Go Down

Topic: problem with a 4-20ma pressure sensor  (Read 3186 times) previous topic - next topic

mikedb

I don't want to be rude , but I worked with probes for the last 15 years and I did it many times .

If I get 16mA direct from the probe I know the tank is 75% full then I connect the probe to the telemetry unit

and drive to the output station if the output is also 16mA I know the system is working.

The 250 ohm is fitted inside the telemetry unit to provide a Voltage for the analog converter.

The probe is a current generator it will generate 4mA always(atmospheric pressure) ,provided the load

resistor is not to big.

mikedb

From the datasheet supplied there is a internal current limiter resistor in series with the output transistor

inside the probe.




MarkT

Bad idea.
The resistance for the mA meter will be much lower than the minimum 250Ω required by the probe, effectively creating a short. That's a terrible idea, and a way to destroy the sensor there and then. You must have a resistor in series.
Wrong, its a constant current source output, so short-circuiting it is absolutely fine.  (Just like
open-circuiting a voltage source is fine).

The short-circuit current is given as 18mA in the specifications even, so its clearly not going to damage
itself or an ammeter.
[ I DO NOT respond to personal messages, I WILL delete them unread, use the forum please ]

mikedb

Thanks MarkT for confirming.

Internal resistance of my multi meter on 20mA scale is 13.6 ohm

and

4.3 ohm on 200mA scale.


wvmarle

Wrong, its a constant current source output, so short-circuiting it is absolutely fine.  (Just like
open-circuiting a voltage source is fine).
That appears to be at odds with #6, where I was told that 47Ω is too low a resistance for the thing to work.
Quality of answers is related to the quality of questions. Good questions will get good answers. Useless answers are a sign of a poor question.

Grumpy_Mike

#20
Apr 09, 2019, 09:54 pm Last Edit: Apr 09, 2019, 09:54 pm by Grumpy_Mike
Quote
Wrong, its a constant current source output, so short-circuiting it is absolutely fine.  (Just like
open-circuiting a voltage source is fine).
So please explain what the load line in the data sheet is telling you and what information it provides.

MarkT

Ah, that's true, there's a thermal limit - use 12.5V supply for extended short circuit operation.

It'll be fine for short tests with a multimeter.
[ I DO NOT respond to personal messages, I WILL delete them unread, use the forum please ]

PNV61

I checked back in to see if the problem had been resolved. There were no further posts from WarLion, but many other interesting responses. I appreciate that everyone is trying to help as best they can, and my intention is not to step on anyone's toes. However, some of the information that has been posted is not correct and for those that are interested in using these devices, I am hoping to steer that information back in the right direction. I apologize, in advance, for the length of this post.

WarLion's problem is a simple one, my logic goes like this:
The circuit consists of a power supply connected to a pressure sensor and a series resistor. In a series circuit the sum of the voltages across the series components must be equal to the power supply voltage.
So, if you are powering the circuit with 12V, and you measure 2.25V across your resistor, that leaves (12V - 2.25V) = 9.75 Volts across the pressure sensor. Since the manufacturer requires at least 12.5 volts across the pressure sensor to operate properly, you cannot expect it to work at 9.75V. This does not exclude that it may still be faulty, but there is no way to verify that if you don't operate it within the manufacturers specified limits (between 12.5V and 24V).
My rule of thumb when testing for a faulty part is to NOT test it near its limits. So, even if testing this sensor with an ammeter, I would use at least 13 or 14 volts (not 12.5V, and certainly not 12V or less).

The sensor is more than just a current source:
The pressure sensor can be thought of, and does operate like, a current source. But there is more inside than just a current source; there is also a strain gauge bridge, a constant current supply for that bridge, a strain gauge amplifier, and an output current driver. All of these internal circuits require some voltage to operate. That voltage must be between 12.5V and 24V. In that range, the device is electrically equivalent to a current source (sourcing between 4mA and 20mA).

Using the manufacturers graph:
The graph provided by the manufacturer, on page 4, is used for determining the appropriate power supply voltage for different load resistors. For example, if you choose a 250 ohm resistor: find 250 ohms on the left side, move across to the right until you intersect the line, and drop down to see that your power supply needs to be at least 17.5 volts. This is the minimum voltage required for the pressure sensor to work properly over its full 4mA-20ma range when using a 250 ohm resistor. Other resistor values mentioned (some discounted) in these posts, such as 47 ohms, 220 ohms, 267 ohms, 13.6 ohms, 4.3 ohms, and 0 ohms are also all perfectly valid choices.

