Geller's specs are not is good as you have quoted. The 10 volts +/-10uv
is very, very short term.
BUT, it really depends what you are looking to do with the voltage
references.....it may meet your requirements.
From his site:
SVR Specifications:
Temperature Coefficient "L" grade: 5 ppm / ºC
Short term transfer accuracy: better than +/- .0005% absolute (+/-
5 ppm).
Short term transfer accuracy applies in the short term at the
reported temperature and at the specified power supply voltage (15.0V
for 10V SVR references) after of warm up period of 30 to 60 minutes, <10
ppm within 5 minutes.
All reference boards and boxes are calibrated to within +/- 10 uV
absolute and guaranteed to remain longer term (6 months) within +/- 500
uV (+/- .005%) of absolute as rated.
Extra trim resistors are added to narrow the trim range for easier
calibration. Calibration data is reported to 10 uV resolution (.0001%).
The AD587 series reference chips are rated at 15 ppm / 1,000 hours of
operation. Very close inspection of most monolithic reference chip
family specification sheets reveals a 1,000 hour rating generally
ranging from 10 to 60 ppm / 1,000 hours.
end of his quotes =======
if you really need a good voltage reference, consider the LM199 series
which is affordable. But you will have to burn it in and have it
calibrated. These are heated buried zeners and were used in the HP3456A
(2ppm) and HP3457A (5ppm) multimeters, and others. The LM399A is
rated at 1 ppm/C and has a time stability of 8-20 ppm a year.
if you have a little more money to spend, consider the LTZ1000, which is
used in the HP3458A multimeter, and the national Instruments PXI-4071
DMM. The LTZ1000 is rated at 0.3ppm/C and has a time stability of 8 PPM
a year.
You will also find, that maintaining voltage references are a little
more difficult then time/frequency - we all have GPS or LORAN to verify
our time standards, etc. You either have to have the equipment to
monitor the voltage references, or send them out for calibration. I
prefer to leave standards/references where they are for stability
reasons, and send the test equipment out for regular calibration. You
have the same environmental problems, you name a stable temperature
environment to operate them in.
Linear, the IC manufacturer, has many applications notes on the
LM199/399 and LTZ1000 series - and they can show you how to build a good
reference and they can save you a lot of common pitfalls. When you get
into measuring microvolts at 6 to 10 volt levels you have to account
Kelvin/Seebeck thermal and galvanic effects.
If your requirements are less than what I have mentioned here, check on
the Xicor X60008 (Intersil has bought them). They make a reference
thats good to a half a millivolt level at 5V, and it has some promising
specs. I have 5 units they sent me for evaluation that have been
burning in for 2 years.
If you have the instrumentation, you can also pick up a standard cell
off of fleabay - but make sure the voltage is good - I've seen too many
of these units sold and they are below specifications for using as a
transfer standard. If they can not tell you the voltage of the cell,
don't buy it. Any body with a decent DVM can check one quickly to see
if its OK. Nominal is 1.0183 volts - and there are differences between
saturated and unsaturated cells that you need to read up on.
You may consider purchasing a HP3456 or 3457A off of fleabay and put it
in a calibration program. I have picked these meters up for around
$150, and the local cal lab here calibrates the 3456A for $200 and the
3457A for $160. They give me before and after data and I cal them every
2 years. You can also pick up Fluke voltage calibrators or differential
voltmeters, too from fleabay...
These multimeters can also measure your "calibrated resistors" with
extremely good accuracy - they can use what is called 4 wire
measurements - where they compensate for voltage drop in the measuring
leads.
In "the good old days" Mallory (Duracell) use to sell a mercury battery
voltage reference that was the workhorse of many labs. It was like 10
cells in series and they brought out each tap. But since mecury is bad
news now...they are a part of history.
Brian KD4FM
Ed Palmer wrote:
> It's nowhere near the idea of a Josephson array, but if a NIST-traceable 10V
> +-10uV reference is good enough to satisfy your voltage-nut urges, you can
> buy it from www.gellerlabs.com for $35.
>
> I also have a few standard resistors (e.g. 1.000002 ohms) that I'd be
> interested in calibrating, but I can't seem to come up with a practical way
> of doing it.
>
> Frustrating, isn't it? :-)
>
> Ed
>
> Scott Burris wrote:
>
>> Now that many of us have a nice 10Mhz reference courtesy of TAPR,
>> I was wondering if there was any way to use that to build a precise
>> voltage or resistance standard?
>>
>> I've got once of those high precision standard resistors with a sticker
>> on it noting the actual measured resistance. Is it still accurate? Who
>> knows?
>>
>> As well, I have a +5v reference that uses an Analog Devices precision
>> reference chip as its source. I have more faith that this reference is
>> correct within the tolerances specified in the datasheet.
>>
>> Now if I could somehow take that frequency reference and derive a
>> voltage standard or the like, I'd be in business. But I can't think
>> of a way that wouldn't require calibration of some sort, and if I had
>> the means to calibrate, I wouldn't need the standard in the first place.
>>
>> Any voltage-nuts or resistance-nuts out there?
>>
>> Scott
>>
>
>
> _______________________________________________
> time-nuts mailing list -- time-nuts@febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
>
_______________________________________________
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
