Build of Deployable Units

[Penri]

Terry

I'll wait until you have everything ready, please.
I've purchased some IP rated boxes but would like to have everything at hand to do trial fits as the boxes are larger than I'd like.

I also need to work on interconnections between the Shopfronts / dioramas and find a suitable mounting scheme.

Hwyl

Penri

[TerryJC]

OK. No problem.

This time I won't make any promises about how long it will take me, but barring more domestic issues (the tumble drier packed up on Thursday evening), I would have thought a few days or so.

[TerryJC]

I decided to try out switching the screens around, as outlined above. The screen at Position 1 was one from the original batch, which I'd never seen any problems with. Positions 2 and 3 were populated with screens from the most recent purchase. I swapped the screens between Position 1 and Position 3 and the flickering went with the screen.

To me this means that more recent builds of these screens were probably manufactured by a different Chinese Sub-contractor, who a lower standard of materials and / or methods. Couple this with the design technique that I used to display on multiple screens and we probably end up working at the limits of their capability. I know I've seen this particular screen working OK when I first tested them, so we may have to reserve any 'dodgy' screens for the single screen installations.

BTW. This may mean that some of the screens from the second batch might be coerced into working, although this is less likely because I'm fairly certain that I tested them individually.

I'll keep you posted.

[TerryJC]

I have (mostly) completed and tested the first of the Single-screen Interface Boards and it works well, even with the screen from the last batch which flickers on the Multi-screen Interface Board. This board is currently missing two Pi retaining screws (to be cut down from the 12 mm screws that we have in abundance) and needs configuration of the WiFi, IP Address and Autostart of the software to be considered complete.

I am about 50% of the way through completion of the second Single-screen Interface Board which should happen tomorrow. However, I have trips to the hospital, the Vaccination Hub and the doctors over the next three days, so this board might get held up again.

I haven't tried them yet but the boards that were all faulty may include two that will work in this configuration. I'll try them out and report back.

[TerryJC]

Penri,

I'm just about finished construction of the second Single-screen Interface Board and should be able to deliver 3-off Multi and 2-off Single Screen Boards to you very soon.

However, I've been thinking about why one of the Screens from the German Supplier is noisy in a Multi-Screen configuration and works when used with a Single-Screen Board. As far as I can tell, the only difference between the two configurations that would effect the display quality is the power supply because the problem occurs all the time and not just when the I2C Bus is active.

I had a look at the 3.3 V power supply output arrangement from the Pi Zero W (see circuit attached) and the 3.3 V power input arrangement on the Screens (see circuit attached). The Pi has a 4.7 uH inductor in series with the output and a 4.7 uF capacitor across it (the circuit appears to show that as non electrolytic). The screen, in turn, has a 100 nF capacitor across the 3.3 V input. I have lots of 10 nF capacitors but unfortunately no 100 nF. Also I have no 4.7 uH inductors and the only 4.7 uF caps that I have are electrolytic.

What are your thoughts on this before I pass everything to you?

Call me if you want to talk this through.

[Penri]

The decoupling and conditioning circuits look fairly good to me on both diagrams.

Can you see any significant noise on that or any other PSU line?

I will have some 100nF ceramics, not sure about 4u7F non-electrolytics or 4u7H inductors but will take a look and let you know.

I'm not around for a few days from Sunday so hold on to the items until the latter part of next week, please.

Penri

[TerryJC]

Penri,

I can get 4.7 uH inductors from Amazon within a few days.

I'll have a look at the PSU lines after I've completed the final Single-screen Board.

My main questions was really; is it worth spending yet more time on this?

[Penri]

I’m happy to install them and see how they perform in the wild!

[TerryJC]

I’m happy to install them and see how they perform in the wild!

I'm currently taking PSU readings on the scope, so maybe that will show something.

Here are the results for a flickery screen on the Single-screen Board, First, straight after boot up with the screen not running, then with the screen showing images:

I'll finish fettling the second Single-Screen Board and then repeat the above measurements with the Multi-screen board.

[Penri]

I suppose I'd like to see the noise at power up to be less non a ideal world.

If I'm reading things correctly the fundamental frequency of the noise at boot up is approx. 1kHz, does that sound about right for the DC-DC converter being used?

The pulses when the display is operating are not what weblike to see but the overall amplitude is not much greater than the power up noise.

Its interesting but if its not causing an insurmountable issue then we could investigate in slow time while I get on with integrating the modules into WMT.

Penri