Circuit Review

Tom,

In the article https://www.element14.com/community/external-link.jspa?url=https%3A%2F%2Fwww.sensorsmag.com%2Fcomponents%2Fimplementing-… , the circuit only uses a general purpose op-amp to perform the differential reading and the single ended power rail should work nicely with Arduino.  The shunt resistors are not that bad, it is just a matter of choosing as low a value as possible to get the voltage range you would like.  Going too high on the shunt resistor runs a risk of existing the drive or compliance voltage on the xmitter side (10K @ 20ma = 200V, which is most likely greater that the drive voltage, 5-28V).

Looking at the Libelium devices, they are using jumpers on the receiver to manage the possibility that the xmitter is un-powered, so you could likely do the same.  I have used any thing from 100Ω to 200Ω (2 to 4V max across the resistor).

Any generalize design will have a lot of variables when you are working with such an open protocol as 4-20mA loops.

Good luck!

Gene

Ok, I`m going to rework the 4-20 area utilizing the LM358. I`ll post the idea later. If you have time can you review this Rs485 circuit? Its intended to connect ModBus type devices.

I added high speed optical separation from the pic to Rs485 chip. I as going to add complete power separation on the RS485 in side but I stopped due to this being a test unit.

Tom,

I am not sure that the isolation that you have designed will do anything.  The 485 transceiver will take the hit if anything happens on the bus.  It is extremely unlikely that anything is going to get past the transceiver.  The only way that the opto-isolators would be helpful is if the transceiver was also isolated on a separate power supply (along with the transceiver side of the optos).

What I would do, is attempt to protect the transceiver with a transzorb(s).  One bidirectional transzorb between the A/B pins of the transceiver, along with optional transzorbs from the A/B pins to the power rails.  Also, I would land a resistor pad(s) across the A/B pins with a jumper, to allow the module to be a terminated if it is the end module on the bus (I like this better than placing the terminator at the connector).

Gene

Does these look right. At 12VDC I used a 16VDC TVS @ 5VDC I used an 8VDC unit

8VDC (Voltage - Reverse Standoff (Typ)

12VDC Voltage - Reverse Standoff (Typ)

I didn't add the terminator this model is designed to always be the Master. I will add this to the other slave model.

I added the 10ohm 1w resistor as an energy adsorbing resistor, always on the receiving side of the TVS

Question, I`ve also been questioning the optical part. You think its needed? Should I leave it? Or save the circuit cost and remove.

T

Tom,

This is a little more like what I was talking about:

This provides line to line and either line to rail protection.  These are bi-directional transzorbs, voltages should be in the VCC range (or slightly less).  The resistor, while normally a good idea, might not be the best way to go with a RS485 transceiver as would imbalance the signal a bit (depending on the terminator value).  The transzorbs are a nice touch, but I know many people do without them and their designs still work.  This is more of a protection against static discharge and gross mis-wiring.

Gene

Ok, I`m with you now. What about the optics VO0601T`s would you keep them?

Picking the right TVS Diode

Look at the Max Driver Output from here you need to pick a diode that works in the range

Figure 1: MAX3232 Driver Operation Voltage

MAX481CSA+-ND

Driver Input Voltage (DI).............................-0.5V to (VCC+ 0.5V)

Driver Output Voltage (A, B)...................................-8V to +12.5V

In this case I took the  SMBJ7.0CA-E3/52SMBJ7.0CA-E3/52 as it will cover both circuits(7 to 12VDC

T

Tom,

I would drop the optoisolators, as they are not really buying you much protection.  Typically, I would only use optoisolator to cross the boundary of a separate power supply (isolating the transceiver which is also on isolated power rails), or to protect against high voltage bleed through  to the microcontroller.

Gene

Tom.

This is a somewhat confusing specification.  The 'Absolute Maximum Ratings' section of a specification is more about the worst possible case or the point were you are most likely going to damage the device.  Typical specifications are usually a better guide as to what range yo want to protect.  If the device you are communicating with is also powered at 5V (not sure if you have stated what Vcc is), it is unlikely that you would see -8 to +12.5V swings.

Gene