ResourciBoard Part 4

This is a continuation of a project. Find the start here, or return to the previous page here.

The first assembled ResourciBoard prototype powered by DC
The first working prototype.
3D rendering of the ResourciBoard Circuit board
The second version of the prototype.

When assembling the first working prototype I discovered a few issues, so a new version is on its way. I also created a few quick subboards to give this a little more purpose. So I can continue on with testing code until the new parts arrive in about 3 – 4 weeks.

A 100% complete board

Parts are starting to arrive en masse, and I finally got some of the super capacitors with the right pins. It all fits perfectly! So here’s a peek at the first 100% complete prototype.

Fully assembled ResourciBoard prototype with Battery and ATtiny85 chips in place.
The now fully assembled prototype with the large capacitor.
ResourciBoard within a PVC enclosure with a clear lid
Here it is comfortably fitting inside its tiny home.

This will allow me to test a bit further as I found an issue with the small capacitor on the MID rail the other day. My hope is the big capacitor will smooth it out otherwise I need to find another solution.

OMG it works

I have been testing this for the last day or so and voltages seem ok, no capacitor fires, etc. I have managed to get a successful OLED hookup to display a message from the net with no trickery so I’m on the right track.

ResourciBoard showed powering an OLED display via battery.
OMG it works

This means the OLED shield I ordered should be ok, and the data pins are connected correctly. The fetching from the net is a bit flaky but the pushing is fine, so next up is the GPS to see if I can get it to post its whereabouts every couple of minutes.

Makeshift system clock & programming

I got a bit sidetracked and wanted to see if I could get a rudamentary clock going on these so that they could synchronise. The answer was to post a timestamp as a header, that the APP chip would detect when it made a web request, and save it back to the bios. In effect setting the system time just like an NTP server. Then once I had the timestamp I could work out the hour, minute, day, etc.

This was quite fiddly and with such a small serial buffer I ended up losing the first few headers of a web request so I had to reorganise those. And the OLED takes up so much precious ram. But eventually success!

The ResourciBoard prototype pushing a timestamp, and the current hour, minute, and second to an OLED - All via a AAA battery!
The ResourciBoard prototype pushing a timestamp, and the current hour, minute, and second to an OLED – All via a AAA battery!

That allows me to to all manner of time based trickery. I also purchased some USBTinyISP boards to free up my Arduino Uno, and to come up with a plan to supply these optionally with these boards should someone want to program the ATTiny85 chips themselves. So I have designed a rough veroboard to make use of it…

USPTinyISP proposed veroboard socket diagram
A USBTinyISP veroboard sketch.

This has been quickly assembled and tried, and it seems to do the trick.

USBTinyISP arduino programmer with custom socket veroboard
The USBTinyISP & a veroboard based socket adapter

This seems to respond well to the Arduino IDE. So with the next print of boards I’ll make a proper adapter board that can be soldered straight on to this for a more user friendly programmer.

I have heard that V2 of the prototype boards are near so hopefully a second prototype can be constructed soon to prove the DC jack powering of this.

V2 has arrived

ATTiny85 based ResourciBoard now on Black PCB with OLED, GPS, and ESP32CAM shields
V2 of the RecourciBoard Prototype along with some sub boards or “shields”.

The new boards arrived and the black colour should make it look more cohesive and hide some of the nasty soldering iron burns on the underside. It gives it a more professional appearance. The matte finish was unexpected, but looks great!

Initial placement fixes seem ok, the holes are much better for the DC jack along with the MOSFET and such. The ESP12F pads are slightly adjusted and look better. The mounting holes unfortunately are too small still so there will be another version needed.

I have to wait for further parts which are nearby, so hopefully they will arrive and I can assemble what the end result will look like with the proper colours.

Until then, I am designing a few more boards for ideas, such as a pager / buzzer board, which is commonly used here to alert you when your order is ready.

The Pager / Buzzer board that I am working on.

Hopefully the next update will be a nice photograph of an assembled board.

