DC Current measuring

I thought that I had come across a trick for measuring the DC current, a couple of years ago, in a circuit without breaking the circuit or using a DC clamp meter.

My failing memory tells me that it was using either diodes or resistors in parallel with a component in the circuit when wiring the circuit up, but I just cannot seem to find any reference to this.

I've run a search on PLCs.net, Patchn and general Google web without luck.

Anybody have an idea what I'm referring to?

... rab

If you are talking about a 4 to 20 ma loop, then placing a diode in the loop, and turn your multi meter to ma, then measure the current by placing your leads on either side of the diode, allows the current to pass thru the meter when you connect up.

Gil47 said:

If you are talking about a 4 to 20 ma loop, then placing a diode in the loop, and turn your multi meter to ma, then measure the current by placing your leads on either side of the diode, allows the current to pass thru the meter when you connect up.

Click to expand...

This is exactly what I am looking for. The diode would be in series, correct? [-]-----|>|-----[+] or [-]-----|<|-----[+] (can never remember the orientation, usually try the circuit and when it works, I'm happy).

Again, if you are refereing to a 4-20ma (or 0-20ma) is this:

Actaully measure the volt drop acrost one of the devices, not the the transmitter, but one of the other devices like a panel meter or analog input. Most loop devices have resistence of 250 ohm, so 4-20ma would produce 1-5vdc. Can verify that by either actaully breaking in to the circuit and measuring DC amps once, or by having some idea of the process variable and the calibration of the xmitter. Anyway, once you know the resitance of the device then can measure the volt drop and infer what the DC milliamps.

RBergmann said:

This is exactly what I am looking for. The diode would be in series, correct? [-]-----|>|-----[+] or [-]-----|<|-----[+] (can never remember the orientation, usually try the circuit and when it works, I'm happy).

Click to expand...

The diode would be in series.
For forward bias I would use choice 2,
also use Google images diodes

After the long wait, its now available!

http://us.fluke.com/usen/products/Fluke+771.htm

Below will be a description/definition of a diode but TO REMEMBER WHAT END IS WHAT, THE BAND IS THE NEGATIVE END. Look at the picture below, the band looks like the bar at the tip of the arrow; so the band end (cathode) connects to negative when you want to forward bias.

In this discussion the term diode and rectifier will be used interchangeably; however, the term diode usually implies a small signal device with current typically in the milliamp range; and a rectifier, a power device, conducting from1 to 1000 amps or even higher. Many diodes or rectifiers are identified as 1NXXXX . A semiconductor diode consists of a PN junction and has two(2) terminals, an anode(+) and a cathode(-). A diode and schematic representation are shown below.

​

rsdoran said:

Below will be a description/definition of a diode but TO REMEMBER WHAT END IS WHAT, THE BAND IS THE NEGATIVE END. Look at the picture below, the band looks like the bar at the tip of the arrow; so the band end (cathode) connects to negative when you want to forward bias.

Click to expand...

Ahh, so now I remember, the diode symbol is in terms of "conventional" current flow (arrow orientation showing [+] to [-] current travel) rather than electron flow ([-] to [+] current travel). So, if I understand correctly, the diode is placed in the current loop circuit to pass current (makes sense or else the circuit wouldn't work). But how does this permit the meter to read DC current as though the circuit was broken to permit the meter to be placed in series?

To explain why I asked this in the first place, we build industrial control panels. On the occassion when an analog device is installed in the system it often has to be validated that a current loop is reading (I know, this is stating the obvious). I can have my panel builders install, as a standard, a circuit that allows easy validation of the signal from say a PLC to the field device. I thought that it might be easy to just solder a diode into the circuit at a know location. Right now, I'm considering the installation of a small terminal block style "switch" that I can use to plug my multi-meter into either side of the switch, then open it, to test the circuit.

The new Fluke 771 meter runs just under $400 from my distibutor (I received notification of this product almost simultaneous with Alaric's post), and for the number of times it is necessary to do this, the cost is not justified. In additon, I may not be the one testing the circuit, it may be an instruction over the phone with a field tech. However, it is justified to install a standard circuit for validation rather than have wires removed and reinstalled when chasing down a problem signal.

... rab

But how does this permit the meter to read DC current as though the circuit was broken to permit the meter to be placed in series?

Click to expand...

The multi-meter in current mode is effectively a short circuit compared with the 0.7V or so voltage drop across the diode. THe voltage drop across the meter will be in the mV region, so the current takes the path of least resistance and flows entirely through the meter.

If you're doing this in a panel, it would be easy to place the diode in circuit and then bring a pair of leads to a couple of bannana plugs on the panel where you could just plug the meter leads in as required.

You can buy terminal blocks with diodes in them:

http://www.pacificparts.com/vends/wago/wago.elecbloks.htm

http://www.google.com/search?q=+diode+terminal+blocks&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-USfficial&client=firefox-a

Here I go assuming....

I am assuming a PLC is involved, if so when the incoming analog values are out of range an alarm light could be turned on. If an HMI is involved the value could be displayed.

The simple solution may be to add an analog or digital panel meter but may cost $50 or more per panel, if multiple analog signals are involved this could get expensive.

Since this is for testing only I might just run the signal wires to a terminal block with the diode connecting them, this would allow you to use the terminals as a testing point and appropriately label it.

RMA said:

The multi-meter in current mode is effectively a short circuit compared with the 0.7V or so voltage drop across the diode. THe voltage drop across the meter will be in the mV region, so the current takes the path of least resistance and flows entirely through the meter.

Click to expand...

Of course! Thanks for reminding me of what I should have learned (hah!) 35+ years ago in Electronics 101 when studying these devices.

rsdoran said:

I am assuming a PLC is involved, if so when the incoming analog values are out of range an alarm light could be turned on. If an HMI is involved the value could be displayed.

The simple solution may be to add an analog or digital panel meter but may cost $50 or more per panel, if multiple analog signals are involved this could get expensive.

Click to expand...

You have assumed correctly about the PLC in the most recent application. This project that what brought the question to mind was an unusual one for us that had (18) analog inputs and (2) analog outputs. there were (7) different types of devices, some self powered and some where the loop power had to be provided.

I really don't like the idea of disconnecting and reconnecting wires to check the loop current, so I thought that maybe we should just build in testing points as a standard when we have these kinds of applications.

Just as a point of infomation the circuit that caused us the most problems was a loop between an analog level transmitter, about 100' from the plant, and a analog input module. We knew we had a full tank and that we had to get some type of reading from the transmitter, but weren't. Used my cheapy signal generator that I bought from Grainger (9V w/4-12-24mA output selections) to validate that we (that is the contractor) had wired everything correctly. Utimately the problem was due to lack of a correctly "wetted" signal, or should I say a mis-wiring of the voltage source.

This brings up another question that I have resolved for myself, but I am always interested in the opinions offered here:

Scaling this many signals from Data Units to Engineering Units. Depending on the PLC being used, I may have diffent options for scaling. Some have scaling parameters (like SCP in RSLogix 500, or XYFS in Idec), many don't (RSLogix 5000, Misubishi GX Developer) so I have to write a scaling routine based on information that I obtained on this site. Usually I have an operator interface available. I currently have adopted the setting of analog conversions through the interface (using data registers), not through an on-board scaling function like Mitsubishi's Beijer units have, or scaling that can be performed when progamming the analog module) rather than hard coding the conversion values into the program. Regardless of the number of analog I/O signals. Any thoughts?

... rab