Hello to all the members
There has been considerable discussion on the time-nuts site regarding the HP3458, which I would like to add to my other HP DVM's. This is a problem due to the relative rarity, and therefore cost in the UK. But I have got a 3457 and a 3456 in my lab.
What is the general opinion of the 3457 - is it a good deal short of the 3458 - I gather it is all down to the voltage reference device.
It would seem that obtaining a good DC voltage standard (mine is a Fluke 731B), is much more difficult than the frequency standard, with a heap of GPS and other gear at inexpensive price levels. Any good designs for this purpose, or is it best to find commercial products.
Any advice on these matters would be welcome.
Roy
Roy Phillips wrote:
Hello to all the members
There has been considerable discussion on the time-nuts site regarding the HP3458, which I would like to add to my other HP DVM's. This is a problem due to the relative rarity, and therefore cost in the UK. But I have got a 3457 and a 3456 in my lab.
What is the general opinion of the 3457 - is it a good deal short of the 3458 - I gather it is all down to the voltage reference device.
It would seem that obtaining a good DC voltage standard (mine is a Fluke 731B), is much more difficult than the frequency standard, with a heap of GPS and other gear at inexpensive price levels. Any good designs for this purpose, or is it best to find commercial products.
Any advice on these matters would be welcome.
Roy
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
Its up to the references and the A/D converters have to be very linear.
I recently had my cal lab to do a test on the 3457A (7-7-2009). I had
them to hook it up to their calibrator system, A Fluke 5700A, which they
rated at 1.0000000V at 8 ppm and I had them to let the meter average the
measurement for 10 minutes. The mean reading was 1.000002V and the STD
Dev was 137.3x10-9. I also had them to read 10V, using their Fluke 732B
which was certified at 9.999959V at +/- 0.4 ppm. The instrument read
9.999956, the STD Dev was 5.358x10-6 for ten minutes. I read somewhere
that HP rates the instrument at 5 ppm, and from the 10V readings that
appears to be very close.
The 3456A is rated 2 ppm, but you can only read 6 digits.
Brian - KD4FM
The seller states it as "untested" and/or claims not to have even
bothered to power it up and run self-test.
...Note that many such boxes have hardware failures. Your original
acquisition cost + repair can easily exceed cost of buying a fully
tested unit.
The seller's photo shows it as powered up but you can see "Err" in small
letters below the main characters in the display.
... "Err" means an error was reported and the seller would need
to hit "Blue Shift" Error to find out which error(s).
Watch out for the Version of its A5 Outguard Controller:
---=====
Old version is 03458-66505 (first released in 1988)
or 03458-66515 if opt. 001
It has 3 SRAMs, each having a non-replaceable internal battery that
lasts 10 to 15 years. When the batteries crash, you'll loose your
cal constants and your DMM won't work. The fix is to either replace
the 3 SRAMS (expensive) or upgrade to the new ver. A5
The SRAMS are qty 2 DS1230Y and qty 1 DS1220Y.
Check the datecodes on them.
Bottom line: If date codes are more than 10 years ago,
you're living on borrowed time!
Firmware might be very old.
...depends on age but might be between ver. 2 to ver. 8.
New version is 03458-66547
or 03458-66548 if opt. 001
Much better! Uses new version Non-volatile RAM
with user replaceable batteries!
Has much newer version firmware. (Probably ver. 8 or later.)
Cheers,
Greg
IMO the 3457A, 3456A and 34401A are all excellent DVMs. All three DVMs must
be adjusted at many cardinal points for DCV, ACV and Ohms. Of course these
adjustments are via physical pots in the 3456A, whereas most are via
electronic cal constants in the 3457A and 34401A.
I would say that the 3457A is a little bit the predecessor of the thinking
that went into the development of the 3458A.
But the 3458A is still king by a thousand miles. Nothing else comes close.
Its internal "ACAL" routines take advantage of the 0.02ppm linearity of its
DAC, establishing ratios from one range to the next (and from its LTZ1000A
zener reference voltage to full-scale on its 10V range). Consequently its
adjustments to the external world require only a short, a 10Vdc standard and
a 10k ohm resistor. It even adjusts its own AC freq. response (during ACAL
AC) by adjusting for the most pure flat top of an internally generated
square wave. Its most accurate AC mode (AC SYNC) digitizes AC sinewaves and
then calculates and displays their voltage.
After a 3458A has been adjusted at "0", 10Vdc and 10k ohms (and SCAL AC if
needed for AC HF frequency response above 2MHz) - and after internal ACAL -
it will meet all published specs. (Or else it's broke and needs repair
because there's nothing else to adjust.) It's had an amazing long life (21
years) so far.
Cheers,
Greg