-2
\$\begingroup\$

I'm very new to the topic but I read that some reference pins of some ICs should not be fed by voltage dividers because of impedance issues.

I need an adjustable +-4V output from an opamp or any suitable IC. The +-4V range adjustment should be done by a single pot. This adjustable voltage will be a reference voltage for the AD8226.

I have +-5V dual power supply to power all the circuit.

Basically I don't want to use heat-sink. I don't know if a voltage reference, voltage regulator or a opamp as a buffer is suitable for such circuit which will not drive much current(such as for a IC ref pin).

The thing is I need to be able to adjust between +4V to -4V just by using a single potentiometer.

Edit:

To follow one of the suggestion safely I need an IC which can output both +5V and -5V as below(what is the generic name of such ICs if exists?):

I found this one: http://www.linear.com/product/LT3032 Would it be precise enough for this case? (Oh noo this requires an already dual supply:(()

enter image description here

\$\endgroup\$
  • \$\begingroup\$ 1 - What are your "raw" input voltages? Do you have +/- 3 volts in, and want each output to vary between 3 (or -3) and zero? 2 - How much current do you need to supply? \$\endgroup\$ – WhatRoughBeast Apr 4 '17 at 14:05
  • \$\begingroup\$ Raw dual supply voltages are +/-5V. I actually need an adjustable output from min -4V to max +4V by using a single IC and a single pot. The output should not exceed 4V or -4V when played with the pot. If not possible I can live with +-3V as well. \$\endgroup\$ – atmnt Apr 4 '17 at 14:11
  • 1
    \$\begingroup\$ "I want to use it for a reference pin of this IC." OK, I'll bite. Which IC? Come on, stop dribbling information. Describe your system in detail and make us stop guessing about what, exactly, you need. Details matter. \$\endgroup\$ – WhatRoughBeast Apr 4 '17 at 14:21
  • 1
    \$\begingroup\$ Will you please answer the questions I ask? Which IC? A link to the data sheet would be even better - that way we don't have to do your work for you. \$\endgroup\$ – WhatRoughBeast Apr 4 '17 at 14:32
  • 1
    \$\begingroup\$ In load cells we need to set/zero the offset after we load the load cell. So adjustable offset would be great for that. But for strain gages Im not sure if I need a +- adjustable offset. Im not sure if a strain gage when installed have zero strain on it. I was thinking to make an amplifier which would work for any bridge sensor type application \$\endgroup\$ – atmnt Apr 4 '17 at 19:29
2
\$\begingroup\$

If i understand correctly, you just want something that can produce +-4 V, but it cannot be a voltage divider.

How about using an opamp as voltage follwer, and just use an voltage divider as input? The IC you are using does not seem to require much current.

schematic

simulate this circuit – Schematic created using CircuitLab

This voltage divider will do the voltage you need.

\$\endgroup\$
  • 1
    \$\begingroup\$ If you check the link, this will not do, and is specifically cautioned against in the data sheet. \$\endgroup\$ – WhatRoughBeast Apr 4 '17 at 15:08
  • \$\begingroup\$ You might be right, but i does not see how this differs from the first solution you made. (I understand that it might be a bad idea if got a bad power supply) \$\endgroup\$ – keffe Apr 4 '17 at 15:14
  • 1
    \$\begingroup\$ Please read the data sheet. It discusses this. \$\endgroup\$ – WhatRoughBeast Apr 4 '17 at 15:16
  • \$\begingroup\$ No, as i see it, it says that this method is correct. I am not trying to be an ass or anything, i just literally does not understand what is incorrent. My text suggests using a voltage follower after this voltage divider, but is not seen on the schematic (Which might be bad). \$\endgroup\$ – keffe Apr 4 '17 at 15:23
  • \$\begingroup\$ isnt there a single chip precision adjustable dual voltage references in the world? \$\endgroup\$ – atmnt Apr 4 '17 at 16:35
2
\$\begingroup\$

