[volt-nuts] Wanted - UK source of a replacement battery (type LX 1634) for HP 3457A 6.5 digit multimeter

[Dr. David Kirkby (Kirkby Microwave Ltd)]

On 25 Jan 2016 03:51, "Tom Miller" <tmiller11147 at verizon.net> wrote:
>
> Looking through the service manual I don't see a connection from the
battery to the NVRAM. I am sure I measured battery voltage on pin28 so I
suspect a schematic error. Can someone that has one open verify that? My
3457 is in cal and I prefer not to break the seals.

NVRAM (non-volatile random access memory), an example of which is flash
ram, doesn't need any power source to hold the data.  The 3457A has SRAM
(static random access memory), which does.

Using my old Tektronix 4.5 digit multimeter I definitely measured a voltage
on positive voltage on pin 28 which was around 200 mV less than the
battery.  I believe it must be via a high impedance path as my meter, which
I assume is 10 M Ohm input impedance,  is loading the circuit and so one
sees the voltage drop slowly although the battery voltage doesn't change.

When I put the mains power on, the voltage on the SRAM rose to about 4.8 V.
When the mains power is removed the voltage stays well over 4 V but
gradually drops in voltage.  I assume that a decoupling capacitor has been
charged to 4.8 V, so the voltages across the SRAM doesn't immediately fall
back to 2.8 V. Without having acess to an electrometer or other very high
input Z multimeter I could not say for sure, but I suspect that if one was
fairly quick (of the order of minute or two) one could probably just cut
the old battery out, solder a new one in, without loosing the contents of
the SRAM. But one would not want to take that chance if one considered it
was important to keep the data.

I am actually thinking whether it is better for *me* to actually purposely
let the SRAM contents be lost, but my reasons for this world form another
thread.

Dave.