comparison: content/opentally.rst
content/opentally.rst
- changeset 125
- 669eb33e410b
- child 128
- aba381b2bac9
equal
deleted
inserted
replaced
| 124:ed357d9b4845 | 125:669eb33e410b |
|---|---|
| 1 ================ | |
| 2 OpenTallyLight | |
| 3 ================ | |
| 4 | |
| 5 :Author: David Douard | |
| 6 :Category: Electronics | |
| 7 :Tags: ITC100, Tally Ligth, Arduino, Firmata | |
| 8 | |
| 9 | |
| 10 A month ago, a friend of mine called me to ask for help with a technical device | |
| 11 that seemed not to work properly. | |
| 12 | |
| 13 He was testing with a friend some video recording setup. They use vMix_ to mix | |
| 14 (live) video streams from 4 (or more) vide cameras (Panasonic GH5 if I'm not | |
| 15 mistaken). They also have a DataVideo ITC100_ intercom system. Now the problem | |
| 16 was: how to control tally lights via the ITC100_ from vMix_? | |
| 17 | |
| 18 Surprisingly there are not so many affordable solutions on the market. | |
| 19 Eventually, they found one device, made by a small US company, that would fit | |
| 20 the needs. It's the setup of this very device that was not working properly | |
| 21 when they called me. | |
| 22 | |
| 23 TL;DR; | |
| 24 ====== | |
| 25 | |
| 26 Go to the `project's gitlab page`_ if you want the kicad project files. | |
| 27 | |
| 28 _`project's gitlab page`: https://framagit.org/douardda/opentallylight | |
| 29 | |
| 30 | |
| 31 The non working tally light controller | |
| 32 ====================================== | |
| 33 | |
| 34 The device itself is a small-factor plastic box with a USB B connector, a LED | |
| 35 indocator, plus a pair of HD15 connectors (for the connection with the | |
| 36 ITC100_). | |
| 37 | |
| 38 The first unsual fact was the LED indicator was not on. But that not the only | |
| 39 bizarre stuf: | |
| 40 | |
| 41 - the USB device was sometimes identified by the PC, sometimes not (and no it | |
| 42 was not a bad contact), | |
| 43 | |
| 44 - when plugged to the ITC100, some tally lights immediately lighed up but quite | |
| 45 dimly, | |
| 46 | |
| 47 - and no attempt to change the current and preview streams did not change | |
| 48 anything on the tally light side. | |
| 49 | |
| 50 So I grabbed the device home, opened it up and tried to identify a possible | |
| 51 problem. At that time, we knew the PCB had a problem (because of the LED | |
| 52 staying off), and the seller sent a new PCB (without the case to save a bit on | |
| 53 the shipment), so I was not too worried of hacking the dead PCB. | |
| 54 | |
| 55 I discovered a short circuit between the pads of the LED, and this was | |
| 56 disturbing the 5V USB supply. In fact, depending on the hability for the USB | |
| 57 ports used on the PC to deliver enought current to actually power the | |
| 58 microcontroller or not, the device was seen by the PC or not. | |
| 59 | |
| 60 I though this was the problem, so I just removed the LED and it's limiting | |
| 61 current resistor: no more over current on the 5V rail, the device was now | |
| 62 properly recongnized by the PC every time weplugged it in. | |
| 63 | |
| 64 I also tried to understand a bit the circuit which was in fact dead simple: a | |
| 65 PIC18F4553 microcontroller is presented as a HID device to the PC and control | |
| 66 directly (via a resistor) the pins of the HD15 connectors. | |
| 67 | |
| 68 So I tried to 'reverse engineer' the "control" protocol and since it was a | |
| 69 simple HID device, I just wrote a small Python script to try all the possible | |
| 70 addresses on the first 2 bytes, using pyhidapi_. And I quickly discovered how | |
| 71 to control individual pins connected controlling the tally lights (via the | |
| 72 ITC100_). On my Linux machine with a simple LED connected directed to the HD15 | |
| 73 connector, it was looking like it was working. | |
| 74 | |
| 75 .. _pyhidapi: https://github.com/apmorton/pyhidapi | |
| 76 | |
| 77 But once plugged on the ITC100_, it was still not working. Same erratic | |
| 78 behavior, same total absence of actual control of the tally lights. | |
| 79 | |
| 80 This time I had a multimeter with me, so I took measurments on the ITC100_ | |
| 81 side: each tally light signal was at 12V! and had to be pulled down to ground | |
| 82 to light the corresponding tally light on. | |
| 83 | |
| 84 Since the device was attempting to control directly these signals from the PIC, | |
| 85 no way this could have ever worked. And this was sold something like 650US$ | |
| 86 plus shipment... | |
| 87 | |
| 88 A first prototype | |
| 89 ================= | |
| 90 | |
| 91 So I explained to my friends there was no possible way of making this device | |
| 92 work as intended: it was just not designed correctly. Meanwhile, one of them | |
| 93 told me that vMix also have support for controlling tally light via | |
| 94 Arduino-based setups using the Firmata_ protocol. | |
| 95 | |
| 96 .. _Firmata: https://firmata.org | |
| 97 | |
| 98 Since the fiasco with the "professional" device they bouight, they were really | |
| 99 on the urgent need for a working solution, like for the day after, so I spent | |
| 100 the evening building a very rough prototype. | |
