How Accurate is Your Test Equipment

How Accurate is Your Test Equipment
Recently we had the task of building a low-cost test fixture to allow our personnel to quickly check the accuracy of the many small digital meters we use in our company. Without incurring the significant cost to have everything individually calibrated, we decided that a simple portable reference would be a good tool to have so we could quickly check each device as needed and cabling connectivity. As a result, we then came up with a simple test fixture that would do the job adequately but would not break the bank. What we finally designed is something every home-brewer should seriously consider. Come devices can be connected just with the cables, you can find coaxial cable at The finished unit is shown in fig. 1 and the schematic in fig. 2. The heart of the test fixture is the Analog Devices AD587 Reference IC. This chip is available in several varieties. However, the one we chose is the AD587JNZ, which provided an output of 10 volts, is accurate to ±10mv, and comes in an easy-to-handle 8-pin DIP package. We also added a couple of other low-cost features as you soon will see. The AD587 takes an input voltage of 15 volts (or more) and produces an output of between 9.990 and 10.010 without calibration. An external pot (10K) can be connected as shown (in dotted lines) to allow the output to be trimmed exactly if a suitable reference for comparison is available. Temperature drift is on the order of 20 parts per million per degrees C, and stability of the output is rated at ±15 parts per million per 1000 hours. What all of this means is that with this chip you can produce a very accurate calibrator for the shack if you wish, and one that easily will provide the level of calibration required for the average home brewer.
test fixture
In our circuit we have added a diode and 15-volt regulator to protect the chip against reverse input polarity and to also allow an unregulated voltage source to be used (such as a surplus 18-volt wall wart). Keep in mind that even a couple of 9-volt transistor-radio batteries in series could also be used for the utmost in portability. You will also notice from the schematic that we have added a number of 0.1% resistors to configure a divider that produces 3.00 volts and 1.00 volt (like the common oscilloscope divider) for voltage range calibration. In addition, 1.001K, 10.01K, and 100.1K resistors (also 0.1%) are provided for ohmmeter calibration purposes. We have even added a 10.00-milliampere current output (derived from the 10.000-volt reference output) via a 1K 0.1% resistor to check the current range. All of this is mounted in a RadioShack box”along with binding posts from the same source. Wiring is not critical. However, all connections to precision components should be mechanically sound and properly soldered. Connections to the binding posts should be tight and for best results, soldered as well. You donwant a 1-ohm resistance in series with a 1000.1-ohm resistor due to a poor mechanical connection to a solder lug! If you want even better accuracy, use the 10K pot, but make sure you have something better than 10.000 volts to use as a reference for comparison. Also keep in mind that this test fixture can only be used (as is) to check DVMs with high-input impedances (usually 10 megohms). Using lower impedance meters will load the resistor divider portion of the circuit and give inaccurate results. The data sheet for the AD587 describes its operation in more detail and also covers the more accurate versions of the chip. The AD587JNZ is in stock at DigiKey Corp.( and costs less than $7.00, a real bargain considering the accuracy it provides. For a total cost of less than $25, this is a tool that every home brewer really should have. If there is enough interest we will be glad to investigate the possibility of having someone make a low-cost commercial version of this device available, so please let us know. If you are a experimenter,”then you should seriously consider actually building one.

Share This Post

Recent Articles

Leave a Reply

You must be Logged in to post comment.

Powered by