Ahh the joys of prototyping tiny little components when your eyes are getting old!

And after my bad soldering effort!

You may also have heard we also had a bit of a shake just after my last post with a 7.8 earthquake north of Kaikoura about 80Km away, It wrecked the lab and workshop which was at peak mess at the time toppling most of my test equipment and monitors on to the floor and spewing thousands of 0805 components around like snow. My old Owon scope cracked its screen but otherwise no physical damage to anything or my house. The joys of living 2M above and 200M away from the Pacific Ocean are easily forgotten when your shaken awake at midnight and have to get to high ground in your underpants!

I am also doing a re-design of the Data Collection Terminal for another possible client. WiFi is proving to be a pain for them on their large sites and the new design will use either the RFM69 or RFM95(lora) 868Mhz modules (which have compatible footprints and are interchangeable) which should allow a single or at most 2 base units to cover the site and improve battery life. Due to the design of the mobile unit the data rate is very low and so it fits quite well with these modules where low data rates = low bandwidths = quite long ranges. I have not yet put the designs up as they are changing frequently as I experiment with different configurations, the mobile enclosure will remain however I will remove the USB and LiPo battery for 2 x NMHi AA ones which should last several weeks (without a proper enclosure the USB connector had to be cut-out and looked unprofessional) I have also changed the Keypad layout to a standard 4×4 matrix (Nicer IMHO). Oh and we are back to a 125Khz RFID but with a $2 EM4100 module (less than half the price of the chip I was using)!. So basically apart from the enclosure, barcode reader and LCD it is completely different! I was going to use a ATMega328PB to keep my arduino friends happy but the PIC18F26K40 just came out and is twice the Flash and RAM and cheaper.. decisions.. decisions..

The bare bones of the MKII Data terminal (Warts and All!)

The new keyboard overlay was far superior to the original one in every way and came to  US$610 for 100 pieces including courier to NZ which is half the price of my other supplier so shout out to Haiwen Membrane Switch Limited who i will definitely use again

Of course the usual suspects want changes to their respective systems and so everything has to be prioritised, with only one of me and vintage looming this is proving troublesome.

Finally My daughter Alex has turned 21 (today! Happy Birthday) and finally moved on to another full time job in AON in corporate Auckland, I’m going to miss her help, my wife has offered to help by doing my accounts and filing which should free up some time for me and fill the gap a bit

The station has a printer and 2 RFID readers which are activated when a packing crate with an RFID tag embedded in it comes into contact with its cradle. As we have two positions we also need to know which station the reading was sent from. I also need to send a command to a box printer to change the lot number and product name printer on the boxes being packed.

If the station was close I would normally use an RS232 reader on different COM ports. (I cannot use USB as I don’t easily know which RFID port is which), however RS232 is really only good for 10M at 9600 baud especially in industrial areas where there is a lot of noise so I needed a way of doing this.

Our solution to this is below (link to project is here)

Two serial RFID readers, are connected to the circuit board which provides power to them and also reads whatever they send. On receival of an RFID tag code (a 10 char string with a header and trailer char) the system appends an “A” or a “B” to the serial string, removes the header and trailer and replaces the trailer with a CR LF and sends it to the PC over RS485 which is good for very long distances and speeds through noisy environments.

Data sent from the PC is passed though to the printer serial port and on to the label printer. however by embedding a command sequence at the start of the string we can also turn on or off the 4 relays, sound an alarm or request a re-transmission of the RFID tag numbers.

By having it on a micro controller we can reduce the complexity of the PC code and achieve a better reliability as well as achieving our goal of extending the distance.

The prototype is shown below (warts and all), the wiring changes were put in so that the RFID signals can trigger an external interrupt which improves the reliability of the soft modem. the project EDA files have all the modifications in for the second version

The Current MPLABX XC8 code for the PIC is here => main.c

Note the code is work in progress so check back for updates, most of it works but its not been tested under pressure, however it should give a feel for what the system does.

The only change I would make in a future version would be to use a switching regulator like an LM2596 which requires an additional 100uH inductor and a diode but then gives proper industrial power supply tolerance. My circuit is running off 9v as it supplies the RFID’s their power so the LDO gets warm when all the relays are engaged and it pulls 250mA peak, however I am only using 1 relay here so max current is <100mA and there is no reason for me to fix this.