Anyway to read RS485 Probes with Micrologix 1200?

JJ1234 · Jan 19, 2022

I have an existing Micrologix 1200 that read some analog values off of temp and dewpoint sensors. I want to replace the current sensors with new ones that communicate using rs485/Modbus RTU. I would need to communicate with two different sensors on different networks so I would need two rs485 ports for the PLC.

Is there anyway to do this? It doesn't appear that there are any expansion cards for the PLC that do RS485. Is there any workarounds? like a card that reads values off of RS485 and transmits an analog signal to the PLC?

Thank you.

danw · Jan 19, 2022

JJ1234 said:
Is there any workarounds? like a card that reads values off of RS485 and transmits an analog signal to the PLC?

I am unaware of any commercial module that is a Modbus RTU master, with a serial port and an analog output.

The typical approaches uses a small PLC that acts as a Modbus master and outputs analog value; like the Red Lion CPU box that has the Data Station comm capabilities (Modbus master) plus an AO card or a Automation Direct CPU with an AO card and whatever it takes to get the serial interface. But then you have two PLCs to maintain instead of one . . .

Ken Roach · Jan 19, 2022

I would need to communicate with two different sensors on different networks so I would need two rs485 ports for the PLC.

Why are the sensors on different networks ?

Modbus RTU is a simple master/slave system, so the MicroLogix 1200 would have to be the only Master for each network. If the sensors are already being polled by a Modbus RTU master, then you've got a different problem.

There was never a 1762-series I/O specialty module with a serial port, so if you really do need more than one Modbus port your only option would be a multi-port gateway like a Red Lion Data Station Plus.

If you just mean that the wiring segments would need to be physically isolated from one another, you could use an RS-232 to RS-485 multi-point isolator. I have had very good luck with the Commfront HUB485-4 isolated multi-point serial converter.

VERY IMPORTANT: only the MicroLogix 1200 Series C controllers support Modbus RTU master communications. The Series A and Series B supported Modbus RTU slave only.

JJ1234 · Jan 20, 2022

Ken Roach said:
Why are the sensors on different networks ?

Modbus RTU is a simple master/slave system, so the MicroLogix 1200 would have to be the only Master for each network. If the sensors are already being polled by a Modbus RTU master, then you've got a different problem.

There was never a 1762-series I/O specialty module with a serial port, so if you really do need more than one Modbus port your only option would be a multi-port gateway like a Red Lion Data Station Plus.

If you just mean that the wiring segments would need to be physically isolated from one another, you could use an RS-232 to RS-485 multi-point isolator. I have had very good luck with the Commfront HUB485-4 isolated multi-point serial converter.

VERY IMPORTANT: only the MicroLogix 1200 Series C controllers support Modbus RTU master communications. The Series A and Series B supported Modbus RTU slave only.

I believe they have to be on separate networks due to the distance I would have to run the cables. The manufacture only recommends 100ft for one network and the sensors are about 100ft away from the PLC in different directions.

THey are Vaisala DMP7 Dew Point and Temperature Probes

danw · Jan 20, 2022

The Vaisala recommendation of 30m (98ft) for an RS-485 cable is really unusual (pg 31 paper, 33 pdf, user manual)

You might call them and confirm that, because it violates the conventional EIA-485 standard: 4000' limit at lower baud rates (higher baud rates work over shorter distances). It can't be a driver power issue; all instrument manufacturers buy a commercial RS-485 chip and I've never heard of a RS-485 chip with such a limited the distance (there's no market for it because 485 is used over 232 for its distance and multidrop functionality).

I suspect that Vaisala is trying to negate the need for a terminating resistor at the end of the 485 bus because they're using M12 connectors and there's no clean, easy way to install a terminating resistor with M12 cordsets. They could have put in an internal switch-enabled termination resistor, but apparently not. Experience shows that the lower baud rates and short distances, the need for a terminating resistor is not critical and they're probably taking advantage of that with their 100 ft bus distance limitation.

With the instrument hanging on the 'end' of an M12 cordset, that cordset length becomes a 'stub' when the network end of the cordset is connected to the incoming bus twisted pair and the outgoing bus twisted pair. If you multidrop the two probes on one bus, shorten the instrument cordset as short as possible to get to a junction box where you can install a 120 ohm terminating resistor, and make connections for the probe, incoming twisted pair and outgoing twisted pair.

Or you could just get the Indigo 500 transmitter, put it where the PLC is, run the 100 foot cables in opposite directions to each probe, and take 4-20mA out of the transmitter for each probe to the PLC and skip the Modbus thing altogether.

JJ1234 · Jan 20, 2022

danw said:
The Vaisala recommendation of 30m (98ft) for an RS-485 cable is really unusual (pg 31 paper, 33 pdf, user manual)

You might call them and confirm that, because it violates the conventional EIA-485 standard: 4000' limit at lower baud rates (higher baud rates work over shorter distances). It can't be a driver power issue; all instrument manufacturers buy a commercial RS-485 chip and I've never heard of a RS-485 chip with such a limited the distance (there's no market for it because 485 is used over 232 for its distance and multidrop functionality).

I suspect that Vaisala is trying to negate the need for a terminating resistor at the end of the 485 bus because they're using M12 connectors and there's no clean, easy way to install a terminating resistor with M12 cordsets. They could have put in an internal switch-enabled termination resistor, but apparently not. Experience shows that the lower baud rates and short distances, the need for a terminating resistor is not critical and they're probably taking advantage of that with their 100 ft bus distance limitation.

With the instrument hanging on the 'end' of an M12 cordset, that cordset length becomes a 'stub' when the network end of the cordset is connected to the incoming bus twisted pair and the outgoing bus twisted pair. If you multidrop the two probes on one bus, shorten the instrument cordset as short as possible to get to a junction box where you can install a 120 ohm terminating resistor, and make connections for the probe, incoming twisted pair and outgoing twisted pair.

Or you could just get the Indigo 500 transmitter, put it where the PLC is, run the 100 foot cables in opposite directions to each probe, and take 4-20mA out of the transmitter for each probe to the PLC and skip the Modbus thing altogether.

Thank you for the help! Sounds like that will work.

Anyway to read RS485 Probes with Micrologix 1200?

JJ1234

Member

danw

Lifetime Supporting Member

Ken Roach

Lifetime Supporting Member + Moderator

JJ1234

Member

danw

Lifetime Supporting Member

JJ1234

Member

Similar Topics