18th May 2026 - Screens Again
The recommended screen to replace the generic Qpas 1602A model was the Waveshare LCD1602 I2C Module. In the search for suppliers and for information about the unit, a review on AliExpress was turned up. The author waxed lyrical about how good the product was, how robust in use and how easy to use compared to the generic models. That sounded encouraging but one review is seldom enough to get a good idea about a product. The main selling point for this project was that the unit did not need a Backpack to drive it as all the logic was ‘onboard’. This was confirmed when the ordered units arrived as there were many more chips visible on the reverse.
 |
Physically the new screens were on backing boards slightly longer and slightly narrower than the Qpas model. The screen available for easy sale were blue background with white lettering compared to yellow background and green lettering for Qpas units. The cabling was also just four wires but this time there was a socket and a plug with four rats’ tails. It was decided that the colours of the Rats’ tails should be copied from the Wago connection to the Pi breakout pins to make life easier in the future. Fortunately, the jumper collection contained sufficient wires of the right colours and correct connections.
 |
With that done it did not take long to wire the first screen to the Pi. When the Pi was switched on the screen lit up with a bright blue light in a similar sort of manner to the Qpas units. The sudo i2cdetect -y 1 command was run again and the message came back that the I2C address was 0x3e. A note of that was made and the second screen was similarly tested. That too was blessed with the I2C address of 0x3e which was not a surprise. The back of the screen was then carefully examined with the magnifying glass to find where the solder point was to change the address. To some frustration no sign of a solder point for adjusting the address was found. This was relayed to Deep Seek and after a short while it was confirmed that changing the address was not possible and every screen of this type would have the address 0x3e.
At this stage the screen had yet to be persuaded to actually display anything. Deep Seek provided a number of scripts to run that were meant to activate the screen. None worked and eventually a Waveshare demo script was found and tested.
 |
That did work and it was confirmed that both screens were operative. However, the problem of them both having the same fixed I2C address remained. I was rather cross that the glowing reference on AliExpress had never mention that problem.
The ever soothing Deep Seek offered the TCA9548A Multiplexer as a solution, but that meant adding another board. With a possible attempt at teasing (AI interfaces are improving!) Deep Seek suggested that the Miltiplexer would allow the addition of a 3rd or 4th screen if required with no need for extra support boards. Thus the internet was scoured for the suggested Multiplexer and a small variety was discovered. Deep Seek suggetsed that one of the AdaFruit boards would be a 'best buy' in that it would work, it would be reliable and only cost slightly more than the anonymous generic article. An order was placed and the board arrive in next to no time, unlike the free time to deal with it which was very difficult to acquire.