While waiting for the test PCB, I have to find something to do. Our laser cutter is only 30cm large for 610cm long, so I find a way to cut our wooden plank it in three ways. I hope that there will be no need to add something, but if the plank bend, we will have to put some wedge in the middle of the plastic plank.
I also have to unbend the wood. We ship a 3m plank that came wound, so I make something like this
I also cut a plastic plank for our test box
And made a better scheme with CorelDraw for those who want to make the same device
At last, I take a look at Ilan code for the beginning of a thread that will check the state of all the box.
We are still waiting for our test PCB, so on Monday I create a HAL file with STM32CubeMX for the test PCB, merge it with Ilan’s code and I will wait that we receive the PCB to check if the merge was ok or not.
Today I’m in Telecom, so I will try to finish the box (I did not have a plastic plank last time) and to begin the box for the final devices.
On Wednesday, I finished the PCB. I used one trick that found Alexis on our test PCB. We have two VCC, one for the coils and the other for all the other components. So, I create a two-parts plan. The VCC is in green and the VCC_COIL is in yellow.
I also have to route all the components. I found a way to make the routing easier, by enforcing a simple rule. When the route is horizontal, it must be on the top face, and when it is vertical, it must be on the bottom face. I also have to change the pins of the multiplexers in order to route them easily.… Read more
On Monday, I routed the 4×4 PCB I placed on Friday. I choose to put one multiplexer at the top of this cell and another one at the right. So the SDA lines (which come from the STM32, not from the multiplexer) will come from the left and the bottom. I also changed the scheme by removing a pull-up I put between the main multiplexer and the other multiplexers.
I try to use some shortcuts I found such as one that swaps the pins, but after reading the manual, I found that without a change in the part editor, I cannot use it, so I finally change the pins in the scheme
Now, I will copy 16 times this cell and continue routing with the other components
On Friday, I continued to place components. I spent some time to put the 4×4 coils precisely (in the same way as for the test PCB. I found some keyboard shortcuts that help me to do this quickly, but of course, I found the best shortcuts on Monday, when most of the work was done. I take a look at Alexis’ work he made on our test PCB and found some toolbars that will be very useful for me.
Now, I will try to place the multiplexer and to route it. Then I will copy this 4×4 grid sixteen times to finish the PCB until placing and routing the unique components.
On Thursday, I went to Telecom for the first time of the week. I took advantage of this journey to print the prototype of our box for the test PCB. But I lack plastic plank for the middle part (in green in other photos), so we need to wait for a week to have it. However, we will have it on time because we won’t have the PCB before.
In the afternoon, I began the final PCB. We found before a component for charging the battery on a USB cable (the LTC4088), that I put in the schemes.
After some hours on it, I finally find how to import a DXF file of our PCB.… Read more
On Monday, I began to have a look at a 16×16 device scheme. For this, I took the 4×4 scheme, I copied it 16 times, and I added two multiplexers. One that will multiplex the SCL signal to the multiplex that controls the Hall effect sensor, and one other that will multiplex the multiplexer that controls the H-bridge drivers. Then, I had some questions.
Should I keep a pull-up or a pull-down between two multiplexers or not? My answers for this moment is to keep a pull-up in order to avoid a clock signal on SCL if there is some latch between the two multiplexers.… Read more
Last week, I made a lot of routing to finish our first test PCB. We choose with Alexis to make a PCB that would work with an alimentation because it was easier, so I had a lot of more things to do, but after some work, I finished this part.
Scheme of the box
In order to understand better the design of our box, I made a scheme of the organization if you want to try to do a Touch at home with this design.
This was not clear in my previous post, but this design allows the marbles to flip when the box is put horizontally and vertically.… Read more
After more tries on Tuesday, I found a way to flip the marbles vertically. I tried to use again the plastic part, without results. I make a huge comparison of all the configuration we tried before and at last, I shift a little the marbles from the coils, and with this, they flip in an easier way, even vertically, with a hole of 9.9mm. After finishing this, we can now concentrate on the PCB and the other problems of conception and programmation.
The shifted box is a box with the coils, a disk of metal, a 2mm plastic plank, a 3mm plastic plank (in green) with holes of 12mm of diameter, shifted of 5mm.… Read more