Dick Cappels' project pages http://www.projects.cappels.org
Return to HOME (more projects)
A Portable Precision Voltage Reference
A plenty accurate, nearly bullet-proof voltage reference

The voltage reference chip is a surface-mounted part under the board.
Most of the parts are to protect the voltage reference chip.


It has been said that a man with one watch knows what time it is, but a man with two watches is never sure.  The same can be said for a person who has more than one voltmeter.  In my situation, I have several voltmeters in each of two different locations a third of the way around the world apart. The best solution would be to send the meters to  NIST traceable calibration labs, but that's too expensive for me. Eying some of the semiconductor manufacturer's web sites, I saw that modern-day precision semiconductor references are accurate, stable, and inexpensive, so I bought some. Here is how I hooked them up.

I went looking for a reference chip between 1 and 2 volts, so I could check the calibration of my digital voltmeters on their 2 volt scale. The Maxim MAX6168Am a 1.8 volt, 5 parts per million per degree C reference was appealing, but I was not able to find one available i small quantities for delivery soon enough for my purpose. After shopping around, I settled on the National Semiconductor LM4140-1.0 precision voltage regulator, which I was able to buy from Digikey for $3.58 each.

The National part, specifically an LM4140CCD-1.0-ND is specified to have a nominal output voltage of 1.024 volts, within 0.1%  -that's accurate to within about 1 millivolt. The temperature dependent drift is better than 10 parts per million per degree C, and the aging drift is less than 60 parts per million for the first 1,000 hours. Since the accuracy is within one digit of my best meters, it is more than adequate from my anticipated purposes.

This basic circuit should be useful for many of the other series connected voltage reference chips out there, though it might be a good idea to change the 78L05 voltage regulator to one with an ouput voltage appropriate for the particular regulator, and to similarly change the output protection zener to one with an appropriate voltage fo the particular regulator used.

The Circuit

Most of the circuitry is there to protect the LM4140 from the user.

A keyed power input connector, series rectifier and a shunt rectifier, both 1N4007, prevent reverse voltage from being applied to the power input. A 27 volt metal oxide varistor clamps the voltage to the 78L05 that follow it, to less than 30 volts, thus protecting the 78L05. An input voltage of between +9.0 volts and about +24 Volts is needed to power the 78L05. The 78L05 in turn, provides 5.0 ± 0.25 volts for the LM4140 precision reference. This has two important benefits: The 78L05 keeps the input to the LM4140 from exceeding the maximum rated input voltage of 5.6 volts. It also keeps the input to the LM4140 constant, thereby removing up to 300 parts per million of output voltage per volt of input voltage of input voltage sensitivity. A 3.3 volt zener diode across the LM4140's 1.024 volt output limits the maximum voltage that can be applied to the LM4140's output by an accidental contact.

This circuit  would work well with other voltage reference chips as long as the input voltage for the reference chip is compatible with the output of the three terminal regulator and of course if the reference voltage was significantly lower than the output protection zener diode's zener voltage.

Varistors are uncommon parts in some labs. The varistor on the input can be omitted, or a zener diode of the appropriate rating could be substituted.

The LED and its associated 2.2 k resistor are only there to indicate that the circuit is on. It too can be eliminated without affecting the performance of the rest of the circuit.


The circuit was hand wired on a piece of phenolic board with one pad per hole. The parts were arranged so only one jumper, which was run across the component side of the board, was needed.

The most difficult part of the construction was the mounting of the small 8 lead surface-mount LM4140.  It turned out that I could place the package on the pre-etched pads on the phenolic board, but I had to split the pad that leads 6 and 7 attached to because they aren't to be connected together. Using a new X-Acto Knife blade and made two paralled cuts across the copper doughnut, and removed a thin slice of copper to form a small gap between the pads. To soldering chip to the pads, I put a tin coating of neutral PH flux on the copper pads, tinned the pads with solder, removed the excess solder with a flux soaked piece of wire braid, then carefully and tediously nudged the package into position then tacked one corner lead down. After several iterations of tacking, inspecting, moving, etc., I managed to get the chip tacked down in the right position, then went around and heated the rest of the pins, adding solder where it looked like it was needed. The solder blobs on some pads appeared when I connected wires to the pads.

The other components on the board were connected by bending their leads over to leads of the other parts they were to connect to, and soldered in place.

The hardest part was mounting the tiny
surface Mounting package. The pad that
pins 6 and 7 are soldered to was split
into two half doughnuts before mounting
of the part.