A few graphics to pass the time

I received a few more new parts this week so it was enough to do a mostly complete assembly and move on to choosing colours of the jumpers that would make the most sense.

A mostly assembled ResourciBoard to test the colouring of pin headers
A mostly assembled ResourciBoard to test the colouring of pin headers

I also put together the shields that sit on top to see what issues they might have as I think a third prototype will be needed, and given the two weeks delay I need to get it sent off soon.

An OLED shield, NEO-6M GPS shield, and an ESP32-CAM shield.
An OLED shield, NEO-6M GPS shield, and an ESP32-CAM shield.

The GPS and ESP32-CAM shields need a few positioning tweaks but they should be good for testing. I am also keeping track of the best order to assemble the ResourciBoards and making a howto document with renders and photos, so here’s a few more renders.

How the completed ResourciBoard goes together, with some colour trials.
How the completed ResourciBoard goes together, with some colour trials.
A more sketchy view of the completed ResourciBoard.
A more sketchy view of the completed ResourciBoard.

So I have enough to know that some positioning changes were successful for V2, I have yet to power it on as I’m waiting for more parts but I might just go for it to save time. Then I can get V3 on its way while I work on firmware and the App side.

Faulty LM317T linear regulators

I have been powering it up and trying a few things, it’s looking good with a few labelling changes and no blown capacitors however I discovered an issue where it used to be able to power the ESP32Cam on the original board, but with the new board it can’t. With a lot of trial and error it seems to be the LM317T regulators.

Testing the LM317T with an ESP32-Cam
Testing the LM317T with an ESP32-Cam

I checked where I got them from and it seems others have found the same issue – the current from them isn’t right. Where it should supply amps, it can only do <200ma. This isn’t enough for the camera.

The brownout detector was firing as there wasn’t enough current on the 5v rail. ESP32cam’s are hungry.

These are the faulty ones – the comments at the bottom reflect the issue I am seeing: https://www.aliexpress.com/item/10PCS-LM317T-TO220-LM317-TO-220-317-1-2V-37V-1-5A-new-and-original-IC/32842362589.html?spm=a2g0s.12269583.0.0.6e1b40da4cJF99
They’re from a store called “CHIPLIJIAYUE”. Best to avoid it.

These are the ones I was using previously which work much better: https://www.tme.com/au/en/details/lm317-cdi/regulated-voltage-regulators/cdil/lm317/

Hope this saves someone the days of trying to work out what is wrong!

Update: I have ordered some more from a local store and they are testing ok. Here’s the comparison:

Original working "CDIL" LM317T on the left, the faulty/fake one in the middle which people are guessing might be an LM317L internally, and a working LM317T on the right.
Original working “CDIL” LM317T on the left, the faulty/fake one in the middle which people are guessing might be an LM317L internally, and a working LM317T on the right.

Refinements

I have spent the day cleaning up some of the issues I have discovered as I continue testing. Most of the issues are cosmetic or physical fitment.

The latest resourciboard, now with friendlier solderpads & more consistent jumper labelling – along with fewer lines to give a cleaner board.

I have also received most of the parts I need for the new pager board so I have been busy designing so that it can be sent along with the other boards for a new print.

The new and improved pager board.
ESP32cam now with clearance to sit lower, and programming headers.
GPS board with the module flipped, clearance to help it sit lower, and optional 7333 if the onboard 3.3v isn’t enough.

These are now looking pretty good. I need to wait for the new LM317t’s to test, and the buzzer component to see if the holes are in the right spot. If those two things work out then it’ll be ready to send off.

I also received a second batch of “Yellow” pin jumpers, read about the debacle here.

Update: Buzzers arrived and look good, and LM317T’s test ok. Also got the ESP32Cam to work correctly so all the above boards have been sent off to manufacture! I can already see one issue but it’ll do for now.

3D rendering of the gerbers, front view
3D rendering of the gerbers, back view

Hopefully this will now result in a working solo board, pager board, ESP32cam board, OLED board, and GPS board!