Martinstown Cricket Club

By the end of their project Martinstown had a fantastic scoreboard, however their project did have some problems and there are some important lesions to learn.

Things that Martinstown CC did well:

  1. They didn’t rush the project; they waited until the off season to start the build
  2. They put together a committed team, who were able to put time in to the build and brought different skills
  3. They had a focus on quality, and doing things right.
  4. They persevered when there were problems

Martinstown first stumbled across the buildyourownscorboard website in May 2016, but decided to wait to October to start their build in earnest.  A build team of Matthew R, Tony J and Jim G formed and began the build.  Jim and Tony focused on building the board, circuits and wiring.  Matthew set up the Arduino and the Raspberry.  Like Ian B and myself, the Martinstown build team had no prior electronics experience, and yet were able to build the most fantastic board!  If you have their patients and methodical approach, you too can build an amazing electronic board.

As you can see Martinstown chose a modular design for their digits layouts, taking inspiration from the Stalybridge St Paul and Potton Town builds.  I love the neat layout and use of connector blocks, which makes re-wiring and troubleshooting easier that the jelly crimp that we used on our build.   That said, the one problem the Martinstown did have was with a loose connection that took a few hours to find and caused their first set of numbers to show 888!  This was solved by calling in Paul H (head groundsman and ex BT field engineer) who found the bad connection and fixed it after about 5 minutes.

dsc_0031

Martinstown mounted all of their electronics to a single board, which again is a module on the bigger board.  You will see that all of the cables are neatly labelled and colour coded which makes debugging very very easy.   You will also notice that they are using an IP67 power supply designed for LEDs. Despite problems for Stuart at Staylbridge, Tony has found a water proof power supply that works for their board perfectly.  You can see the exact version on ebay here, if you want a similar device.

dsc_0216

Here you can see the front of the board, which has 4 coats of black paint on it.  It looks amazing.  Martinstown also built a cover which is put over the front of the board when it isn’t in use, which also doubles as an analogue scoreboard in the event of problems.

dsc_0221

Here you can see the finished article from the back.  Again, I am blown away by the time taken to make the build neat and tidy.  Well played!

Here you can see the finished board.  The lamps look pretty powerful to me, and the lenses make the score very well defined.  All in all Tony says that Martinstown spent 1 morning a week for around ten weeks to build their board.  All in all, they spent around £800 building their board, but this included all of the timber (including the cover) and a ball proof cover (like the one used on the Stalybridge build).

final

Problems!!!

The board was installed for the start of the season, and worked well for a few weeks before some problems began to occur.  Initially, the board started suffering with some random occasional digit corruptions and then over time got worse and worse until the board stopped working usefully.  Thankfully the manual scoreboard cover came to the rescue, but the club were £800 and a hundred hours down!

To debug the problems we focused on the bottom row by:

1. Isolating each  of the digits on the bottom row, by removing cables from both sides of the screw clamps (top and bottom), as per the bullet instructions below, on every digit.

  • Remove the chained yellow 5 volt cable from every digit on the bottom row.
  • Remove the white serial in cable for every digit on the bottom row.
  • Remove the blue RSK cable for every digit on the bottom row.
  • Remove the green SRCK cable from every digit on the bottom row.

2. Carefully rewired in the Overs Ten digit by reconnecting the cables as per the bullets below (only this first digit).

  • The chained yellow 5 volt cable
  • The white serial in cable
  • The blue RSK cable
  • The green SRCK cable

3. Use the Bradford web interface to enable the test mode.  The first digit on the bottom row should count from off through to nine.  Cancel the test mode once this is working.

4.Wire in the next digit (Overs Ones) by reconnecting the cables as per the bullets.

  • The chained yellow 5 volt cable
  • The white serial in cable
  • The blue RSK cable
  • The green SRCK cable

5. Use the Bradford web interface to enable the test mode. You should see the first and second digit on the bottom row count from off through to nine.  Cancel the test mode once this is working.

6. Repeat steps 4 and 5 for each subsequent digit, until a problem occurs.  When you hit a problem (the display being corrupted!) then stop and check all the connections on the last circuit that was added to the string.  The problem is either at that digit, or the one before.

From this testing, a number of faulty circuits were found, despite looking OK and testing with a multimeter passing.

After a lot of investigation Jim concluded that the problems were associated with the condition of the soldering on the circuit boards.  Having carefully examined the boards that failed there was evidence of a light corrosion on some joints which has compromised the circuit integrity to the point that after warming up they short with joint alongside. This is not a disaster as they seem to be easily cleaned and operate faultlessly when this has been carried out.  In addition there were a number of joints that were brittle.

Jim explained the issues were related to:

When we were gathering materials a club member gave us some solder. It was fluxless. I soon deducted that I needed to add flux, I did this by whetting the joints with a spot of flux using a cocktail stick. This gave me some success in soldering the shifter blocks to the boards but it soon became clear that there was too much flux applied and after a day or so this turned green and looked a mess. I cleaned the offending boards by applying methylated spirits with a cotton bud and this seemed to work, at least the joints were clean.  This caused some bad joints.

I then bought some flux cored solder and continued with that which made things much easier. I found the shifter blocks relatively easy to do as the joints were stable and was able to develop a technique (the more I did the better they got).

The soldering iron that I used was a small electrical one with a 4mm copper bit. I filed the bit to a chisel point about 2mm wide at the end. We found some brittle joints because of me applying the heat for too long on joints .

My technique improved fairly quickly as I went on so the number of brittle joints were small. I must say that the soldering of the wire to the copper was more difficult than the shifter blocks and I was glad to have done all of the shifter blocks first.

I must say that everything that I did I tested out with a meter many times over, so I was confident that the joints were good at that stage but always mindful that it was possible that they could break down due to handling etc.

Finally

The board is now back to full operation due to the dedication of the team at  Martinstown they have a brilliant build.  I know it was a difficult time during the debugging and repair, and I would like to thank the team for sharing their results with the project!