Admin Group
Joined: 31/12/2012
Location: New ZealandPosts: 9375
| Posted: 07:45am 20 Apr 2024 |
Thanks, Jim.
I have replaced the LCD female headers, but nothing has changed.
Guru
Joined: 31/01/2014
Location: United StatesPosts: 326
| Posted: 08:51am 20 Apr 2024 |
The one thing I noticed looking at the images is the GND and CS pins appear to be connected to a copper pour. Is that section of copper pour truly grounded?
--Curtis
Guru
Joined: 20/06/2011
Location: AustraliaPosts: 2924
| Posted: 09:43am 20 Apr 2024 |
GDay Grogster, All,
Check that the TFT GND pin is actually connected to GND, by adding the switch you ‘may’ have an isolated GND pour meaning that GND on the TFT is at actually system GND.
It is hard to follow the traces with the colour choices but it is basically the same as the original except for the switch and different HC12 layout.
Kind Regards,
Mick (the big one)
.
Admin Group
Joined: 31/12/2012
Location: New ZealandPosts: 9375
| Posted: 11:34pm 20 Apr 2024 |
Good point, but yes - the LCD GND and the CS pin are connected to the common ground.
Once I unearth my DSO next week, I will try to get a screenshot of the D/C line that Jim wanted.
Guru
Joined: 20/06/2011
Location: AustraliaPosts: 2924
| Posted: 01:01am 21 Apr 2024 |
Hi Grogster,
Looking further it does seem that a GND does connect to the Pic and the TFT but that image above is your old version without the new switch.
This is a curious issue.
Suggestions are to minimise it by removing sections you don’t need (like the SD connector and any HC12)
Not much to minimise though.
Are you able to post images of top and bottom copper without the copper pours and individual copper layers (with pours) and without the other layer showing? That would make tracing the wiring easier.
Regards,
Mick
.
Edited 2024-04-21 11:01 by bigmik
Guru
Joined: 20/06/2011
Location: AustraliaPosts: 2924
| Posted: 04:02am 21 Apr 2024 |
Grogster,
Another thought.
Are the 2 capacitors on the VReg actually 10uF? You may have accidentally used 10nf maybe?
I can’t see any logical reason for one board to work and the other not to.
Mick
Guru
Joined: 24/12/2017
Location: AustraliaPosts: 1632
| Posted: 04:16am 21 Apr 2024 |
I did ask about the capacitor and Grogster said it was OK. The PIC does run and drive the LEDs but I guess it could be marginal? Could be worth trying another.
Bill
Guru
Joined: 05/10/2019
Location: United KingdomPosts: 7106
| Posted: 09:50pm 21 Apr 2024 |
The regulator pads look right for the SOT-223 package. I hope that's the one used here as the SOT-89 is different.
Admin Group
Joined: 31/12/2012
Location: New ZealandPosts: 9375
| Posted: 09:53pm 21 Apr 2024 |
Yes, SOT-223.
I'll upload the images Mick asked for later, but have some other work to do today, so that will be later on tonight my time.
Guru
Joined: 31/07/2019
Location: AustraliaPosts: 507
| Posted: 01:35am 26 Apr 2024 |
Any resolution ?
Without a deep dive my only thought is that the blue pour connections running between the 170 pins seem thin, maybe one is intermittent or high resistance.
Displays can be a pain.
(ed - what size is the display ? I got the impression it was 2.2 ? As it happens I have a similar display (2.2") that works OK but the controller is ILI9225.)
Edited 2024-04-26 11:46 by zeitfest
Admin Group
Joined: 31/12/2012
Location: New ZealandPosts: 9375
| Posted: 11:17pm 26 Apr 2024 |
No, and as it is relatively low-priority compared to other paid work I needed to get done, it has basically just sat on the bench for the last few days.
I have not forgotten it though, and will get back to it when I can - which should be sometime next week when I expect to have a little more time to tinker.
LCD is an ILI9341, 2.2" without touch.
Admin Group
Joined: 31/12/2012
Location: New ZealandPosts: 9375
| Posted: 02:58am 08 May 2024 |
OK, I have this working perfectly now.
It was a SCHOOLBOY mistake - anyone wanna guess what it was, before I tell you what it was?
Nothing to do with the PCB, the traces or the layout.
Anyone wanna take a stab at what the problem was?
Regular Member
Joined: 12/10/2014
Location: New ZealandPosts: 93
| Posted: 03:11am 08 May 2024 |
Something to do with the power supply?
Guru
Joined: 24/12/2017
Location: AustraliaPosts: 1632
| Posted: 03:17am 08 May 2024 |
I can only think of it being the wrong value component but the MM2 and the Microbridge work and the backlight is working. I give up.
Bill
Admin Group
Joined: 31/12/2012
Location: New ZealandPosts: 9375
| Posted: 03:27am 08 May 2024 |
You win.
I installed the wrong regulator!
Supposed to be MCP1703A-33(3v3), I put in a MCP1703A-50(5v)
As soon as I spotted that, I had a bit of an "A-HA!" moment.
Replaced the 5v regulator with the correct 3v3 regulator, and everything sprung to life - on the original pin-out etc.
I took it from the static sheilded bag containing regs that SHOULD be 3v3, but I must have let one 5v slip into that bag, when I sorted them. I will now check all regs in both the 3v3 bag, and the 5v bag, so as not to run into that again!
I guess I also should have been checking the part number on the reg itself, rather then just ASSUMING that every one in there, were the correct output voltage...
Again, a pretty schoolboy mistake, and I should have detected that WAY before now, and everyone also knows what assumptions are the mother of...
That mistake, would have meant that the MM2 chip was firing 5v logic(or close enough to 5v) at the LCD, and I guess it upset the SPI port on the LCD module, whose I/O is 3v3 max I seem to recall.
I'm probably pretty lucky that I did not kill the LCD, but it still is working fine!
Probably lucky that I did not kill the 170 MM2 chip too, as it is only rated for 3.6v maximum!
Edited 2024-05-08 13:30 by Grogster
Guru
Joined: 18/11/2011
Location: United KingdomPosts: 3907
| Posted: 06:32am 08 May 2024 |
Finding it ... persistence paid off!
John
Admin Group
Joined: 31/12/2012
Location: New ZealandPosts: 9375
| Posted: 06:39am 08 May 2024 |
The only thing I am a bit sorry about, is starting this thread, and it taking three pages.......to find out it was my f-ing mistake all along.
I could certainly have simply NOT posted the fix - but then, that does not give the rest of you a chance to have a giggle at my expense!
Guru
Joined: 05/10/2019
Location: United KingdomPosts: 7106
| Posted: 07:58am 08 May 2024 |
It's never easy to spot your own errors, you are far too close to the design. This is something I learned while working as an electrical draughtsman. If possible you should always check each other's work. Sometimes it only needs one wrong wire number to start hours of fault finding.
Guru
Joined: 17/05/2016
Location: United StatesPosts: 3186
| Posted: 11:34am 08 May 2024 |
Shades of the days of program "desk-checking" by a group (in a period when a new compilation required because of a typo might mean a 4-hour turnaround time).
But is there a single one of us active on this thread who has another shedder available to look at physical parts? Having a worldwide group is great for solving software problems or design problems, but more limited when the solution is to look at a part with a magnifying glass (while knowing what you are looking at and why).
Congrats on finding the solution--and for sharing.
Guru
Joined: 05/03/2018
Location: NetherlandsPosts: 4529
| Posted: 06:38pm 08 May 2024 |
Grogster,
You can always say that the box with 1703 regulators also fell on the floor, at the same time the 10-100-1k resistors fell...
Volhout