You can certainly do what you want, but it can be simple or complicated depending on your system. The big question is, how accurate and stable are your +/- 5 volt supplies compared to your variable voltage requirements? If the the answer is that you trust your power supplies, then you can simply do

schematic

simulate this circuit – Schematic created using CircuitLab

However, as I say you have to really trust your power supplies. Any noise or drift on these lines will be coupled to the Vref of your AD8226, and this is probably a Bad Thing. Assuming you want bullet-proof, you need something a bit more complicated, such as

schematic

simulate this circuit

Note that R1 and D1 are not specified. They can be a resistor/zener combination or a voltage reference IC. At any rate, the output is inverted by OA1 and the gain adjusted by choosing R3 to give -4 volts. This -4 is applied to one side of the pot. The -4 is then inverted by OA2 to give +4, and this is applied to the other side of the pot. The pot output is buffered as in the simple version.

\$\endgroup\$
  • \$\begingroup\$ I was planning to use a 5V single supply to make a dual power supply by using this converter: cds.linear.com/docs/en/datasheet/1046fb.pdf So I would have a dual 5V supply. But now you say it matters a lot in this case how good the rail voltages are. Your second circuit suggests 3 opamps in a row. Two questions: 1-) Where is the "single pot" which will make +-4V output in the second circuit, is R3 that pot? 2-) How about using OP4177 in my case which has four OP1177 in one package?:analog.com/media/en/technical-documentation/data-sheets/… \$\endgroup\$ – atmnt Apr 4 '17 at 16:11
  • \$\begingroup\$ oh sorry I just saw the pot R6. But I dint get what R3 is about \$\endgroup\$ – atmnt Apr 4 '17 at 16:23
  • \$\begingroup\$ Note that R1 and D1 are not specified. What zener value it should be? \$\endgroup\$ – atmnt Apr 4 '17 at 16:34
  • \$\begingroup\$ That is up to you. Obviously the zener (if you use one) must be less than 5 volts. Choose one, and determine the resistance based on the specified test current of the zener. But there are also lots of voltage reference ICs which will take 5 volts and give out a very precise and stable voltage such as 1.25 or 2.5 volts. \$\endgroup\$ – WhatRoughBeast Apr 4 '17 at 17:50
  • \$\begingroup\$ The gain of OA1 is -R3/R2. So you need to pick R3 to give a proper -4 volts from whatever voltage reference you pick. \$\endgroup\$ – WhatRoughBeast Apr 4 '17 at 17:51
0
\$\begingroup\$

You don't specify if you have both positive and negative voltages available and if the rails will be loaded or not. I assume you have load, otherwise see WhatRoughBeast solution.

One can dream up many possible solutions. Here is one over-simplified concept:

schematic

simulate this circuit – Schematic created using CircuitLab

OA2 needs negative supply too.

Other possible solutions include a DC/DC converter, possibly flyback, with two identical secondary windings, one positive and one negative. Regulate on the positive as per usual to your desired pot voltage and make sure the load on both rails are near identical and leakage inductance low and you will have an adjustable +-4 V supply set from a single pot.

schematic

simulate this circuit

\$\endgroup\$
  • \$\begingroup\$ TL 08x op amps will not give 4 volts from +/- 5 volt supply. \$\endgroup\$ – WhatRoughBeast Apr 4 '17 at 15:07
  • \$\begingroup\$ @WhatRoughBeast No, they are generic. I'll edit it. \$\endgroup\$ – winny Apr 4 '17 at 15:13
  • \$\begingroup\$ Yeah, but the point is that "generic" op amps are not rail-to-rail, and you need to specify that. It is unlikely that the poster is aware of the difference. \$\endgroup\$ – WhatRoughBeast Apr 4 '17 at 15:15
  • \$\begingroup\$ @WhatRoughBeast Well, I'm not feeding him/her a finalized ready for production solution. He/she needs to work out some details too. \$\endgroup\$ – winny Apr 4 '17 at 15:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.