Precision Voltage and Current Reference

A little something to check your digital multimeter accuracy occasionally.

Public Chat
Similar projects worth following
A small box that outputs 10.000V, 5.000V, 2.5000V, 1.0000V, 0.1000V, and 10.000mA, 1.000mA, and 100.0µA. This should be handy to check my Keysight 34461A 6.5-digit DMM from time to time.

Why am I doing this?:

I'm waiting for delivery of my Keysight 34461A 6.5-digit digital multimeter (DMM). I got to thinking that this would be the perfect time to build a precision voltage reference to check the DMM for parameter drift. This is a bit daunting. The voltage accuracy of the meter is specified as:

After 1 year the meter should measure a 100mV reference with an error of less than ±8.5µV, 1V±47µV and 10V±1mV. The real question is which is correct -- the meter or the reference?

Wandering through the reference wilderness:

There is a lot of passion about this subject. The absolute best ever reference is a Josephson Voltage Standard, but expect to pay ~$300k for one.

The next best thing is a heated buried zener diode, ala LTZ1000. This voltage output of the diode can hardly be called a reference since it is usually somewhere around 7.1 V. You don’t build one of these for accuracy. Its claim to fame is low temperature drift and incredible stability (i.e. it doesn’t change much with time). But the LTZ1000 is a bit pricey as well — around $60 just for the hermetically sealed metal can that encloses it. And the surrounding circuitry is a nightmare of concern. Not for me with just my lowly 6.5-digit DMM.

After you discard the top two, you’re left with a bazillion choices in a more reasonable price range: $1-$12. The problem is to choose wisely. In order to do that I had to decide what is important for my purposes. Here’s a list of primary specifications:

  • Temperature drift. How much is too much? The other factor is how does the drift change when the reference voltage is trimmed, which I intend to to. Really good drift is 1ppm/C, but you pay dearly for it. <10ppm/C is pretty common and cheaper. I figure that the temperature in my shop varies from 50°F (10°C) in winter to not more than 80°F (27°C) in summer, for a total range of 17°C. That’s not a lot, but with 10ppm/C drift the voltages vary 0.017%, or 1.7mV @10V. That’s a bit more than desired.
  • Stability (or Aging). This is the biggy. The lower the better. Specifications range from 10ppm/1000hrs to 100ppm/1000hrs. No design tricks to fix this.
  • Noise. This is a big deal if you are designing a reference for a data converter (ADC or DAC), but I think this doesn’t matter in this application because the DMM filters the input and averages many readings to obtain a good result. Any noise above 1 Hz should disappear.
  • Line regulation. All of the specs are pretty good: usually < 5ppm/V. This translates to an output variation of 5µV of change at the output for every 1V change on the supply.
  • Load regulation. Usually in the range of 10-30 ppm/mA. Doesn’t matter. The 34461A has a highZ input that puts a negligible load on the reference output. Even the 10MegΩ input resistance of a lower grade DMM won’t be of any significance.

Other Stuff to Worry About:

  • Temperature effects on the other components can destroy the integrity of the reference. Temperature drift of discrete resistors is not guaranteed to match. Discrete resistor tolerances below 0.1% get very expensive, along with tempcos less than 10ppm/C.
  • Using an amplifier to generate a separate reference, either higher or lower in value, incurs offset errors and offset drift, resistor matching tolerances and drift. Also possible loading effects on the reference source.
  • Thermoelectric noise in electrical connections of dissimilar metals. This is caused by temperature differentials between circuit connections that generate small voltages. The trick is to use metals (like gold)...
Read more »

Gerber files of the Heater PCB. No need for a BOM for this board; there's only two components -- 2512 resistors, and the SOT23 AO3400 FET.

Zip Archive - 34.34 kB - 10/09/2019 at 18:00



Bill of Materials for the voltage reference.

ms-excel - 9.50 kB - 09/23/2019 at 21:43


Gerber Files of the Voltage Reference PCB.

