Some 732a notes - updated with new measurements and fix for sensitive pot

[martinr33]

I am working on a 732a. This unit was apparently used at a Fluke lab in Schaumburg, but has not been calibrated since 1987. I wonder if maybe it failed.
Anyhow, the charge circuit had a shorted transistor, which I replaced. After a lot of messing around, I found that the high current charge threshhold voltage pot is incredibly sensitive. The circuit works as expected, but only on a tiny, tiny range of the pot that is hard to set.

This nonexistent range does not seem right, so I have to figure that out or button it up and see how it goes.

The circuit is stuffed with constant current diodes. These are Siliconix parts, but with Fluke house numbers. I measured a few of these and annotated the schematic. It is impossible to understand the operation of the circuit without these values. Replacements from Mouser are $20 each, but I found some at the local junk stores (none of these were bad, it was down to that supersensitive pot). I have measured some of the units from  the board and recorded their values on my schematic.

I think that these constant current diodes are in the design for thermal stability.

This unit has a number of undocumented mods. In particular, the float voltage is set by 26.4V of zener diodes and a lower value pot. This is presumably more precise than trying to use a trimpot and resistors. 

Working on it is tricky, as the board has to be in the frame to test the circuits. Next step, figure out how to broaden the operating range on that pot.

Link to picture of the board
https://photos.app.goo.gl/AL5Y661G8UcJbCQs5

Link to the album:
https://photos.app.goo.gl/BTQYcmHNpSp5CQTi7

I have attached an annotated schematic.

One thing  to check on these units is that the 18.6 V rail is set correctly. This is the pot on the voltage regulator board. It may affect the oven temperature, which will cause the voltage to drift.

UPDATE: I have it working now. Of all things, the 500 ohm pot was measuring 700 ohms, and had a bad spot - right at the set point - where it added 1k of series resistance into the wiper

Adjustment seems finicky - the charge point and the float point seem to have a little interaction, which I didn't expect.

Anyhow:

Charge current 0.33A
Float current 14mA
Charge trip voltage 29.7V (29.8V at the connector. I have moved this down a couple of volts - 32V is too high for the batteries I have)
Float voltage 27.4V (27.3V at rear power connector)

These voltages are measured at the battery. The connector voltages, with a series diode and parallel resistor, measure 0.1V lower.

And one more update.
I had a misunderstanding about the operation of the circuit. I had assumed that it would not enter fast charge mode if the battery was already at its float voltage. However, the circuit appears to always start up in fast charge mode, and then drop to trickle mode once the threshhold voltage is hit. In other words, it does not care about the initial state of the battery. This approach changes my dim understanding of the circuit somewhat. Those constant current diodes are tricky.