Once the circuit had been built up and tested, I added four metal standoffs to hold the board off the workbench, and then cut a small piece of phenolic board to shape and hot-glued it to the back of the circuit board to protect the LM4140 and the wiring from being accidentally physically damaged. When applying the hot melt glue, I was careful to run it around the perimeter of the back of the board, keeping it off of the LM1440 and component leads as much as possible. I went around twice, building up a rim to hold the phenolic board off of the circuit board. As the glue had hardened, I placed the small piece of phenolic board on the rim and added more hot melt glue to hold the board in place.


I measured the LM4104 on my two Fluke digital voltmeters, both about 10 years old and not calibrated since they left the factory. I was amazed to see that they both measured 1.024 volts. Another voltmeter, made by a manufacturer whose name one would recognize, measured the LM4140 as 1.031 volt. So, at this point, I would tend to believe the Flukes.

I have an older voltage reference, which was based on Design Idea article in EDN Magazine back in the 1970's -its an LM723 voltage regulator chip connected so that the internal pass transistor acts as a heater, and the internal current limit transistor serves as a temperature sensor, so the chip could provide a stable temperature for the on-chip zener reference voltage.  Back on March 1, 1979, I calibrated this against a differential voltmeter with calibration traceable to the NBS (Now NIST). I had set the output of this LM723 based reference to 1.00019 volts. I connected this LM723 based reference to a bench supply and watched it on one of my Fluke DVM's. After a minutes of warm-up time, the reading settled at 1.000 volts.

This result tends to reinforce the belief that the two Fluke voltmeters are right on, and the lesser brand is also pretty close. More significantly, based on my older LM723-based reference, the new LM4140 based reference is accurate to within a millivolt. My faith in National Semiconductor's specification for the LM4104's accuracy was also reinforced.

With them I can calibrate my test equipment and my projects, and more importantly, know more accurately what voltage values are in circuits I am testing. This is a bigger deal than it might first appear. Often, I'm left wondering what happened to the odd few tens of millivolts when my measurements don't add up. Now, I'll have a much better idea of how much of that error to assign to the meter. 

The precision reference can also be used with a low-offset opamp and a precision resistor to make a precision current source, which could come in handy. It could also be used to generate other reference voltages using a precision divider in a opamp's feedback look.

With two of these LM4140 references, I can have one at each of my two locations, and if need be, bring them  together occasinally for comparison. 

Example: Summary of  accuracy of my DVMs on the 2000 millivolts scales:
Meter Name
% Error
Fluke 1
1.024 V
Fluke 2
1.024 V
1.031 V
1.025 V
1.018 V

I think using the Fluke metres in Arizona, I don't have much to worry about, but
now I see there is a 1.2% spread between the two meters I use most in Thailand,
and I need to take this spread into account.

HOME (More Projects)
Contents ©2005 Richard Cappels All Rights Reserved. http://www.projects.cappels.org/

First posted in August, 2005, updated February 2012.

You can send  email to me at projects(at)cappels.org. Replace "(at)" with "@" before mailing.

Use of information presented on this page is for personal, nonprofit educational and noncommercial use only. This material (including object files) is copyrighted by Richard Cappels and may not be republished or used directly for commercial purposes. For commercial license, click here.
  Liability Disclaimer and intellectual property notice
(Summary: No warranties, use these pages at your own risk. You may use the information provided here for personal and educational purposes but you may not republish or use this information for any commercial purpose without explicit permission.) I neither express nor imply any warranty for the quality, fitness for any particular purpose or  user, or freedom from patents or other restrictions on the rights of use of any software, firmware, hardware, design, service,information, or advice provided, mentioned,or made reference to in these pages. By utilizing or relying on software, firmware, hardware, design, service,information, or advice provided, mentioned, or made reference to in these pages, the user takes responsibility to assume all risk and associated with said activity and hold Richard Cappels harmless in the event of any loss or expense associated with said activity. The contents of this web site, unless otherwise noted, is copyrighted by Richard Cappels. Use of information presented on this site for personal, nonprofit educational and noncommercial use is encouraged, but unless explicitly stated with respect to particular material, the material itself may not be republished or used directly for commercial purposes. For the purposes of this notice, copying binary data resulting from program files, including assembly source code and object (hex) files into semiconductor memories for personal, nonprofit educational or other noncommercial use is not considered republishing. Entities desiring to use any material published in this pages for commercial purposes should contact the respective copyright holder(s).