1. Home
  2. Electronics Design
  3. 4-20mA Precision Source Ideas
  4. Page 3

4-20mA Precision Source Ideas

Thanks for the reply! But I'm not sure I follow. R5 puts a load on the LM317 to keep it happy. The ADJ current gets sunk by U2 input. The idea wasn't that the LM317 would provide the output current, but it would act as a cheapo reference so that 0 volts would give a 4 mA output. LM317s in TO-92 packages are really inexpensive! Unfortunately it's moot, as the OP mentioned below having 4 mA at 0 volts input wasn't really his requirement.

This LM317 model seems to misbehave in the standard voltage regulator circuit too.

Reply to
Bitrex
Loading thread data ...

How about this one?

ftp://jjlarkin.lmi.net/22S220A_20.pdf

John

Reply to
John Larkin

In theory, and perhaps even in Spice, rail-to-rail OpAmps should work. However real world OpAmps don't go completely to the rails. So I'd expect at the very least some offset, maybe failure to queue up.

Let us know how it works out. ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |

      Remember: Once you go over the hill, you pick up speed
Reply to
Jim Thompson 

--
Bad:

       |
       |
-------o-------
       |
       |



Good:

      |
      | 
------o-o------
        |
        |
Reply to
John Fields

Ugly! We've never had a problem with connection dots falling off the page.

John

Reply to
John Larkin

Nonsense. It's perfectly readable as it is.

Makes a mess of any complicated schematic. Most people aren't blind.

Reply to
krw

relatively speaking!

Jamie

Reply to
Jamie

Fifty years ago, when people still used chain printers, this was a problem. He's likely not worked with anything newer since.

Reply to
krw

a

LM317 directs its several mA of quiescent current to the output, so your ~10k load is not enough for the poor thing to do anything else bar saturate?

As for the adj current, 22k there seems high, but I didn't look at a datasheet. As a rule with normal LM317 circuits I'll use a 120R output to adjust resistor to make sure the minimum load requirement is met.

240R for the baby one.

My comment on your circuit was from looking at the high value resistors around the '317 are not letting the thing work properly as you're not letting the quiescent current get out so the chip can operate properly.

Does either explanation make more sense now?

I'd expect it to if you have a habit of using such high value resistors around it, OTOH I've not used LM317 in a simulation yet. Maybe the model's off?

Grant.

Reply to
Grant

WTF? Where is it hiding? That's more like the cabinet end of measurement stuff I used to work on. 'cos it's got way too much for a two wire sensor or sender unit ;) Where does the coffee come out?

Grant.

Reply to
Grant

Me neither, but we had to for pcad schematics to meet their design rules, the angle the crossroads to make two dots. What a pain.

Reply to
Grant

MC4558 has +/- 22V max rails, it's a cheap dual. Only problem is datasheet says pinout compatible with LM358., but it does not have common mode to Gnd.

Grant.

Reply to
Grant

It has a bunch of modes.

formatting link

This board has 13 ARM processors, 12 on the isolated channels and one overall supervisor. It's used to simulate processes to big control systems, or to do control, or whatever.

John

Reply to
John Larkin

s:

tes:

o

with a

ms

ted

red

I

The inputs usually do go to the rail, the outputs never do. In John Devereux's application, you could pick a P-MOSFET with a high enough minimum gate-to-source voltage to let the output stick within its guaranteed range.

If he actually needed a 0mA output it might get tricky, but 4 to 20mA shouldn't pose a problem.

-- Bill Sloman, Nijmegen

Reply to
Bill Sloman 

--
That's not what it's for; it's to make sure that crossovers don't
become connections because of an inadvertent addition of a dot at the
crossover point.

In the newer system, where there are no dots and wires merely touch
other wires in order to indicate connection, it's essential to offset
connections in order to be able to distinguish a connection from a
crossover.
Reply to
John Fields

with a

I tested the LM317 model in a standard circuit with low value resistors and a 100 mA load against a table of output voltages. With 240 from OUT to ADJ, and 910 to ground, the '317 should give about 6 volts output. The model gives more like 6.3, which I think is far enough off the mark to indicate a problem. Also, the current out of the ADJ pin is 48 uA, which is only half the value specified on the datasheet. When using large value resistors the discrepancies become more pronounced.

I wouldn't usually use such high value resistors around the '317, but the idea in the circuit was to have close to 0 volts at the output of the op amp U2 (which is like an upscale LM324) when the input is at 0 volts - unless the load on the op amp is relatively large the output low voltage will rise. The resistors around the LM317 would then set things up so that at 0 volts input there would be 4 mA through resistor R2. Increasing the input voltage would then increase the current from the quiescent value.

I have some LM317s somewhere, I think I'll breadboard the circuit and try it out!

Reply to
Bitrex

It's not remotely new either. Other problem.. it's sure as heck not "precision" with +/-5000uV Vos at 25C and minimum gain only 50,000.

Possible contenders available in duals.. LT6011 (limited differential input voltage), OP2177... but they're more than 10x more expensive.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com
Reply to
Spehro Pefhany

Not new, not 'good', but higher voltage than the norm as only claim to fame ;)

What about TLE2022 dual, TLE2024 quad? +/- 40V max, 100uV offset? LT1113?

That's just selecting for high voltage, then precision, from Farnell's online catalog, for duals or quads. Seeing what's in stock.

Grant.

Reply to
Grant

People have been known to inadvertently swallow a fork.

Newer systems with no dots? That's crazy. I wouldn't buy an EDA systems that was so cheap it didn't come with lots of dots.

Did yours run out or something?

John

Reply to
John Larkin

A line crossing another is ALWAYS to be considered to NOT be a connection. No little loop over required, and one should NEVER expect to see a connection node "dot" at such an intersection/crossing.

That IS the rule. You do know what rules are, right, John? It isn't an electrical thing, it is a schematic representation/expectation/standard thing.

The proper way to join a node to a line schematically will always be a "T" or three way join. Never a 4 way intersection, and it is even questionable to bring in another line on an angle other than 90 degrees to an existing three way "T" node.

Reply to
Chieftain of the Carpet Crawle

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.