Zip Archive - 57.17 kB - 09/23/2019 at 21:43



Bill of Materials for the Current Reference.

ms-excel - 8.50 kB - 05/16/2019 at 04:30


Gerber files of the Current Reference PCB.

Zip Archive - 55.85 kB - 05/16/2019 at 04:19


View all 6 files

  • 1+ Year

    Bud Bennett07/23/2020 at 23:13 2 comments

    It's been more than 1 year since I purchased the Keysight 34461A DMM. It is officially out of calibration now. Time to do a quick measurement. I powered up the voltage calibration box (the Digikey box) and the DMM for six hours prior to taking a measurement. Here's the results:

    DIgikey Box:

    Ambient Temp: 68°F(20C)

    DMM settings: 100PLC, Auto Input Z (High Z)

    Reading (not averaged)    Scale

    99.9994mV (nulled)          100mV

    1.000011V                          1V

    2.49976V                          10V

    4.99940V                          10V

    10.00003V                         10V

    These are not very far off from the six month readings.

    AliExpress Box:

    (same settings and ambient temp as above, soak time 4 hours)

    Reading                                            Scale

    100.00181mV (avg/20 samples)        100mV Nulled

    1.000035V (avg/20 samples)           1V

    2.50014V                                           10V

    10.000114V (avg/20 samples)              10V

    Current Reference(all readings 100plc, averaged with 20 samples):

    Reading             Scale

    10.001474mA     10mA

    0.999977mA     1mA

    99.99313µA       100µA

    General Conclusions:

    All of the voltage references seem to be very stable now. No significant difference between the USA components and the Chinese components. The current readings are drifting a bit, but the specs of the 34461A are quite a bit looser for current than voltage. The only way to tell for sure is to compare readings to a calibrated DMM with >6.5digit capability. That's not happening anytime soon.

  • 6+ Months

    Bud Bennett02/20/2020 at 20:51 0 comments

    All of the references have been sitting in a drawer, collecting dust, until today. I has been a few months since the 1000 hour tests were completed. I let the references and the DVM warm up for at least 2 hours before I took any measurements. At this point in time the DVM is expected to be within 155ppm @ 100mV, and within 80-90ppm for the 1V and 10V ranges.

    The Voltage Reference using Digikey parts:

    2020-02-20 2hrs warmup:

    Set Point: = 0.6278V

    Value          Range     Change
    100.0166    100mV    (+101ppm)
    1.000036     1V         (+13ppm)
    2.50013      10V        (+80ppm)
    4.99932     10V        (+20ppm)
    10.00005   10V        (-2ppm)

    I also calculated the change, in ppm, from the last readings taken at the 1000 hour point. The big disappointment was the 2.5V reference and the 100mV reference, which is derived from the 2.5V. The reading is outside the allowed range of the DVM. 80% of the change appears to be due to a big shift in the 2.5V reference voltage.

    There are some differences: I believe that the 2.5V reference is a band-gap reference, where the other two references are derived from buried Zener diodes; the 2.5V reference is the only reference that is soldered to the PCB -- the other two are plastic DIP packages that are socketed.

    The 10V reference is truly stellar with only a -2ppm drift.

    The Voltage Reference using Chinese parts:

    2020-02-20 after 2 hrs warmup:

    Value              Range      Change
    100.0043mV   100mV    (+118ppm)
    1.000020V       1V          (+42ppm)
    2.50021V        10V         (+92ppm)
    10.00017V       10V          (-4ppm)

    Again, the big loser is the TI 2.5V reference (both of the TI reference ICs were sourced by Digikey); and the winner is the LT1021-10. The amount of change in the 2.5V reference is nearly the same percent (and direction) between the two PCBs. 

    Current Reference:

    9.99935mA (-13ppm change)

    0.999983mA (-11ppm change)

    100.002µA (-12ppm change)

    The current reference is based upon the LT1021-10 reference, so it is not surprising that the differences are quite small.


    Stick with the buried Zener reference and don't solder them into the PCB.

    [Edit 2020-02-21: Added the following.]


    The day after getting the above results I took more measurements and got essentially the same results. But the 2.5V reference was not changing with the application of power, so I'm thinking it might be worthwhile to just recalibrate the 100mV reference and continue onward, which is what I did. Here's the new readings to compare with future data:

    Digikey Box:

    99.99952mV (averaged)





    AliExpress Box:

    100.00077mV (averaged)




    I discovered that the DVM will calculate mean and SD from the readings. It is nigh impossible to get consistent readings from the 100mV source. Small temperature variations cause galvanic errors and then there is noise. Expect ±2µV uncertainty.

    The reference boxes are going back on the shelf for a few months. We'll take them out and measure again near the first anniversary of the DVM.

  • Do cheap components make a difference?

    Bud Bennett08/28/2019 at 19:55 0 comments

    I decided to make another voltage reference, this time using the cheapest components that I could find on eBay or Aliexpress. I even ordered the enclosure from an Asian source. Here are the results after more than 1000hrs.

    2019-07-14 after running for 2 hours at temperature

    Meter params: HighZ, 100PLC, AZ on

    Value Range
    100.0004mV 100mV
    1.000007V 1V
    2.50014V 10V
    9.9999995V 10V (last digit bouncing)

    2019-07-21 after running for 168 hours at temperature

    Meter params: HighZ, 100PLC, AZ on

    Value Range
    99.9912mV 100mV
    1.000014V 1V
    2.50001V 10V
    10.00008V 10V 

    2019-08-28 1096 hrs:

    Value Range
    99.9925mV 100mV  (79ppm)
    0.999978V 1V (36ppm)
    2.49998V 10V (64ppm)
    10.00021V 10V  (21ppm)

    There doesn't appear to be any significant difference.

  • 1000 Hrs!

    Bud Bennett07/02/2019 at 02:52 0 comments

    The voltage reference has been baking for 1000 hours, give or take a few. The power went down for a few hours during the test, but I don't think it matters much. Here's the results, using the same conditions as when the test began:

    2019-07-01 1000hrs at temp

    Set Point: = 0.627049

    Value         Range    Aging Drift
    100.0065    100mV   9 ppm ±2ppm
    1.000023   1V           18 ppm ±2ppm
    2.49993     10V         16 ppm ±8ppm
    4.99922     10V         18 ppm ±4ppm
    10.00007    10V          1 ppm ±2ppm

    I find it a bit suspicious that all of the voltages drifted upward over time. That would indicate a drift in the measurement tool more than the references. I put tolerances on the aging terms just to account for the uncertainty of the last digit over two measurements. 

    I don't really see a need to open the box to perform any trimming.  

  • 168 Hr Reference Aging Data

    Bud Bennett05/28/2019 at 21:21 0 comments

    It has been a week with the references running at the heater temp continuously. Here's the data (taken after the DVM ran for 1 hr at room temp):

    2019-05-28 168hrs at temp

    Meter params: HighZ, 100PLC, AZ on (Null for 1V and 100mV ranges)

    Set Point: = 0.62785

    Value            Range
    100.0068      100mV
    1.000029      1V
    2.49991         10V
    4.99923        10V
    10.00006       10V

    So far so good. The 100mV reference drifted up by 1.1µV (0.0011%), tracking the upward movement of the 2.5V reference. The 5V and 1V references drifted up by 0.002%. The 10V reference did not move.

    The box will remain powered for another 5 weeks or so.

  • Voltage Reference Results!

    Bud Bennett05/21/2019 at 19:15 1 comment


    I was careful to solder all of the surface mounted parts except the REF5025, which I hand soldered to the PCB using only the 5 pins that were connected to the circuit. I made sure to remove all of the flux with alcohol and a toothbrush. The through-hole parts were then hand soldered and the flux cleaned up as well. The LT1021 references were inserted into gold plated sockets. When power was applied all of the circuitry functioned as designed at first glance.

    Two Problems:

    I spent a couple of days chasing down bad assumptions.

    The 2.5V Reference:

    The TI REF5025 reference datasheet calls out the following circuit for the trim, which I copied without question:

    When I first applied power to the board the two LTC1021 references measured within a few mV of the target value. The REF5025 was sitting at about 3.5V. My first thought was that I had a defective IC. I was able to trim the part in the range of 2.5V, but the trim was twitchy and even small tweaks to the trim pot generated several mV of change to the 2.5V output. 

    I Googled "REF5025 trim error" and the search brought forth many references to the data sheet -- even a data sheet with a Burr-Brown logo, which must have been published back in the last millennium. Nobody was complaining about the trim circuit.

    I then removed the 470Ω ressitor from the circuit and measured the untrimmed output voltage -- which was within a couple mV of the correct value. I also measured the resistance of the TRIM/NR pid to GND with the circuit unpowered: 1MegΩ!

    So now I'm thinking that there is some serious error in the data sheet and that the 470Ω resistor should be in the range of 470kΩ instead. So I soldered a 470kΩ resistor to replace the 470Ω and viola! (I used to play the viola) the trim became much more manageable. I could now trim the 2.5V output to less than a 50µV error.

    The Heater:

    The entire heater/reference system was supposed to be assembled into a small box with insulating foam to isolate the PCBs from the ambient temperature. When I tested the heater circuitry I used an open cell foam which probably does not have a very good R-factor. I used a styrofoam packing material when I assembled the two PCBs into the 100x68x50mm plastic box that I intended to use for this fixture.

    The component values that worked for the bench testing did not work for the new box configuration. The improved insulation and enclosed box allowed for much lower power dissipation in the heater to reach the same temperature for the set point.

    The first cycle with the breadboard component values required about two hours to settle to the set point temperature, but the step response indicated that the system was pretty stable with a phase margin of better than 45 degrees. But 2 hours is too long, so I replaced the resistor, R59, that was setting the integrator time constant with a smaller value, 100kΩ instead of 680kΩ, to decrease the overall time to settle. The result was a shorter settling time, but less phase margin (maybe less than 25 degrees). This is what the response to a perturbation looks like:

    Pretty ringy-dingy, but it still settles to the correct set point in about 30 minutes. This data is produced by the Keysight 34461A DMM, which doesn't attach a timestamp to the data -- go figure -- I had to assume a sample rate for the above plot to get the time information for the x-axis. If this was a design meant to be for production I would never go for this, but for a one-off it might just work. It's pretty clear that my attempt to model the physical and electrical system was way off, but it's not worth the time to figure it out for just one instance.

    OTJ Learning:

    I trimmed the references to the target values, at room temperature (about 12°C in my shop), and then enclosed the heater/reference PCBs within the box. I let it...

    Read more »

  • Results!

    Bud Bennett05/15/2019 at 23:22 0 comments

    The Keysight 34461A arrived today! I completed the current reference board yesterday...almost perfect timing. I plugged in the 34461A and let it warm up for a couple of hours at room temp (72°F/22.2°C). I adjusted the current reference potentiometer to put the three current reading errors approximately centered between the three readings. I took these photos of the front panel for each of the three  current reference values, with Aperture set to 100 60Hz line cycles, and Auto Zero on.

    I received a calibration certificate with the unit. Here's the relevant data:

         DCI 10mA on 10mA range. Reading: 0.0100000 Error: -0.0003%. 

    I don't know what these guys are smoking, but the above line of text doesn't make any sense. I believe that if they are truly reading 10mA on the 10mA range then the reading should be 9.99997. It seems that they have the DMM range set to 1A, which would yield the correct reading of 0.01 for a 10mA current. They did not calibrate for lower values than 10mA! Let's assume that the 34461A is perfectly accurate and go from there. 

    The first reading on the 100µA range has an error of +0.0018% (90 day spec = ±0.065%), that's 1.8nA. The second reading on the 1mA range has an error of -0.0028% (90 day spec = 0.036%. And the third reading on the 10mA range has an error of -0.0052% (90 day spec = 0.032%).

    So it appears that the current reference accuracy is at least 6x better than the tolerance of the 34461A. Not bad!

    [Edit 2019-05-21]:

    I found a major flaw in the layout when I proceeded to take a 10mA measurement with my old HP972A DMM. The ground lead of the 1000µF filter capacitor is dangerously close to the common probe pin. The probe slipped off the lead and shorted the meter across the 15V supply. That blew the protection fuse and killed the BAV70 diode clamp on the 10mA output. That was the first time that I'd ever blow a fuse on a meter...since the 1960's!

    I desoldered the filter cap and installed insulating sleeves over both leads before reattaching it. 

    The other problem that I noticed is that the HP972A measures the 1mA source inaccurately -- about 20% low on the 4000µA range, but gets the correct value (0nly two digits: 0.99mA) on the 40mA range. It turns out that the meter uses a 550Ω shunt resistor to measure the current on the 400µA and 4000µA ranges, and an 8Ω resistor to measure current on the 40mA and 400mA ranges. The 550Ω resistor develops 0.55V across it when measuring 1mA -- this is enough to turn on the BAV70 diode and siphon off some current to the supply. The solution is to put a small battery in series with the DMM to boost the voltage above the supply. A 1.5V cell should do the trick. 

    If I was going to redesign the circuit I would add another BAV70 diode between the anodes of the three clamp diodes and the positive supply rail. That would provide a compliance of nearly 1V.

  • Experimenting With The Heater PID Controller

    Bud Bennett05/03/2019 at 00:01 2 comments

    Before I committed to fabricating the PCBs for the voltage reference I decided to conduct a simple experiment to get a handle on possible problems controlling the temperature of the enclosure. I chose two CA3140 opamps from my inventory (probably 40 years old) for the two opamps. I also found some LM34 temperature sensors in metal cans that probably hailed from the 1970s. A IRLZ44 FET completed the component search and I integrated them all into the following schematic.

    The LM34 was taped to the middle of the array of resistors that made up R4. U2 provides a 1.13V reference for the LM34, which should cause the PID to regulate the temperature of R4 at about 113°F. 

    The 4.5V drain voltage on M1 in combination with R4 should generate 5.4W of power across R4 to lessen the amount of time required for the heater to reach temperature. 

    I wired the LM34 with 8 inches of lead wire and taped it to the center of the PCB containing the R4 resistors.  I then wrapped the enclosure with Duck tape to seal the edges of the foam enclosure around the R4 PCB.

    I flipped the switch on the two power supplies and watched the show. It was pretty boring. The temperature of R4 slowly ramped from 56°F to about 115°F in about 15 minutes. When my temperature probe indicated 120°F the gate of M1 started to drop to about 5.6V and the PID took over the temperature regulation of R4.

    For the next 15 minutes I watched the display of the temperature probe. It varied from about 119.6°F to 120°F. That's only 0.22°C. I observed no anomalous behavior or strange oscillation of either the gate voltage of M1 or the temperature reading of the probe.

    It appears that the PID controller is ready for prime time.

    [Edit 2019-05-12]:

    Later in the day I attempted the same experiment with the PID temperature controller. The shop temperature had increased to around 60°F by then. When I applied power to the bread board it oscillated rail-to-rail -- not what I expected. I determined that the cause for the stable response was that the Z44 FET was in its triode region and the system gain was relatively low. When the ambient temperature increased, the FET came out of its triode (linear) region, which increased the open-loop gain.

    I also found that the opamp U4 was oscillating because of the load capacitance -- the drain-gate capacitance of the FET. This was fixed with a resistor divider 10K on top, 12k on bottom, between the opamp output and the FET gate. Unfortunately, the system now oscillated rail-to-rail with a 5 minute period. After 2 days of fiddling with the breadboard I determined that the problem was a 60 second delay in the heater. A delay is really bad news since it is an exponential phase error. A 60 second delay is totally awful. I was unable to get a stable system by substituting components and waiting 20 minutes for the system to settle into a steady-state bang-bang oscillation.