| 101 | |
| 102 I wanted to provide them a better solution, besides a working one. Like with | |
| 103 proper galvanic isolation between the microcontroller and the ITC100_ part of | |
| 104 the circuit. | |
| 105 | |
| 106 Since I had no time nor enough optocouplers in my stash of junk parts, so I | |
| 107 just used regular non isolated 2N2222 to control the ITC100_ tally light lines. | |
| 108 | |
| 109 .. image:: {static}images/opentally/proto.jpg | |
| 110 :alt: A very rough protoype of OpenTallyLight | |
| 111 | |
| 112 It's a very crude protoype, but i worked! Believe it or not, but they have been | |
| 113 able to record several events with this board! | |
| 114 | |
| 115 A (slightly) better protoype | |
| 116 ============================ | |
| 117 | |
| 118 Since we validated the Arduino and Firmata based approach, I proposed them to | |
| 119 design and build a better device. | |
| 120 | |
| 121 At first, I kept the idea of using an existion arduino compatible board and | |
| 122 only design a daughter board for the optocouplers and the connectors. | |
| 123 | |
| 124 I chose the Sparkfun Pro Micro: it was one of the smallest board with enough | |
| 125 pins to drive the 8 tally lights (each with 2 LEDs, the green for preview | |
| 126 signaling, and the red for signaling the recording in progress). | |
| 127 | |
| 128 So I designed a first PCB (in fact several of them, but meh) using Kicad_. | |
| 129 | |
| 130 In the mean time, I also spent quite some time looking for a nice aluminium | |
| 131 enclosure for the PCB. I finally found a small extruded one for a 80x50mm PCB | |
| 132 on aliexpress. It was perfect, so I ordered 5 of them. | |
| 133 | |
| 134 I also decided to replace the HD15 connectors by RJ45 ones, then assemblng | |
| 135 custom cables from shielded RJ4 Cat6 ones. | |
| 136 | |
| 137 So I ended up with this design: | |
| 138 | |
| 139 .. image:: {static}images/opentally/opentally-pcb-1.jpg | |
| 140 :alt: A slightly better design of OpenTallyLight | |
| 141 | |
| 142 | |
| 143 Looks neat, but... I did not realize I put the RJ45 socket backwards! Doh! | |
| 144 | |
| 145 .. image:: {static}images/opentally/proto2.jpg | |
| 146 :alt: A (not so) better proto of OpenTallyLight | |
| 147 | |
| 148 In order to test is nometheless, I put the Pro Micro on the back side of the | |
| 149 PCB (otherwise I was not able to plug the RJ45 cable in). | |
| 150 | |
| 151 Unfortunately, if this board worked fine on Linux (I mean the Firmata protocol | |
| 152 part, since vMix does not work on Linux), the Firmata protocol on Pro Micro is | |
| 153 not working on Windows. vMix was not able to establish the connection with the | |
| 154 Pro Micro, neither did the Firmata test program available on the Firmata_ web | |
| 155 site. I'm not sure why it did not work on Windows, but I had no time nor the | |
| 156 desire to debug this Windows evil misbehavior. | |
| 157 | |
| 158 I looked in my tray of development boards, and found a few Nucleo ones that are | |
| 159 supported by the Arduino environment and have almost compatible pinout with the | |
| 160 Pro Micro. I had to budge a pair of tracks on the PCB, but I could make it work. | |
| 161 | |
| 162 OpenTallyLight | |
| 163 ============== | |
| 164 | |
| 165 So I designed a new version of the board. This time I stick to a good known | |
| 166 working pair: an ATmega328P + FT232RL based solution (the Pro Micro is based on | |
| 167 the ATmega32U4 which embeds the USB controller). | |
| 168 | |
| 169 I wanted to use a USB B connector rather than a micro USB one (too fragile for | |
| 170 this kind of environment), and since I also wanted the design to stay small | |
| 171 enough to fit in my cute little enclosures, I decided to implement the arduino | |
| 172 on the PCB itself. | |
| 173 | |
| 174 The PCB design looks like: | |
| 175 | |
| 176 .. image:: {static}images/opentally/opentally-v3.jpg | |
| 177 :alt: A better design for OpenTallyLight | |
| 178 | |
| 179 Which gives, in real life: | |
| 180 | |
| 181 .. image:: {static}images/opentally/front.jpg | |
| 182 :alt: Front side of the PCB | |
| 183 | |
| 184 .. image:: {static}images/opentally/back.jpg | |
| 185 :alt: Back side of the PCB | |
| 186 | |
| 187 And the result once installed in the enclosure: | |
| 188 | |
| 189 .. image:: {static}images/opentally/complete-unit.jpg | |
| 190 :alt: The completed OpenTallyLight | |
| 191 | |
| 192 .. image:: {static}images/opentally/enclosure.jpg | |
| 193 :alt: The completed OpenTallyLight (enclosure front) | |
| 194 | |
| 195 .. image:: {static}images/opentally/enclosure-2.jpg | |
| 196 :alt: The completed OpenTallyLight (enclosure back) | |
| 197 | |
| 198 | |
| 199 The good news is it worked immediately. The bad news is I forgot to put an ICSP | |
| 200 connector on this PCB. So I had to burn an Arduino bootloader on the ATmega | |
| 201 chip before I soldered it on the PCB. | |
| 202 | |
| 203 | |
| 204 | |
| 205 | |
| 206 | |
| 207 | |
| 208 | |
| 209 | |
| 210 | |
| 211 | |
| 212 .. _vMix: https://www.vmix.com/ | |
| 213 .. _ITC100: https://www.datavideo.com/eu/product/ITC-100 | |
| 214 .. _kicad: https://kicad.org |
