[volt-nuts] Old HP3458A - SN: 2823A 03939

[Frank Stellmach]

Hello,

to update such an old instrument you need the following components:

1EA DS1220AD-150+, about 8$ from maxim/Dallas or distributor
2EA DS1230Y-150+, about 30$
6EA 27C512-100, about 20$, STM
4EA 62256-80, about 10$
3 precision DIL sockets for nvSRAMs

I assume, the firmware will be the earliest one, (#2 ??)  and also the 
calibration will be old, but still valid, i.e. the calibration constants 
in the nv-SRAMs might not be corrupted yet.

You will have to read out the CALRAM content first over HPIB (described 
here and elsewhere), then (easily) desolder the 3 nv-SRAMs.
Replace these by the precision DIL sockets.
Copy the 2kB CALRAM content into the fresh one, and assemble these 3 RAMs.
The content of the 2x32KB RAMs will be initialized automatically.
And the calibration, especially for the higher frequency AC 
measurements, will be conserved for the next 10..20 years.
Also plug in these 4 SRAMs for cheap opt. 1. (148K memory)

After these updates - and only if you really need a fresh calibration, 
i.e. if you don't have access to precise 10V and 10kOhm, you may spend 
these  500€ for calibration by a HP or Fluke lab.

Firmware 9 you may buy from Keysight, but order the single EPROM (4MBit) 
version for about  30$, or so, but you have to split that content 
bytewise into the six EPROMs (all FW versions are obviously compatible 
concerning the different HW versions).
They take around 200$ for the old 6 EPROMs version only..

Maybe a friendly volt-nuts is able to assist at that point ;-)


Yep, the fan is quite loud, but might still work all right, so keep it.
Nobody has identified a better (less noisier) part yet.

If the plastic filter is intact, you only have to disassemble and clean 
it carefully, and that's it.
(That has to be done on regular basis anyhow, so that Ti <= (Ta + 15°C) 
at any time).

Service Note 18-A,  formally applies to a certain batch of agilent 
instruments only, S/N US28031400 - US28032927, as it is indicated on 
that note. On that batch, the LTZ1000 was unstable, and the A/D 
converter caused short-term out-of-spec.

We have discussed that, and it's not clear, whether the first issue 
might also apply to all LTZ references in general, if they are not 
powered continuously.. Remember that the LTZ1000 @ 95°C is at least 8 
times less stable than at the regular 65°C, and it is hysteretic at that 
elevated temperature, so the 1yr. stability of 8ppm or even 4ppm can 
only be guaranteed by pre-selection of the reference module.

But I strongly assume, that this old HP instrument is not affected at 
all by SN 18A.

After that, the instrument will be as good as new, best case for 70$ in 
total.

Frank