    Now I turned to LTSpice to simulate the problem. LTSpice will yield a reasonably accurate result (if your models are accurate) in a few seconds instead of waiting 15 minutes on the bench. Here's what I eventually came up with as a working model:

    The heater is modeled with a behavioral voltage source, B1, that converts power to voltage to emulate what the LM34 is doing. B1 also inlcudes the 60-second delay in the heater. The RC network tries to model the time constants involved with converting power to temperature. Heat transfer is a linear differential equation which can be modeled as a simple RC time constant between each physical interface. Note that C2 is now 200µF, a considerable increase from the initial 4.7µF. V3 simulates a changing ambient temperature from 50°F to 80°F to get a feel for how the loop gain affects stability as the ambient temperature changes. This circuit accurately modeled what I was seeing on the bench...

    Read more »

  • A Current Reference Too

    Bud Bennett05/01/2019 at 23:00 0 comments

    It was inevitable. Checking voltage accuracy is all well and good, but not enough. I have been thinking about how to provide a set of reference currents that are accurate enough to keep the 34461A honest for a few years. It's tough, but doable. The 34461A specs for DC current are:

    I tried to design for four reference currents, but gave up on the 100mA current -- too much power dissipation (1W) with the approach that I chose. So the current reference will provide only three currents: 10mA, 1mA and 100µA. In much of my work the accuracy of currents exceeding 10mA is usually not critical. The circuit below is what I came up with after a few revisions.

    Current is just V/R. You need a good V and some good R's. This is a lot different that making an accurate voltage reference where ratios of resistors are important. In this case the absolute value of the voltage reference and the resistances determine the accuracy of the current. I chose the LT1021C-10 for the voltage reference. It has an initial accuracy of 10V±5mV. I added a trim to improve it, hoping to get 0.1mV trimmed accuracy.

    I entertained the thought of just having the DMM measure the current through the resistor directly, but the burden voltage is too high and injects an unspecified error into the measurement. In the circuit above the DMM will connect between the COM pin and the current source to be measured — there is no need for large voltage swings. This makes the design a lot easier than a general purpose current source. Note that when the a current source is not being measured its current flows into a diode to limit the voltage swing to less than 1V below the supply rail, keeping the opamp and FET operating normally. When the DMM is measuring the current source the voltage across the diode will only be a few mV and the diode will have a negligible effect on the measurement.

    The resistors, R1-R3, are expensive, 0.01% initial tolerance, 10ppm/°C tempco, through-hole types (~$7/each from Digikey). If the opamp has low offset and drift then the current in the drain of the FET transistors, M1-M3, should be very close to V/R.  I chose the MCP6V51 chopper stabilized amplifiers for this reasoning: low Vos (15µV max), low Vos drift (36nV°C), very low input bias current (80pA), single supply operation to 45V, with moderate GBW, low supply current (470µA), in a small SOT23-5 package. 

    With a reference trimmed to 0.001% and resistors specified to 0.01% I expect the current sources to be on the order of 0.01%, if I don't screw up the PCB layout. This assumes that there is no significant leakage from gate-to-source or gate-to-drain from the AO3400 FET ( I checked -- the data sheet says < 100nA @ 12V VGS), so I'm hoping that there is less than 1nA of gate leakage current with room temperature conditions and lower gate-source voltage.

    There are a lot of things to worry about with the PCB layout -- is the reference voltage compromised by load currents through traces?, or is the trace resistance low enough to not degrade the tolerance of the expensive resistors? Or are there parasitic leakage currents from nearby traces?

