Welcome to
Azur Electronics

REPAIR HP 1630D LOGIC ANALYSER
(Actually a 1630G!)
June 2007

The HP 1630D Logic Analyser had been damaged in transit and the inside was very dirty, so it was completely stripped down and all the parts thoroughly cleaned up.

Top view of Chassis showing guide rails for 4 plug-in cards and space for CRT on left hand side.

Rear Panel with Fan

Various bits of plastic mounting lugs were found inside the Chassis, these were glued back on to the Front Panel.

A2 Mother Board with connectors (top to bottom):
Connector for Service Maintenance;
Connector to A1 Power Supply Board;
Connector to A11 State Slave Board;
Connector to A5 Timing Master Board;
Connector to A4 State Master Board;
Connector to A3 CPU Board; and
Connector to A13 Analogue Board (1831A-D only).

A7 Key Board

A11 State Slave Board

A5 Timing Master Board

A4 State Master Board

A3 CPU Board with external connections (top to bottom):
Accessory Power BNC Connector
HP-IB Switch Settings
HP-IB Connector
HP-IL Connectors (HP Interface Loop)
Port BNC Connector
Plus 2 internal ribbon cable connections to A7 Key Board and A6 Display Board
 
A6 Display Board and CRT, both made by Hitachi
A1 Power Supply Board

Only 1 fault was found, a high voltage capacitor had exploded in the power supply. C12 is a special 22nF 250Vac decoupling capacitor made by RIFA (missing in this photo). A replacement was obtained and refitted.

This board was modified several times over its life because of reliability issues which are detailed in the Manual.

With the Front Panel repaired, the Logic Analyzer is re-assembled. The construction is quite clever. Behind the Front Panel is the A7 Keyboard Board and then the A2 Mother Board, into which connects from bottom to top: A3 CPU Board; A4 State Master Board; A5 Timing Master Board; A11 State Slave Board.

The A1 Power Supply Board is fitted at the top (as in photo, complete with new capacitor).

Note that the A8 Timing Slave Board is not fitted with this configuration as it is replaced by the A11 State Slave Board. Also the A13 Analogue Board is not fitted, this provides an oscilloscope function for the HP 1631A-D version.

The A6 Display Board is shown fitted to the left hand side panel. This is actually made by Hitachi, as is the CRT, and contains the high voltage power supply and display adjustments.

Just behind the top rail on the right hand side is a 100 way (2 rows of 50 pins) plug. This is part of the A2 Mother Board and allows any of the plug-in CPU, State, or Timing Boards to be connected for maintenance purposes, so that both sides of the pcb are accessible - clever.

Powered up on a Variac ok and "System Specification" screen obtained (green text on black background). With Rear Panel Switch SW6 in position "1" self-tests all pass. Initially some "Acquisition errors" were displayed, this was due to the pod connectors not making a secure connection to the State Master & Slave Boards. A temporary cardboard rear plate was made, this retains all the connectors to the pods. This will be replaced with an aluminium plate if I get round to it, or more likely if the connectors prove to be making an intermittant contact!

June 2009

Made up an aluminium panel (thanks Glen) to retain the plug-in connectors from the Pods.

August 2009

Access to the 4 plug-in Boards is available from the Rear Panel. Remove the Pod connectors retaining panel (2 screws) and the aluminium extrusion (4 screws).

3 screws hold the A3 CPU Board at the Rear Panel and the bottom cover also has to be removed to gain access to the 2 ribbon cables connected to the front of the Board.

According to the Service Manual, this is the current ROM set for the 1630G.

This Logic Analyser has probably had the HP upgrade from 1630D to 1630G which cost $3,400 in 1987.

Note that these are different ROMs from a 1630A or D version.

Therefore it is not possible to swop State Slave and Timing Slave Boards as required to convert from G to D versions.

Spent several days looking for an intermittant fault on the A1 Power Supply Board which was causing it to shut down all output rails, then after a while starting up again. I'm not a fan of switched mode power supplies! They may be more efficient than linear circuits but not as easy to work on and live mains voltages all over the place. At least I know how this particular circuit works now, but far too complicated and needing fault protection circuits which in themselves cause unreliability. Possibly the thermal switch was malfunctioning but after hours of soak testing all seems to be ok for now!