HP 34401A settle time that high?

[Penguin36]

Hi,

I bought a used HP 34401A on ebay and compared it to my good old 3478A using a DIY 5V and 10V voltage reference based on REF102.
After 15min warm up the 3478A is spot on and drifts only by 1-2 digits after 2 hours, the 34401A drifts too and needs 2 hours to stay nearly stable. See table ...

It there a problem with my 34401A or is this normal?

34401A performs all tests with no errors, LM399 is warm, I assume 40°C (but not hot!), ref voltage measured on U400 pin 3 and 7 is 6.9401V and stable, REF10 is at about 9.904V.

Any hints?

[Dr. Frank]

Hi,

I bought a used HP 34401A on ebay and compared it to my good old 3478A using a DIY 5V and 10V voltage reference based on REF102.
After 15min warm up the 3478A is spot on and drifts only by 1-2 digits after 2 hours, the 34401A drifts too and needs 2 hours to stay nearly stable. See table ...

It there a problem with my 34401A or is this normal?

34401A performs all tests with no errors, LM399 is warm, I assume 40°C (but not hot!), ref voltage measured on U400 pin 3 and 7 is 6.9401V and stable, REF10 is at about 9.904V.

Any hints?

You're comparing a 5 1/2 digit vs. a 6 1/2 digit DMM, and you're comparing a REF102 (2.5ppm/°C) vs. an LM399 (<1ppm/°C, essentially). That's not a fair comparison.

So from your measurements, the unknown stability of your REF102, and also from the unknown and probably uncontrolled ambient conditions, it's not possible to tell, whether your new 34401A drifts excessively, or - more probably - your REF102.

The 34401A usually needs about 2h to fully stabilize - but you need constant room temperature to achieve ~ppm stability.

Unfortunately, I don't have a warm-up diagram, vs. a really stable LTZ1000 reference for comparison available.

Frank

[Kleinstein]

When the new baught DMM was not used for a longer time, it may shown some initial drift for the first few days or weeks of use.

It naturally takes longer to get stable to 6 digits than to get stable to 5 digits. A warmup time of 2 hours is kind of normal for precision instruments.

[Penguin36]

Ok, the seller told me that the 34401A was a backup DMM for his lab, but he never used it. It was made in 2000, so if it really was not used it was not used for 17+ years! The DMM looks absolutely new.

So, should I just leave it powered on for the next days or weeks?

[Dr. Frank]

Ok, the seller told me that the 34401A was a backup DMM for his lab, but he never used it. It was made in 2000, so if it really was not used it was not used for 17+ years! The DMM looks absolutely new.

So, should I just leave it powered on for the next days or weeks?

The correct question, what do you want to measure, or how stable do you want to measure?
The 34401A is stable to a few tenths of a ppm, if the RT does not change.
But you also need a superior voltage reference to detect that.

As this is well below the specification limit of this instrument (about 20ppm ball park), don't waste money for energy and lifetime of the DMM.
The instrument seems to stabilize well within 2h to a few ppm, nothing more should be expected.
Also I doubt, that the LM399 reference shows such a big hysteresis from idle state, especially compared to the specification.


Frank

[chickenHeadKnob]

Ok, the seller told me that the 34401A was a backup DMM for his lab, but he never used it. It was made in 2000, so if it really was not used it was not used for 17+ years! The DMM looks absolutely new.

So, should I just leave it powered on for the next days or weeks?

What people are telling you is that every time you go for a cold start it will take the instrument 2hrs + to settle its internal temperature. Actually I observe my HP3457a maybe takes 3 hours or more.
Letting your unit run after a long time in storage will help in drying out the internal moisture levels. This will take about a week uptime or longer but will not change the cold start warm-up latency period.

[Penguin36]

Ok, I understand. I will let it warm up for +2h before use.

BTW, any quick hint which USB-RS232 adapter will work with the DMM? Just the cheap CH340 based ones?
Yes I know, GPIB is faster, but for now I want to start with RS232 :-)