Minor kinks in layout + BOM

[vguttmann]

I just assembled my PCB, and I noticed a few improvements and also a few errors in the BOM:

• the polarity markings for C47 (electrolytic capacitor) are missing
• the button to step through the program has the pin spacing for a header; so the buttons from the BOM won't work
• the resistors that should be 22K are actually 2K in the BOM
• the switches linked in the BOM are already a quantity of four, so already more than necessary

Then, there are a few suggestions I have that would make assembling easier and not require a look at the BOM/at KiCAD:

• the markings for C58 are covered by the IC/socket it's next to
• the types of diodes should be designated everywhere, not just in the control unit
• the Potentiometer linked from the BOM is intended to be panel-mounted and not soldered to a PCB
• lastly, adding the capacitances for at least the caps that are not 100 µF would also speed up assembly.

If you could just post the links for the new BOM items below, I can add them to the shopping cart at Mouser, and remove the old ones.

[edson-acordi]

Glad you finished the PCB assembly!
Thank you for all the suggestions.

• It's good to have polarity identification, I will put the polarity (C47) mark on PCB. However, it is easy to identify the polarity with a multimeter, measuring the continuity of one side with GND.

• All the connections are made with a pin-header, thus, give you more freedom to use any type of button/switches. However, most types of Tactile Switches, have 5 mm of pin spacing. I will update the layout of SW1 and SW2 in the next PCB revision. Sorry about that.
  For now, you can use it like this (or you can try to shape the pins with pliers if you want to put it directly on PCB),

• I will update the correct resistor for 22k (https://br.mouser.com/ProductDetail/603-CFR-25JB-52-22K)

• In the BOM I have made, the quantity of switches is 2, for Step-by-step execution and Reset, so, it is ok.

• I will update the pcb with corrected C58 text position.

• It is nice to have all types of diodes with its own designation value, I will update it on PCB.

• The potentiometer, as well as all components that's need to be accessible, are put on pin-header connections, to give you more freedom to use.

• I will try to identify the value of all the capacitors on the PCB. However, it is easy to identify all the decoupling capacitors (100 nF), as they are close to the first and last pins of the IC's, always in the same position in relation to the IC's.

[vguttmann]

With all due respect: potentiometer on a wire is well and good, but in the end: It's just a potentiometer. Anyone that would care that much about the choice of potentiometer will have the skills to modify the PCB files.

Furthermore, a potentiometer that's just flopping about could accidentally short circuit something, and with the BOM cost of close to 200$, I'd prefer if no one accidentally ruined theirs.

I mean, you can keep the pin headers for both buttons and potentiometers, but I'd also add footprints for PCB-mount options (ideally, the header holes would be covered by the component so it's clear you don't have to install both.

When I'm home this weekend, I'll bring an Arduino with me, so I can do some long-term testing. Since I do have a rather advanced scope (well, as far as hobbyist scopes go, it has 2 analog and 16 digital channels), I could set up a long-term test with fault monitoring, I'd just need your help in designing an appropriate test scenario.

[edson-acordi]

You could try this type of potentiometer (it's tiny), and it would fit under the footprint of the current pin header,

https://mouser.com/ProductDetail/652-3310C-001-105L

It's not difficult to modify the PCB to fit another type like this one (I may update it on the next PCB revision),

https://mouser.com/ProductDetail/652-PTV09A4020SB105

So, you can decide to use pin-header or a potentiometer.

I'm finalizing a doc that explains how to transfer programs using Arduino, so you can do lots of tests. Likewise, I will prepare a set of tests to run for a long period of time and confirm whether or not the hardware will fail at full speed.

For now, I'm doing tests with a 16x2 LCD (HD44780), and I can tell you that, with some hacks, it is running at 4 MHz. But need more observations for a long time.

[vguttmann]

As I said, I modified my order slightly to add mounting holes for the poti you may add in the new revision.

If you are going to make the new revision soon, I'll add this to the basket now, but if it will take a couple of months, then I think I'll wait till then.