DEC PDT-11/150 – Part 4. Self tests and the drive controller

I have been running some tests and learning a lot about this little machine and its ways.

It has two layers of self-test. The first is a mandatory Power On Self Test (POST). The second layer is an optional interactive test made available by setting switch 4 on the main DIP switch block and activated by the toggle on the back panel (replaced in part 1)

In my second post I couldn’t get to the optional test no matter how I set switch 4 inside and the toggle switch on the back. Instead I was getting an octal number and being dropped into the ODT.

I naively thought that this was some kind of RAM value. It wasn’t. This was in fact an error code from the POST. The value I was given, 170732 means there is a RAM fault on the main board. I have spent some time re-seating all of the socketed chips and I think that’s what fixed this.

So the reason I couldn’t get to the optional test suite was that the POST was failing.

Past the POST.

Now that the basic POST is passing, I can get beyond that to the attempts to boot I showed in the last post.

At this point it doesn’t make any obvious attempt to access the disk drives. I would expect a little graunching to find track 0 or something similar but there is nothing.

Test mode.

I can now try setting the DIP switch and toggle to enable the test mode. With this done and a reboot I get nothing on the terminal at all. Flick the toggle for normal mode and it’s back to the boot prompt. Set the toggle to test and nothing.

The PDT-11 has a number of status LEDS on both the main board and the disk controller. Setting for test mode again, opening up the lid and separating the boards shows us a number of clues.

The main board switches are showing 01111 which means “1 Disk controller module” or “2 Cable G1”. The cable looks fine so it’s the disk controller module.

A look across to the disk controller and its lights are showing 0011 which means  that the LSB RAM IC is faulty. If that’s true, switching the two RAM chips over should move the problem to the MSB.

A quick switch later and the error code on the LEDs has changed to 0010 which means the MSB is now faulty.

These chips are 4bit x 256 bits and some new ones are on order. 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *