Cricket Scoreboard Repair

Well, it’s game related and has electronics in it. The story shall be told in chronological order to pad it out…

Evening One

Whilst watching my kids running around at the cricket club, I find out the scoreboard is broken – it’s been like that for a few weeks. “Blown fuse” apparently. I volunteer to have a look…

Evening Two

It’s not a fuse. Lots of fuses did pop, but they had a scoreboard specialist company out, who had two guys spend five hours looking at it. The processors are dead… and encrypted, apparently. I take a brief look anyway. There’s a control board, which is alive and three ‘slave boards’ which each drive a load of segments. The one observation I can make is that there’s no processor in one of the slave boards – it later transpires it was taken away to see if it could be read. I’m told they tested it and the control signals were making it to the slave boards.

The system is by Digitext Display Systems. It’s from 2004, so not that old. Company is no longer around though.

I walk away to think about it. Bit daunted now. Some wine later and I vow to sort it, one way or another…

Evening Three

The missing processor has been returned. With it placed back, I can see an LED flashing on each slave unit when buttons are pressed on the console downstairs. The lights come from the processors, so they can’t be that ‘dead’. I take one away to read (87C51). It is encrypted, but there’s a signature there. So, again, maybe not so ‘dead’.

At this stage, I wonder if the slaves are all alive and they’re just getting duff signals.

The next stage is to get an oscilloscope on it. Only problem is there’s no power in the score room and it’s at the other end of a cricket pitch to the nearest mains outlet (and I’m not putting my precious scope onto a generator).

Before I go, I take a look inside the control panel. There’s an exploded capacitor. No-one’s mentioned that before.

Evening Four

I see the note from the specialists who declared the unit unrepairable:

Reading the letter, I feel a bit disheartened. They’ve clearly had a scope on it and have observed data pulses. They’ve also gone right down to the nitty-gritty of the components and found it broken. The guy’s also put letters after his name, so he’s ejucated.

I try the console now I’ve changed the cap. No change. Not really that surprised – it wasn’t a ‘smoking gun’. I did think it was a bit crazy having 25V rated caps and a 7805 on a 24V battery, but it seems it’s OK.

I then find out a bit more… the system stopped working, so someone had a look and accidentally wired the 24V battery backwards. That explains the blown capacitor.

Anyway, now armed with 5 extension cords and a 100m reel of twin-and-earth 1.5mm mains cable, I get power to my scope.

All the slaves are getting valid-looking RS232 signals at 1200bps. It’s fair to say something is likely up with the slaves. The circuit seems one-way only. Slaves only receive from the master and each one has a node, set by jumpers.

One thing I don’t get on the letter is the bit about shorting out the transistors to make it work. It just doesn’t seem right – shorting collector to emitter should do something. I used the diode setting on the multimeter to check stuff – always good to check for dead bits by comparing readings for similar pins/components. All the processor pins read consistent – inputs, outputs, the lot – nothing is shorted to rails or open-circuit. What isn’t right is the transistors on the processor outputs (the ones they’d been shorting out). Almost every transistor on every board (36x) is dead – collector shorted to the base. One even has a smashed case. Thankfully there’s 1k resistors between the processor pins and the transistors (the transistors are visible on the 87C51 photo above – TO92 devices).

Evening Five

With 36 Darlington transistors replaced, I try again.

The numbers clack to new values when commanded.

Success, at last!

I will claim my 2 pints of beer reward tomorrow. Shame they had to pay over £500 for the previous non-repair, but at least I saved them 10x that on a new scoreboard.

John Bennett

Not Phil