CAPACITORS ARE THE BANE OF MY EXISTENCE

I discoved after the previous failure that I had blown the fuse on the regulator board so I replaced it with a spare. It also seems like a lot of issues with electronics are related to polarised capacitors going bad and I wasn't going to take any further chances here. Eventually the two large cans for the regulator board arrived. Nichicon - because I'm worth it!

I installed them and a bunch of others while I was at it. The pair of radial 1000uF electrolytics had actually leaked brown goo so I'm glad I didn't have it powered up too long initially.

Once the regulator board was repaired, I put it back into the chassis and powred it up without cards again using a car lamp as a load. All was good when I probed the voltages so I ran it for half an hour or so. I nervously reinserted the CPU, Triple Option, and a core memory board back into the chassis and powered it up again. I was relieved that it was all behaving as normal again and that nothing had broken. I then put in the Cassette I/O board in again, powered it up, and it started misbehaving again like last time. I removed all the cards and checked the regulator board only to discover that the fuse had blown again. This was causing the weird behaviour. I reseated the regulator minus the rest of the cards and started probing voltages. The VINH +15V line was gone. With this information, I buzzed out the Cassette I/O board only to discover it had a short between VINH and ground. That board was to blame. I suspected a shorted axial tantalum capacitor so I removed it, probed again, and the short was gone. I had a Nichicon axial electrolytic with a similar value and soldered it in where the dodgy one was.

Righto! I replaced the fuse, slid the Cassette I/O card back in, and powered up the machine. Everything was back to normal. I hooked up a serial terminal and toggled in a simple loopback test that spits out whatever is read from the keyboard to the monitor. I started the program and randomly bashed away on the keyboard and all was good.

The next step was to load in a program from a paper tape dump. This is a two step process - firstly load the Binary Loader as the built in loader is limited, and then load the program binary you actually want to run. I chose to try out Single User Basic as it was small and looked like fun. Initial attempts to do this failed as I was using the "Binary Loader" tape and then I read that it was the wrong format for the built-in Program Load option. I then switched to using the "Self-Loading Bootstrap Loader" as that was the required tape. It loaded in a split second and the CPU halted. I hit Continue on the front console and sent the Basic tape from the terminal program. And then...

As you can imagine, my excitement level rose to the point that I started cheering. This was all well and good but I needed something better - Extended Basic. It was a much bigger program, in fact the largest paper tape program I had, so it was a perfect contender to test the system. Plus it would test the reliability of the serial link considering it was set to 9600 which was as high as it would go. So...

It worked perfectly! So now I have a working Nova 3 with MUL/DIV, MMPU, serial console, RTC, and 32k words of core. I have another two 16k word core boards that need looking into and fixing so maybe I should tackle those next.