HP 419A excessive input bias current

[guenthert]

I recently acquired a HP 419A in reasonable good nick (neons work perfectly, NiCd batteries still hold charge, Hg battery for bucket voltage is dead, glass of MCM lens is fractured), but if I'm not mistaken, the input bias current is more than 2nA (indicated voltage ~220uV across a 100kOhm R).  Afaict, the input bias current isn't specified for the HP 419A, but 2nA seems excessive for a null voltmeter, particularly if it is designed to also measure currents in the 30pA range.  At first glance it looks quite clean and pretty (gold traces!) inside (and any contamination shouldn't cause a current that large).  I need of course a new battery for the bucket voltage generator, but since the bucket voltage spans only 120% of the measuring range, it won't fix the issue.  What else could it be?
[shoot -- just realized, this should have gone into the 'repair' section]

[doktor pyta]

Hi,
I had similar issues with mine.
The cause was corroded battery holder (for small internal mercury battery used for offset adjust).
I cut off cables going to this holder as I'm not using this functionality.

[beanflying]

Just making another 1.35 Reference for mine. MAX 6341 divided across 3 Vishay low tempco 1.3k resistors with a trim pot similar to what I did for my 740B https://www.eevblog.com/forum/testgear/hewlett-packard-740b-dc-standard-digital-voltmeter-(and-740a)/msg1438536/#msg1438536

Also doing a swap to NiMh as my Nicads were dead but fortunately hadn't leaked.

Haven't looked at things like bias currents yet but mine should be up and going in a couple of days. @ArthurDent has one in running order might be worth a PM?

And in case you haven't read it yet this thread doktor pyta started is worth a read https://www.eevblog.com/forum/metrology/null-voltmeter/

[pelule]

beanflying: Just making another 1.35 Reference for mine.

Why such a effort?
Do just the same "update" as HP did for the later revisions, replacing the mercury cell by an 3.6V lithium battery.
It lasts very very long time and the advantage is: schematic and required values for change are available from HP.
/PeLuLe

[beanflying]

beanflying: Just making another 1.35 Reference for mine.

Why such a effort?
Do just the same "update" as HP did for the later revisions, replacing the mercury cell by an 3.6V lithium battery.
It lasts very very long time and the advantage is: schematic and required values for change are available from HP.
/PeLuLe

I have never seen the change option? Can you post it here?

The reason for the use of Mercury cells was there very very flat voltage across a large part of their capacity compared to other battery Chemistrys. The 740B of mine has a pair of smaller Mercury Cells for similar purposes to the Null Set on the 419A so it seems fairly simple to be heretical and add an IC to a HP Classic 

[guenthert]

Hi,
I have similar issues with mine.
The cause was corroded battery holder (for small internal mercury battery used for offset adjust).

Uh, yes, the unit is pretty clean, except in the area around the Hg cell: there's bad corrosion and some toxic looking stuff.  I got to get rubber gloves before getting close to that, yuck.

I cut off cables going to this holder as I'm not using this functionality.

Hmmh, yes, for low impedance sources it won't be needed.  How low is sufficient certainly depends on the desired measurement error and the input bias of the meter.  Would you happen to know, what to expect from a healthy 419A in VM/AM mode?

[guenthert]

Hi,
I have similar issues with mine.
The cause was corroded battery holder (for small internal mercury battery used for offset adjust).

Uh, yes, the unit is pretty clean, except in the area around the Hg cell: there's bad corrosion and some toxic looking stuff.  I got to get rubber gloves before getting close to that, yuck.

Ah, *shrugs*, live dangerously, die young and beautiful.  I desoldered those cables connected to the Hg cell and the apparent input bias current went down to some 100pA (10uV indicated across a 100kOhm R).  Clearly much better and certainly usable for some applications (feel like I owe you a beer or two), but not quite good just yet, but then it's getting into the region where it's difficult to measure already.

Later this week I should have some time to clean that up and do some more tests.  I'm particularly curious about the AC rejection and whether I'd have to replace some capacitors.

[doktor pyta]

Would you happen to know, what to expect from a healthy 419A in VM/AM mode?

Just for comparison:
my HP419A after 1h warmup on 3uV/30pA range, after zeroing with shorted input:

-input shorted: needle bounces between -0.1uV and +0.1uV
-input open: needle settles at +0.6uV (or +6pA)

[guenthert]

beanflying: Just making another 1.35 Reference for mine.

Why such a effort?
Do just the same "update" as HP did for the later revisions, replacing the mercury cell by an 3.6V lithium battery.
It lasts very very long time and the advantage is: schematic and required values for change are available from HP.
/PeLuLe

I replaced the battery with a CR123 Lithium primary cell, but my google-fu is failing me now:  I couldn't find the new schematics.  I suppose I could calculate the new values myself, but so far I didn't touch those resistors.  It kind of, sort of, works with the 3.6V (or so) Li battery instead of the 1.35V of the Hg one, but the NULL potentiometer is now very finicky and there is some drift (no, or not much, in VM mode).

Would you happen to remember where you saw the updated schematics?

[beanflying]

beanflying: Just making another 1.35 Reference for mine.

Why such a effort?
Do just the same "update" as HP did for the later revisions, replacing the mercury cell by an 3.6V lithium battery.
It lasts very very long time and the advantage is: schematic and required values for change are available from HP.
/PeLuLe

I replaced the battery with a CR123 Lithium primary cell, but my google-fu is failing me now:  I couldn't find the new schematics.  I suppose I could calculate the new values myself, but so far I didn't touch those resistors.  It kind of, sort of, works with the 3.6V (or so) Li battery instead of the 1.35V of the Hg one, but the NULL potentiometer is now very finicky and there is some drift (no, or not much, in VM mode).

Would you happen to remember where you saw the updated schematics?

You can use a Zinc Air battery (Wein Cell) they are the 'normal' replacement for Mercury Cells in the photographic area for older gear and fairly close to the performance and voltage of the original. They are not as good but might be a lower fuss alternate than dropping the 3.6V?

I am choosing to remove the extra battery as I had the parts and it is one less thing to have to worry about in the longer term, it will also be more accurate over time than a battery solution.

[guenthert]

You can use a Zinc Air battery (Wein Cell) they are the 'normal' replacement for Mercury Cells in the photographic area for older gear and fairly close to the performance and voltage of the original. They are not as good but might be a lower fuss alternate than dropping the 3.6V?

I am choosing to remove the extra battery as I had the parts and it is one less thing to have to worry about in the longer term, it will also be more accurate over time than a battery solution.

Zinc Oxide?  I know all about Zinc Oxide thanks to

Oh, Zinc Air.  Well, didn't think of that, but the primary cells (not the upcoming solution of all energy problems, the zinc oxide rechargeable battery) I've seen so far are either tiny (meant for hearing aids) or too large (for lanterns).
AA or CR123 seems to be the right form factor and a lithium primary (3V, not as I earlier erroneously wrote 3.6V) of that size has seemingly a capacity of 1300 to 1500mAh.  Couldn't find anything comparable with Zinc Air just yet (even though the voltage would be perfect).

[beanflying]

The current drawn on the Mercury cell or any substitute is fairly small. Seems you can buy up to 600mAh+ in Zinc Air chemistry https://www.master-instruments.com.au/category/Zinc_Air_Batteries/1538 should be available in most other countries.

[duak]

I have a few old cameras and wristwatches that were designed for mercury cells.  May I suggest Silver Oxide cells with a nominal 1.55 V?  One common size is the 357 that is used in 80's manual cameras and in many digital calipers.

Is it true that zinc-air cells for hearing aids are only good for a month or so?

Cheers,

[Rax]

beanflying: Just making another 1.35 Reference for mine.

Why such a effort?
Do just the same "update" as HP did for the later revisions, replacing the mercury cell by an 3.6V lithium battery.
It lasts very very long time and the advantage is: schematic and required values for change are available from HP.
/PeLuLe

Could you please point to the HP info - schematic etc. - on switching over to the new battery?

[beanflying]

Time is tight currently and I haven't ever done a diagram for it but the basic replacement lash up is here https://www.eevblog.com/forum/testgear/hewlett-packard-740b-dc-standard-digital-voltmeter-(and-740a)/msg1437987/#msg1437987 and the next few posts after.

The dumbed down version is still the Battery option I put above which should be ok too.