Video Design Idea: Build your own laboratory precision voltage reference
Mark Thoren, mixed-signal application engineering manager with Linear Technology, demonstrates an amplifier-based circuit design for a relatively inexpensive precision voltage source.
By Staff -- EDN, May 24, 2007
| Bonnie Baker, senior applications engineer at Texas Instruments and regular EDN columnist, demonstrates a simple way to add DAC functionality to a microcontroller-based system using only an op amp and two passive components.; Bonnie Baker; DAC; EDN.com; analog design; op amp; video design idea; Mark Thoren, mixed-signal application engineering manager with Linear Technology, demonstrates an amplifier-based circuit design for a relatively inexpensive precision voltage source.; Mark Thoren; amplifier-based circuit design; linear technology; mixed-signal application; precision voltage source; Jim Williams, staff scientist with Linear Technology, explains why PC clocks are invariably wrong, and how engineers can surmount the extreme measurement challenge involved in solving the problem.; cell phones; computer clocks; jim williams; linear technology; nanoamps; quartz crystals; video design idea; EDN Tech Clips deliver technical depth and tutorial design information for engineers involved in analog circuit design, power management, embedded-system design, board-level design, signal integrity, and more.
http://www.edn.com/video/video.php/?bclid=1028763155&bctid=1026280202 |
Designers working on analog circuits in the lab often need accurate voltage sources to test the operation of op amps and other components. A lab power supply isn't a very good choice for such work because most such supplies aren't stable over temperature and time. Of course you could use a 7-decade Kelvin-Varley precision voltage reference although such instruments are relatively expensive and perhaps overkill for the task at hand. Or you can build your own relatively inexpensive 5-decade reference and power it with batteries to ensure that no line noise corrupts the precise output.
-
Check out these related EDN links:
Precision voltage reference delivers 80 mA (12/1/2006)
Anatomy of a precision-voltage reference (6/24/2004)
Voltage reference is software-programmable (8/18/2005)
Low-voltage reference uses the ΔVBE circuit (10/28/2004) -
Here is the Dekapot precision resistor used in this design.
-
For comments, questions, or suggestions relative to EDN Video Design Ideas, contact Maury Wright, 858-748-6785, mgwright@edn.com
-
Nice idea and well executed. I liked the video, this was really well done.
J. Mason - 2008-6-11 02:23:00 PST -
Hi Mark,
Great design idea , very useful. But I would like to have a BOM if it is available, specially for the KVD, my lab would be really a better place to work with that precision voltage reference !!
Thank you,
Marc,
Marc-Olivier Guerin - 2008-7-10 18:24:00 PDT -
Thanks for the comments. I never knew a glorified unity gain buffer could be so much fun!
On inspection, I see that the part numbers for the transistors are all messed up. The NPNs are 2N3904, and the PNPs are 2N3906.
Q7 is a 3904, too. Didn't feel like adding another part to the BOM.
I'm in the process of making green boards for this - sure enough, my box is always on someone else's bench!
There is also an error in the schematic. Q6 should be a PNP, with the 100k resistor tied to the emitter. This produces a 93 microamp (or so) reference current for Q1, which is then mirrored in Q2 and Q3.
Another improvement I'm making is to add gain to IC2A (Vout driving an input resistor to IC2A's inverting input, appropriate feedback resistor,and the non-inverting input connected as-is). The window set by R1 and R2 will then be widened appropriately. This will let the accuracy of the out-of-regulation detection circuit depend on the LT1881, not the LT1017.
Mark Thoren - 2007-5-7 16:28:00 PDT -
Impressive, useful design, well analysed - my pick for best idea so far. Schematic has a minor error (no Vref source shown), and the PN for Q7 and the Kelvin Varley Divider would be helpful. Is there a BOM available?
d blaine - 2007-8-6 11:36:00 PDT -
Nice job on both the project design! The EDN project video and related files are great.
What is the expected absolute accuracy and temperature coefficient of the project as presented?
What types of DACs (Linear Technology or other) might have suitable accuracy (100 ppm? 10 ppm?) to replace the potentiometer with a digital divider?
Joe Geller - 2007-24-5 15:35:00 PDT