For those interested, the graph was constructed to ensure that there is at least 12.5V across the sensor as it varies its current between 4ma and 20 mA. It is based on the following simple formula:   Vsupply = (12.5V) + (20mA) x (Rload)

Using a 0 ohm Load:
Of special interest to this discussion is that 0 ohms is a perfectly acceptable load. Using the graph again: find 0 ohms on the left side (it's at the very bottom) and follow it across to find that it intersects at 12.5 volts. So, 12.5V is the minimum voltage required to operate with a 0 ohm load resistor. The pressure sensor can operate indefinitely with a 0 ohm load, and there is absolutely no danger of damaging it, even if you decide to operate it at 24V instead. 0 ohms, in fact, is the ideal load for any current source since the voltage across a 0 ohm load does not change when the current changes.
For those wondering how you can possibly measure the current if your load resistor is 0 ohms; you can terminate the pressure sensor into a virtual ground node, a current mirror, or a Norton amplifier, which are all electrically equivalent to a 0 ohm load. There is, however, a 0.6 Volt offset for both the current mirror and the Norton amplifier, so the minimum power supply voltage would need to be 0.6V higher, or 13.1V in those cases.

The Short circuit current specification of 18mA:
The specification sheet covers 3 versions of pressure sensors;

1. SPT mA version - This version contains a 4-20mA current transducer, and is the subject of this discussion. It is a 2-terminal device.

2. SPT 4V version - This version outputs a voltage between 1V and 5 Volts rather than a current. It is a 3-terminal device.

3. SPT mV version - This version contains only the pressure membrane, and the strain gauge bridge (no active circuits). It is a 4-terminal device.

The short-circuit current specification of 18 mA, given in the spec sheet (page 2), is for the SPT 4V version, not the SPT mA version that we are discussing. If you think about it, the 4-20mA version already limits its current to between 4 mA and 20 mA, depending on the pressure. It also makes no sense to limit the current to 18 mA when the upper maximum is actually 20 mA.

Equivalent circuit showing a "current limiter resistor":
The "Equivalent Basic Circuit" from the spec sheet (top of page 4) shows a "square device" in series with the output transistor. That is NOT a "current limiter resistor". The "square device" simply indicates that additional circuits exist between the output transistor and the +IN terminal. Those additional circuits may be a single resistor, but there is likely more than that, but we don't know. It is drawn that way as a quick way of indicating that the output transistor is not the only device that limits the minimum voltage between -IN and +IN. I can elaborate on this further if anyone is interested, but it is a common practice when drawing block diagrams.

"thermal limit" and "extended short circuit operation":
These terms were used in one of the posts, but there is no mention in the spec sheet of either of them. Their use suggests that there is a temperature or power limitation as well as a time limitation when operating with a 0 ohm load resistor. As discussed earlier, you can use a 0 ohm load at 12.5V or 24V  (or anywhere in between) indefinitely. I realize that the earlier discussion did not include anything about the thermal properties of the device. The fact that it is not mentioned in the spec sheet should be a strong indication that it is not important. But, in all fairness, it is possible to provide a convincing case for why it is not important, but I will simply end it here for now. I can provide more on this if there is interest.


I hope this brings some clarity to the discussion and I am interested in hearing from WarLion, did you get your pressure sensor working?

neiklot

I hope this brings some clarity to the discussion
It was never unclear; I think what you brought is unnecessary detail (how ever correct it may be).


PNV61

neiklot,
you make a very good point.

WarLion,
best of luck on your project.

mikedb

PNV61:

Did you also notice that:

1.When you post a little info ,someone wants more info.

2.When you post your 1500 line sketch ,someone tells you to make it shorter ,more specific to the problem.

3.When you post everything with pic's ,someone complains about not ''auto formatting'' before posting.

4.When the problem is resolved ,everybody already know what the problem was right from the beginning.

5.When you post a long detailed explanation ,someone complains about to much info.

Go Up