Close

Things just can't be that simple, can they?

A project log for UxV ground station

creative name pending

mkdxdxmkdxdx 07/27/2024 at 16:320 Comments

So, that old screen holder was mostly for show - it is functional but it's ugly and it does not do everything i want it to, it's not rigid enough and mounting it on the deck is troublesome, requiring internal access.

I had to redesign it and after that, there was another adventure. Lots of pics under the break:

Second screen holder iteration was easy to mount and looked fairly slick, but it also looked hella boring. 


And it still wasn't rigid enough, i saw how it flexes when i was applying just a bit of pressure.

And so, fast forward 3 days or almost sleepless design and printing and borrowing time from touching grass, ive got this:

A competent, fairly sturdy and solid screen holder.
And because i cannot escape adding "modularity", whatever that may be, it also has two compartments with heat set nut slots to mount something there or cover it with a stub panel.
And it is still detachable and can stand on it's own and be used as separate monitor!
But after putting it all together and talking the deck to a brag-walk, turned out, big screen shows garbage:
On one hand I was not really surprised - after all this panel is constantly remounted on new case prototypes, as well as it's tender connectors are tugged every time, but on the other hand - wtf, was it a time for a replacement?
Walking the fault chain and excluding points one by one, it turned out it was not RPi issue (it could show picture on this port on other monitor fine), and it was not a display issue because it worked fine with other source, and it was not a FFC issue that sends data from connector to an extension connector on the panel:
But it was the female HDMI connector itself that had a lifted pin (it's 6th from the left):
And running a soldering iron over that header + some luck, i was able to fix it without causing bridging:
And indeed it worked!
But there was also an issue of reboot loop after a while. At some point i've decided that this deck will accept 2S-6S lithium packs and because of that it needs a wide range DC/DC which currently is LM2596HVS which can accept up to 50v and output up to 2A safely:
But hovering my finger over it immediately revealed that DC/DC was constantly entering thermal shutdown:
A bit of backstory - i'm currently employing a hack that allows to power big display with 5v to save on DC/DC boards count in my system. To do that, i hacked in a little 5v connector that is soldered on after internal screen DC/DC, to it's output, thinking it probably won't be much of a big deal.
I was very surprised when measuring how much current everything draws, screen was doing 1.35A for some reason at it's 5v. And i was surprised again, when after disconnecting everything and connecting a proper 12v supply to it's proper connector - DC/DC decided to pop with smoke and sparks. So i guess routing power around converters is not always a good idea. It popped so hard, it blew off it's input pin with a signature package bulge.
So naturally, I removed the converter chip thinking i might get away with just properly powering everything without it.
And after that, panel shows nothing. I see that controller works - red/green power leds cycle properly when pressing power button, RPi identifies the display fine, but there is no picture. My current best guess is that it's either EN signal was severed to somewhere when i've desoldered DC/DC chip, or maybe something else was shorted whole time DC/DC was dying. I'm no expert in power circuits designs, unfortunately.

So, yea, it is time to buy a replacement, becase i'm kinda done for now with hacking.

Discussions