Proper use of voltage references

Hello,

I'm working on a project for which I need to input an offset of 2.5V into an op-amp, before doing that I decided to (unsuccessfully) test the voltage reference (LM385Z-2.5) I'm planning to use. For that I made the following simple circuit:

I set Rs according to the following formulas (1) and (2) of this document:

Rs_min = (Vin_Max - Vout) / (I_Load_min + Iq_Max)
Rs_max = (Vin_Min - Vout) / (I_load_max + Iq_Min)

I took the following considerations:

Rload = 320 - 340 Ohms

Vin_Min = 4V
Vin_Max = 5V
I_load_min = 0.007352 A
I_load_max = 0.007812 A
Iq_Min = 0.00002 A
Iq_Max = 0.02 A

Resulting in Rs_min = 19.40 and Rs_max = 191.52.

According to the previously cited document, Rs must be between Rs_min and Rs_max, so I choose a value of 150 Ohms. But when measuring the voltage across Rload I didn't get the expected 2.5V, but 3.233V at Rload and 1.480V at Rs, as if everything was just a simple voltage divider.

I'm not sure how I ended up in this situation as according to the DS everything must be quite straight forward, but I guess there's something I'm missing and I hope you could help me out.

Best regards!

Why do you need RL ?

What's the volage without the 330 ohm resistor ?

Biasing mid-voltage (VCC/2) is more common with opamps and Arduino's A/D.
What is so special about your project that it needs a voltage reference.
Leo..

Your schematic looks fine. Are you sure you didn't wire it backwards? Or maybe it's just defective?

(They show the same schematic symbol as a Zener and a reversed Zener would give you a lower voltage, but I don't know what happens if that thing is reversed...)

3.233V at Rload and 1.480V at Rs

You mean across Rload and across Rs. I know what you mean, but "at" is not clear. You can say "at the junction of Rload, Rs and voltage reference" or "at the 5V terminal" and that would be clear (with ground the assumed reference). If you're not measuring relative to ground you should say "across", or specify the reference.

Voltage references are typically designed for micro to milliamp loads so you setup doesn’t pass the common sense “what am I trying to do” smell test.

I’d start over by eliminating the load resistor, set Rs at 10k and see what voltage output you get.

BTW, you have min and max I swapped.

Rs_min = (Vin_Max - Vout) / (I_Load_min + Iq_Max)
Rs_max = (Vin_Min - Vout) / (I_load_max + Iq_Min)

The above formulas are correct.

The values below appear to be swapped and/or incorrect. The LM385-2.5 data sheet does not give a min value, but the typical minimum current for proper operation Imin is given as 8 uA, max 20 uA. For loads on the order of several mA, you can ignore that.

Iq_Max = 0.00002 A
Iq_Min = 0.02 A

Assuming 5V in and 7.5 mA load current, Rs should be 330 Ohms. And guess what: you get a divide-by-2 voltage divider with your original circuit, and the voltage reference has nothing to do!

What is the actual circuit and the expected load current?

hammy:
Why do you need RL ?

I used it to calculate the current limits and to measure the expected 2.5 V

raschemmel:
What's the volage without the 330 ohm resistor ?

0mV across RS and 4.7V at the diode

Wawa:
Biasing mid-voltage (VCC/2) is more common with opamps and Arduino's A/D.
What is so special about your project that it needs a voltage reference.
Leo..

There's no particular reason. I just want to do it this way with learning purposes.

DVDdoug:
Your schematic looks fine. Are you sure you didn't wire it backwards? Or maybe it's just defective?

(They show the same schematic symbol as a Zener and a reversed Zener would give you a lower voltage, but I don't know what happens if that thing is reversed...)

I have checked (a lot of times jajaja) and it's not wired backwards

DVDdoug:
You mean across Rload and across Rs. I know what you mean, but "at" is not clear. You can say "at the junction of Rload, Rs and voltage reference" or "at the 5V terminal" and that would be clear (with ground the assumed reference). If you're not measuring relative to ground you should say "across", or specify the reference.

Thank you very much for the grammar lesson! You're right. I meant across Rs and at Rload.

WattsThat:
BTW, you have min and max I swapped.

jremington:
The values below appear to be swapped and/or incorrect.

You were right. I swapped the values in the post and are now fixed

jremington:
The LM385-2.5 data sheet does not give a min value, but the typical minimum current for proper operation Imin is given as 8 uA, max 20 uA. For loads on the order of several mA, you can ignore that.

According to the datasheet, the device has an operating Current of 20 μA to 20 mA. I took these values as Iq_Min and Iq_Max, respectively.

jremington:
What is the actual circuit and the expected load current?

So far it's just that. A voltage reference connected to a 330 Ohms resistor with an expected load current of 7.5mA... I have no final circuit, one idea is to use it to add an offset to an opamp but I could also use it for the ADC...

BRI0:
...but I could also use it for the ADC...

If you need an A/D with reference, then buy an A/D with build-in reference, like the ADS1115.
Or change the default Aref of the Arduino to one of the internal ones.
Most basic Arduinos have a buildin reference of 1.1volt, some have more.
Note that a fixed reference is not always what you want, it depends on the sensor used.
Leo..

Wawa:
If you need an A/D with reference, then buy an A/D with build-in reference, like the ADS1115.
Or change the default Aref of the Arduino to one of the internal ones.
Most basic Arduinos have a buildin reference of 1.1volt, some have more.
Note that a fixed reference is not always what you want, it depends on the sensor used.
Leo..

Thank you for the recommendation. But as I said, I don't really have a final circuit and the idea for the ADC was just an example.

So far I just want is to see an output of 2.5V using the voltage reference :slight_smile:

Dear all.

Thank you very much for your support and advices. I just wanted to tell you that apparently the problem was related to the protoboard. I used another one with the very same components and got the expected voltage!

Did you forget the power rail jumpers ?

raschemmel:
Did you forget the power rail jumpers ?

In fact, the problem was that the contacts of the breadboard weren't doing a proper contact with the voltage reference, so it was kind of floating... so in the end I had just a voltage divider.

Is it an old breadboard that might have had dust
on it or is it possible you didn't push the pins all the way in ?