If it is an AC error, that is normally caused by bad power supplies and/or relays. (A relay with high contact resistance can act like an antenna and pick up noise generated by the power supplies which combine to cause the AC self test error.)
(Do you know that if the LOCAL/REMOTE GUARD switch is in REMOTE that you'll get an error 7?)
Check the power supply capacitors, especially C6 on the REAR PCB, and C45 to C50 on the ANALOG PCB. (I use an ESI 253 for these measurements, except for C7 on the REAR PCB, in which case I use a GenRad 1689M.)
How I test the relays depends on what options are fitted. If it has FRONT/REAR terminals, place the REMOTE/LOCAL GUARD switch in REMOTE and 2-WIRE/4-WIRE switch in 4-WIRE, power on the instrument, set the FRONT/REAR switch to FRONT, and measure from the front terminals to a connector that attaches to the REAR PCB. In the top left rear, there is a 4 pin connector that connects I-, Lo, V-Guard, and ?-Guard to the REAR PCB. (I think I use a 22 AWG pin to banana adapter to insert into the connector after removing it from the REAR PCB.) On the bottom side, there is a 2 pin connector that connects Hi and I+ to the REAR PCB. I check all these for < 1 ?. I also check continuity from the rear terminals, but the self tests are from the front terminals. This checks the FRONT/REAR relays.
For the (input) relays on the AC PCB, I pull the AC PCB input connector and check from the connector to points on the PCB (using the schematic and with the instrument powered on so that I can switch the relays). To check all the contacts, both AC and AC+DC modes need to be tested (some tests go through both input relays). I suppose that the relays could be tested on the relay pins, but I usually leave the insulation attached unless I have to burnish a relay.