Understanding the 34470A LTZ1000A Reference

[new299]

Hi,

I've been trying to better understand the LTZ1000 reference in the 34470A from the images of the board Dave posted. In particular, I'm wondering if it's possible to attach a 3458A A9 reference to this header via an adapter. As the same header is present on the 34465A (but partly unpopulated) I'm thinking this might be a nice upgrade for this instrument.

However, while the design of the 34470A reference board follows the datasheet quite closely, the header appears to be quite different from the 3458A A9 reference.

In particular, the following two connections are present on the 34470A header, but not the on 3458A:

A connection to the 2n3904 emitter.
A connection to pin 4 of the LTZ1000.

Does anyone have any intuition as to what these might be used for? I'm guess they might be part of the ACAL system?

I've attached my notes on the header. Some of the pins are occluded so it's possible I'm missing a AGND connection. Looking over the PCB, I also think I've traced pin 1 incorrectly (which doesn't make sense as shown in my notes) but the trace is hidden under a label so I can't see where it goes exactly.

[TiN]

These could be used for monitoring LTZ zener current and LTZ oven heater operation. Perhaps for some 2nd order compensation, or just simply for self-test procedures in 34470A.
You don't need these connections if you just want to replace reference in 34465A. But I'd suggest to feed heater and signal section separately (meaning heater 3904 collector and heater diode connect by own +/- wires to +12...+15VDC at the clean linear supply in 34465A, and signal power and ground connect by own set of wires. Then output sense from pin 3 (+) and pin 7 (-) 7V reference connect to former LM399 zener voltage. Make sure to disconnect LM399 current drive circuit from these pins, so you don't inject current into the LTZ.

Perhaps you will have clear understanding if you instead reverse engineer all connections to those pins on your 34465A board. Just use fast DMM with buzzer to trace all connections out.

I never saw/touch 3446x/7x myself, so can't help you on this. I did similar modify on HP 3245A few years ago, replacing LM399 with my own LTZ KX module, and it's happily working ever since.

[new299]

Thanks TiN, I've spent a lot of time reviewing your site. I've built a KX reference too (though I'm waiting on the Z-foil resistors to complete it, comes up with normal Vishay foils too).

I will start my investigation by probing out a 34465A when I have access to one (I should have one I don't care about the cal on, for this project). I guess there must be switching to switch in/out the LTZ1000 board sense pins too. It would be nice to better understand what's going on there and I will try and investigate.