Voltage regulator and voltage reference

Thread Starter

andriasm

Joined Aug 2, 2022
25
What is the difference between voltage precision reference and a voltage regulator?

I need to supply the PMFc720P1 sensor. Its supply range is between 8 and 14 VDC.
Can I supply it with 12 VDC generate by a voltsge regulador or it must be by a voltage reference?
It is good to remember that this sensor is used in medical equipment.
 

Ian0

Joined Aug 7, 2020
9,667
There is some overlap between the two.
It depends on the amount of precision and the output current.
A brief look at the datasheet of the PMFc720P1 suggests that the output is not referenced to the supply, and it has its own built-in reference, so it can be supplied from any supply. At a guess, having a quiet supply would be more advantageous than having an accurate supply.
 

BobTPH

Joined Jun 5, 2013
8,808
A voltage reference is designed to provide a very accurate voltage at low current. It is not designed to power circuits, only to provide a signal.

A voltage regulator is designed to provide power at a less accurate votage.

As Ian said, there is some overlap. If your circuit needs very little power, it might be supplied by a reference. And if you need a reference signal that does not have to be extremely accurate, you can supply it from a regulator. This is often done using the same regulator that is used for power.
 

DickCappels

Joined Aug 21, 2008
10,153
A very important parameter for a voltage reference for some applications is its temperature stability, sometimes noise is an important consideration. Those are both important for voltage regulators, but those parts sold as voltage references have much tighter specifications.

With respect to your application, not knowing more than is in the datasheet (attached) I would go with a tight-tolerance precision voltage reference. I would not power one of those from a typical voltage regulator.
 

Attachments

crutschow

Joined Mar 14, 2008
34,281
I agree with Dick.
It is tested with a 10±0.01VDC supply, which implies a high sensitivity to supply voltage, so an accurate and stable voltage supply should be used, i.e. a voltage reference.
 

Ian0

Joined Aug 7, 2020
9,667
I agree with Dick.
It is tested with a 10±0.01VDC supply, which implies a high sensitivity to supply voltage, so an accurate and stable voltage supply should be used, i.e. a voltage reference.
Depends how you read that datasheet.
How can it have an input voltage of 10V±10mV but a supply voltage that can vary between 8V and 12V when it has only 3 terminals?
 

Thread Starter

andriasm

Joined Aug 2, 2022
25
Depends how you read that datasheet.
How can it have an input voltage of 10V±10mV but a supply voltage that can vary between 8V and 12V when it has only 3 terminals?
Yes. Thats a point to consider. I will make a test varying the supply between 8 and 14 VDC. Does the output voltage must be the same in this voltage variation, right? I will feedback you the results.
 

Ian0

Joined Aug 7, 2020
9,667
Yes. Thats a point to consider. I will make a test varying the supply between 8 and 14 VDC. Does the output voltage must be the same in this voltage variation, right? I will feedback you the results.
That’s one to ask the technical help line for ”clarification”.
 

BobTPH

Joined Jun 5, 2013
8,808
It is hard to imagine why it would need a very tightly regulated power supply to achieve an output accuracy of ±2%
 

WBahn

Joined Mar 31, 2012
29,976
(Forgot to hit Post yesterday, so this is mostly OBE.)

What are the current requirements for the sensor?

A reference is intended to be a reference -- it provides a precision voltage value that you compare other voltages to. A voltage regulator regulates a supply voltage that provides power to something.

In practice, the distinction can be gray as manufacturers make parts that can provide both services at the same time. But it's a pretty good bet that a voltage reference is probably going to do a better job acting as a precision reference when it isn't also providing a bunch of current.

In looking at the sensor data sheet, the voltage input is specified as being acceptable between 8 V and 14 V. The output mapping, however, is specifically spec'ed at 10 V at 25C. Details are not very clear, but it states in the verbiage (which is often written by marketing types) that the accuracy is better than 2% full scale and 3% over the entire temperature range. It is silent on the effect of supply voltage.

You can either try to get clarification from the manufacturer, or you may just have to take some of your own measurements. But, if I had to make a guess, I would say that it should be good over the entire input voltage range. The max signal out is 5 V, so there's quite a bit of overhead even at the minimum supply voltage of 8 V. I'm guessing there's an internal reference that's good enough to meet spec over the full supply range.
 

DickCappels

Joined Aug 21, 2008
10,153
The implication from the test condition in the datasheet is that to get the output to agree with that described, you will need a very good +10V reference.

Try something like a ST Microelectronics L78S05 voltage regulator and if you are happy with the results, go ahead and use it. If I were designing a product I would study the mass flow sensor on the workbench before designing it into anything, and then I would probably make a power source based on a precision voltage reference. The $3 (U.S.) savings can be used to make the downstream circuit adjustable to compensate for errors in the gain and offset.
 

Thread Starter

andriasm

Joined Aug 2, 2022
25
The implication from the test condition in the datasheet is that to get the output to agree with that described, you will need a very good +10V reference.

Try something like a ST Microelectronics L78S05 voltage regulator and if you are happy with the results, go ahead and use it. If I were designing a product I would study the mass flow sensor on the workbench before designing it into anything, and then I would probably make a power source based on a precision voltage reference. The $3 (U.S.) savings can be used to make the downstream circuit adjustable to compensate for errors in the gain and offset.
Dick, there is some example that you can give me to make a circuit adjustable to compensate for errors in the gain and offset?
 

DickCappels

Joined Aug 21, 2008
10,153
Attached is the excellent Designing Gain and Offset in Thirty Seconds from Texas Instruments.

Methods getting a single supply opamp to go to or below ground:
If you are trying to get your offset down to zero volts and you only have a positive power supply, you can use a rail-to-rail opamp to get very close to ground (0 volts) but to your 12 bit A-to-D it might not look like zero because rail-to-rail opamps aren't really rail-to-rail. You might need to place a resistor from the output of the opamp to ground to "help" the output get all the way to zero volts. Better would be to add a negative power supply with an ICL7660. Try a bare rail-to-rail opamp first if it doesn't go to zero you can use one of the techniques mentioned above to solve the problem.

A very low current negative current source made of R1 and Q1 can pull the output of some opamps below ground.
1660760302393.png
Alternative you can look up Avalanche Breakdown Photoemission and the Photoelectric Effect in Bipolar Transistor for a less popular negative power supply
 

Attachments

Last edited:
Top