The Last Parts

I completed the sensor connection PCB this weekend. I need to cover it with heatshrink to protect it but I need to test the full system before I can do that. Here it is connected to the gun complete with curly cable leading to the sensors:



Here is the complete assembly:



To test the system I need to construct another Milestag PCB. As mentioned before I bought 4 of these so I used my PCB jig to help me populate the boards:



This device holds multiple PCBs, allowing me to place components. When populated, a foam covered board is placed over the components and clamped in place. The jig then rotates to allow the components to be soldered whilst the foam prevents them from falling out:



Here are the completed boards:


I intend wire IR LEDs and trigger switches to the boards so I do a quick test. Before I can do anything with the remaining boards though I need some more gun housings, and for that I need my own RepRap......

Completed Sensors

Today I attached the two sensor housings to the new RJ10 PCB I made. Here is a detail of the junction PCB showing the two sensor leads attached with tie wrap strain relief:



Here is the assembly showing the two sensors attached:



I decided to attach the sensors to clothing with velcro. This allows me to flexibly configure the sensor positioning as required. I bought some 50mm velcro and stamped holes in it to allow me access to the sensor screws.



I've also designed a small PCB to hang from the gun to allow me to connect a RJ10 curly lead to the sensors:



I just need to make this and the gun is complete.

Sensor Connections

I've been trying to find the best way to mount the Milestag sensors to a headband and connect them to the gun.

I've decided that I'm going to mount a loop strip of velcro on a baseball cap or elastic headband and attach hook velcro to the back of the sensors. This gives me the most flexibility for sensor placement.

I've been unhappy with the way the sensors connect. They are traditionally soldered in parallel with a connection to the gun. I had concerns about the reliability of this as soldered connections break when flexed.

I decided to compromise and designed a PCB to connect two sensors to a RJ10 socket. Why RJ10 when others are using RJ11? Because RJ10 curly leads are available for less than 1UKP and these are ideal to connect the sensors to the gun.

Here is the headband sensor schematic:




It's very simple. I intend to place it in a small, velcro mounted box.


Here is a set of 8 PCBs ready for cutting and drilling:

Finishing the sensors

Today I drilled my sensor PCBs. There were only two hole sizes, 0.8mm for the TSOP IR detectors and cable holes and 0.6mm for vias.

I use high speed tungsten drill bits with this 10000rpm PCB drill I blagged in a broken state from my previous employer:



I just populated two boards, which is sufficient to complete one gun. Here are the boards, one enclosed in a transparent sensor dome:



The next step is to mount these on a headband and connect them to the gun.

Milestag sensor PCBs

As I've mentioned, I was fortunate to be in the USA last year and was able to get an American friend to order Milestag RevH PCBs for me, which saved me about 35UKP on postage. However, Jim was out of stock of sensor PCBs at the time. So I still needed to source these.

I decided to make my own. First I used Kicad to implement the simple schematic published on the Milestag site:



Then I designed a PCB around it. As I intend to make this myself, I changed the diameter to 31mm as this is the internal diameter of the 35mm holesaw that I have, and I replaced the two through hole resistors with surface mount ones to give me some more real estate to work with. Here is the design:



I repeated this 8 times onto an area of 160x100mm which is a standard dual layer size supplied by Maplin.

I then printed the top and bottom layers onto tracing paper with my laser printer:



The top layer is cut and overlayed onto the bottom layer. I align them using a lightbox to make sure the alignment is as accurate as possible:




When aligned, I fastened them together to form a pocket using cellotape and inserted the pre-sensitised PCB, fastening it with tape to prevent it from moving:



I then exposed the PCB to UV light from a single sided UV source, turning the board to expose the other side.

I develop the PCB with commercial developer, in the past I have used 99% sodium hydroxide crystals but this is nasty stuff. I got this from Mega Electronics. They do commercial grade PCB stuff but are willing to deal with individuals:



I etch my PCBs in a home made tank:

It's tuppaware with tropical fish tank heater and air pump with a wooden surround. Basic, but it works. God help me if it ever leaks though:



Here is the etched PCB:



I then cut each unit out with the holecutter:



After cleaning with wire wool, I used this stuff from Mega Electronics:



To tin plate the copper tracks to prevent oxidisation.

Here is the finished sensor PCB:

Full assembly

I experimented with IR focus and the results were interesting. The circuit published in my previous post could detect IR from the gun at about 50m without a lens to focus it. My plan was to get as far away from the test circuit as I could until it failed to detect a transmission, then adjust the focus until the sensor saw the IR pulse again, then move further away.

Unfortunately, I was unable to get far enough away from the sensor in my test environment! I resorted to the IR camera and got a satisfactory beam focus by visually examining the size of the beam on my monitor.

I haven't built the IR sensors yet, but I decided to build the milestag gun up fully. It went together well. I also fitted a picatinny rail to the top of the gun and fitted a cheap red dot sight I got from ebay.

It looks pretty good. Trouble is, I now need to paint 60% of it bright orange to conform with UK imitation gun laws. Ah well.....

Focus, focus, focus

I've been trying to optimise the focus of the Milestag IR beam today. At first I thought the best way would be to observe the beam, and try to get the brightest dot at the furthest distance.

To achieve this I created this unholy device:



This is a super sensitive IR camera with a battery powered CCTV alignment monitor + the Milestag IR focusing section of my gun with faux reload and trigger switches.

My idea was to see how far I could get the beam to be a minimum diameter.

It worked very well, except that at the distances I want I could hardly detect the IR light on the small monitor.

Also I looked a bit of a cock as I walked about my street in darkness with this thing.

So I sat back and thought 'what am I trying to achieve?'. Answer: IR reception at maximum distance so I came up with this:



It's the TSOP4856 detector the Milestag system uses connected via a transistor to a LED. When it detects a 57kHz IR signal the LED illuminates. With the aid of a helper I should be able to optimise the IR focus. The further away the LED lights, the better the focus.

Here is the test circuit built up: