HP/Agilent 6632B Repair

This is the second 6632B, that I bought from eBay in condition "For parts or not working". I repaired it about 2 years ago, but after receiving some questions about the 6632B via email, I thought this article could be helpful to others.

The inside of the unit was very dusty, but after cleaning there was no visible damage on any part. After reconnecting the transformer tabs from 110V to 230V mains, the unit seemed to work almost properly.

There was just one problem in the high Current range. It constantly showed an offset of about -25 mA in the high current range and the low current measurement was only working for currents smaller than 9 mA vs. 20 mA mentioned in the manual. It's obvious that this couldn't be just a calibration problem.

6632b schematic current range — Fig.1: Schematic of the 50m and 72Ohm shunts for HI- and LO current range

After taking a closer look at the circuit in Fig.1 dedicated to short the 72 \(\Omega\) shunt R403, consisting of an MTP2955V NMOS Q313 and a NTP18N06LG PMOS Q314, I discovered that the PMOS failed completely short between Gate-Source and Drain-Source. Normally in the high current Range, the 72 \(\Omega\) Shunt has to be shorted. The current would flow through the body diode of Q313 and the conducting PMOS Q314, whose Gate Voltage would therefore be set to about -11.5V by the MC34074 Quad-opamp U304.

With the shorted PMOS, U304 tries to set the -11.5V, which would lead to a sink current of \({11.5 V \over 200 \Omega }~= 58 mA\) through R499, but the protection circuit of the Opamp actually limits the current to the displayed -25mA. The MC34073 datasheet Fig.2 even shows the implemented current limiting circuit. If the output current gets to high \(I_{out} \cdot R7 > 0.7 V\), the pnp Q18 starts to conduct, which reduces the the Base-Emitter Voltage of the the upper Power Transistor in this totem pole output stage.

mc34074 schematic — Fig.2: MC34074 schematic from https://www.onsemi.com/pub/Collateral/MC34071-D.PDF

In the low current range Q314 should not conduct. The current flows through the body diode of the Q313, and is clamped to 0.7V. This explains the 9 mA max measured current.

\[ I_{max}={U_{SD,nmos} + U_{DS,pmos} \over R_{shunt}} = {0.7 V + 0V \over 72 \Omega} \approx 9 mA\]

After replacing the defective and obsolete MTP2955V with an IRF9Z24N, everything works fine now.

two mosfets — Fig.3: MTP2955V and IRF9Z24N (TO-220)

The rotary encoder was also broken, and replaced by a new ALPS EC16.

Datasheet: keysight.com
Manual: keysight.com
Datasheet MTP2955V: onsemi.com
Datasheet IRF9Z24N: irf.com
Datasheet MC34074: onsemi.com