[time-nuts] Voltage standards

WB6BNQ wb6bnq at cox.net
Thu May 26 23:18:10 UTC 2011


Hello Will,

I think you meant to post this on the "Volt-Nuts" list ?

The reference parts used in the Fluke 731/732 are not "off" the shelf parts, they
are selected.  While they are referred to as transfer devices, if working
correctly, they exceeded their specs versus time on the 10 volt port.  I base
this on the history of over 50 units that were tracked by Navy's Primary Lab, as
these were used to feed the second level labs in the Navy's system.

Bill....WB6BNQ


Will Matney wrote:

> Hello all,
>
> I was reading some old posts about voltage standards from Dr. Kirby, and
> some others, and one that was mentioned was the older Fluke standards. I
> did some research on this, going back to the Fluke 730A transfer standard,
> which I have one of, and found out what the voltage reference and the
> differential amplifier was from an old Motorola parts book dating from
> 1968.
>
> The voltage reference in the Fluke 730A was most likely the Motorola part
> number: MCA1914N or the MCA1924N in a can package. Both are rated at 6.8
> Vdc, but there is a difference in the maximum voltage change between the
> two. The MCA1914N is the tightest, with a change of only 0.005 volts from 0
> to 75 deg C. The MCA1924N is made for a broader temperature, with a change
> of only 0.010 volts from -55 to 100 deg C. These are not an oven zener, but
> a voltage reference. Motorola first produced these using two diodes and a
> transistor sealed in a can, literally. They used a temperature compensated
> zener diode, consisting of a zener, and a regular diode, turned with their
> cathodes facing, and connected together. These were placed inside the can,
> along with a transistor across from them, then sealed. They used this same
> reference transistor on up through several transfer standards, including
> the 732A.
>
> The Fluke 730A did not use an op-amp after the reference, but used a dual
> transistor in a can, and created their own comparator amp. Motorola made
> several of these along with Fairchild. There are two possibilities here in
> the parts. I think they used a MD918A or MD918B, which are dual NPN silicon
> annular transistors, with a Vceo of 15 Vdc, and an Ic of 50 mA. The MD918
> series was to be used for "differential amplifier applications requiring a
> matched pair of transistors with a high degree of parameter uniformity
> under varying environmental conditions".
>
> The other possibility on the dual transistor was a 2N2914, which was a
> "dual npn silicon annular transistor, especially designed for low-level,
> low noise, differential amplifier applications, featuring very high Beta,
> guaranteed from 10 uAdc to 1 mAdc, and excellent noise characteristics".
>
> Next, the Fluke 731B used the same voltage reference, but it did not use
> the dual transistor amp. Fluke changed this to an op-amp, which turns out
> to be a LM308. This op-amp was then used in the newer models after the
> 731B. Also, this op amp used a few of the same resistor values, and
> circuitry, as the two transistor amp.
>
> Now, a word about "voltage transfer standards". They are not meant to be as
> stable as a true primary voltage standard, as they are meant to be used to
> "transfer" a voltage from a known source to other equipment being
> calibrated. The transfer standard is made to be recalibrated in one year
> increments or closer. Tight tolerances in the short term, in one year or
> less, are all that is needed. The Fluke 730A, 731B, and 732A and B are all
> transfer units. The small voltage transfer standard boards, now sold on
> ebay, are meant for calibration every six months, and are not meant to be a
> long term primary standard. They should be calibrated from another transfer
> standard, which has been calibrated from a primary voltage standard, such
> as a Josephson type, by Fluke, or the NIST. The ones calibrated by using a
> DMM, no matter who made it, are not calibrated correctly. To calibrate a
> board correctly requires the use of a voltage transfer standard, that was
> calibrated by a primary voltage standard, and a very sensitive null
> meter-detector such as a Fluke 845A, etc.
>
> Hope this helps the ones who had a few questions.
>
> Will
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.




More information about the time-nuts mailing list