    I added guard rings around the sensitive nodes -- the probe points used for the currents, and the high-voltage terminals of the R1-R3 resistors. Elsewhere I made sure that the PCB trace resistance was negligible where it made a difference (the GND terminations of R1-R3). The GND pin of the LT1021 connects to the GND trace of R1-R3 with little current flowing in the trace, so voltage error should be minimized. Also, the connections of the opamp and the source of the FET are Kelvin to the top pin of the resistor.

    If the PCB parasitics are eliminated (or at least minimized) then all that’s left is the errors from the voltage sources (the Vref and opamps). The 80 pA input bias current of the opamp, and the value of the resistor. I expect the Vref to be trimmed to around ±0.1mV (±0.001%)....

    Read more »

  • Adding A 1V Reference Voltage

    Bud Bennett04/29/2019 at 21:44 0 comments

    I’ve been bothered about the choices for reference voltages lately. My new Keysight 34461A has voltage ranges of 0.1V, 1V, and 10V. My old HP972A (a freebie) has ranges of 40mV, 400mV, 4V, 40V, etc. If there is not another reference around 1.0V the 34461A I will not be able to test that range effectively. The 972A can use the 0.1V, 2.5V, and 10V references to cover most of its ranges. It’s interesting that neither the HP or Keysight DMMs need the 5V reference for calibration.

    The easiest way to get 1V is to divide it down from another reference voltage. I chose to make a 1:5 divider from 5V to get the 1V reference. The divider uses the same 20kΩ resistors that went into the 0.1V reference. It will not use 50 resistors (there’s no room on the PCB) and will require the loading of the DMM to be higher than 10MegΩ — therefore only useful to the 34461A. 

    The accuracy of the divider is expected to be only around 0.05% (0.1%/√5). But the 5V reference can be trimmed to yield the correct value at the 1V output if the 5V output is not expected to be used as a voltage reference for calibration. That’s the plan.

    Either the 5V or 1V references can be brought out to terminals with the current PCB layout, but I expect to not use the 5V option, given the ranges of my current collection of DMMs.

    Both the layout and schematic have been updated in the project details section to reflect this change.

View all 14 project logs

Enjoy this project?



gelubotezan wrote 02/19/2020 at 16:50 point

Reading the logs I noticed you were wondering why your reference was suddenly (after a black out or something) beginning to drift upwards.
I remember I saw  in a DS  from DA or LT that after 100-200 hours the big drift ceased and  the reference began a drunkard's walk. They say this  can't be controlled but it will stay up/down in the accuracy limits of the device. A similar report I found regarding the OCXOs some time ago. Following, interested, nice! 

  Are you sure? yes | no

Bud Bennett wrote 02/19/2020 at 22:24 point

Thanks for the info. I only ran both reference boxes for 1000 hours. They've been resting, unpowered, since. It's about time to test  them against the DVM again.

  Are you sure? yes | no

Bud Bennett wrote 02/20/2020 at 16:05 point

That's a pretty short App Note. It only talks about long term drift with under continuous conditions. I was wondering how things would go when the references were turned on and off occasionally after the initial 1000 hour burn-in. I found this Design Note that gives a hint as to what to expect: 

  Are you sure? yes | no

Ken Yap wrote 05/02/2019 at 04:36 point

What will you use for the initial calibration check? The meter? So it's to check later if the meter has drifted and you're assuming your reference hasn't?

  Are you sure? yes | no

Bud Bennett wrote 05/02/2019 at 05:08 point

Yes. The meter will be fresh from the factory calibration. I’ll use the meter’s readings to trim the reference’s initial values. Then I intend to run the heated voltage reference for at least 1000hrs and re-trim it if it has drifted. After that I’ll check the DMM occasionally against the references. If the DMM measurement is out of spec then I can trundle down to my old employer where they maintain calibrated equipment to see which unit is out of calibration. If the DMM keeps measuring reasonable values from the references then I’ll continue onward blissfully ignorant.

  Are you sure? yes | no

Ken Yap wrote 05/02/2019 at 05:19 point

Fair enough. I guess the chance of a false negative is far less than that of a false positive.

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates