatively you can use digital pnps to drive the common tx high and then use a digital npn on the master rx.
If they are Prop chips then you can just combine the rx and tx and have the one wire bus that is driven by the selected slave with the others floating their tx though.
2020-10-10 - 13:24:13
Why not switch to open drain. Instead of driving outa use dira. At least saves the diodes.
2020-10-10 - 17:49:28
Strictly a speak-when-spoken to arrangement. I have these devices, distributed on a full duplex 422/485 bus but I now want to have 3 or 4 on one small board and it seems silly to have a MAX491 for each device.
no kidding....as I was typing the OP, I was thinking "Peter's gonna suggest OW again"
Not all devices are Props and so I'm stuck with RX/TX.
I actually prefer this arrangement because, no matter what, the master can always talk. The slaves are running motors and should one get its knickers in a twist and tie-up the shared response line, the master can still broadcast a shutdown command.
Many thanks for the input
2020-10-10 - 23:30:44
edited 2020-10-10 - 23:31:32
Oh, now that makes a difference if it normally does hookup with drivers but there is a big difference between RS422 and RS485 though. True RS485 will be multidrop half-duplex and RS422 is typically FD P2P although both have been hacked to death in practice. Sounds like a FD protocol is being used though and your diodes plus stronger PUs if necessary should work just fine.
btw, I use to write all kinds of complicated protocols for RS485 but I have been quite happy with the simple ping-pong type protocol I have been using on P1 and mean to implement it on P2 which should be very easy. There is never any lockup and with a P2P virtual connection on a multidrop line both ends have equal access and are self pacing and establishes a virtual "full-duplex" connection.
2020-10-11 - 07:54:33
485/422 hack: I hear you.
Probably better described as a FD, multidrop, 4-wire network, based on 422/485 transceivers.
I really enjoy using this stuff