Multiplexer with 4-20ma inputs

So neither lead of a 4-to-20mA device are ground?

...Jim Thompson

To circuit ground. Typical supply 24V, compliance 10-20V+ typically.

Here's one way to do very high accuracy mux with a single precision sense resistor and cheap parts:

(to other channels) (+) 1 | | o | | | | | | 1N4148 | | | +--->|-------+ 1K | | ___ |/ | /SEL_1 o-|___|--| 2N | |> 4401 | | | Output | +--------o === | GND-A .-. | | | |R '-' | === GND-B

Some care as to layout is required as almost 1/3A might be flowing through GND-A with 16 channels, and R might be a couple hundred ohms.

Best regards, Spehro Pefhany

Yes, for "two-wire" transmitters. Some have an internal power supply and just look like an isolated current source, so you just ground the (-) terminal and connect the (+) to the resistor, and end up at about the same place.

Best regards, Spehro Pefhany

Not directly. The usual way is to convert it to 1-5 V through a 250 ohm resistor. That's it, assuming the negatives can all be tied together. If they can't, often the case in industrial installations, you need to follow it with a differential buffer. If you take that road add some off-set and gain to have it output 0-5 V.

- YD.

someone mentioned current steering, I pictured something like this:

you don't have to use discretes the transistors could be a ULN2003 the diodes a common-cathode 8-pack and address decoding done by an active-low address decoder or decade counter, but that's still 3 parts instead of one. you

That was me. You have pictured it correctly for load tied to ground. Just put some resistors in the transistor bases ;-)

...Jim Thompson

Hello Jim,

Or use FETs...

"Joerg" schreef in bericht news:zBwzg.142$ snipped-for-privacy@newssvr13.news.prodigy.com...

Or use ULN2003 etc, as Jasen suggested in the first place.

Although it won't affect things under normal conditions, the fact that some current can leak through from other unselected channels under a zero-mA fault condition on the selected channel might dissuade me from using a switch with such a high voltage drop at 20mA. A DMOS driver (eg. TI) might be better, and could guarantee