trim igloo
I’ve been trying for days to get these yeyeass displays working again. They just magically stopped working out of nowhere when I changed the EFIS cables (I know it doesn’t make any sense, and it really doesn’t).
Here’s the list of things I’ve already tried:
-Soldering ceramic capacitors onto the buses
-Improving the ground routing
-Replacing the 5V car charger converter with an industrial buck step-down converter
-Running configuration with only the displays
-Adding an extra Arduino dedicated only to the displays
-Testing only the displays connected in the setup
-Testing with another USB cable
-Testing with another Arduino
-Reducing the configuration execution speed
-Connecting only one display
-Testing with different brightness levels
-Testing changes in test mode
-Replacing the MAX7219
-Replacing the display wiring
-Using another power supply
-Testing with a brand-new Arduino (again)
-Testing without external power (even with just one display connected at minimum brightness, which should be enough)
-Testing without a capacitor
-Soldering wires directly onto the Arduino board
-Testing one digit at a time
-Testing without a GND bridge between the Arduino and the module
I’ve also read through the MobiFlight documentation:
-MAX7219 issues
-7-segment modules
-Troubleshooting
I’ve also researched multiple ChatGPT discussions about adding 10µF capacitors and ceramic ones — with or without them, the problem remains the same. The MAX chips and Arduinos are brand new, straight out of the box.
here are some reference pictures:
What intrigues me the most is that it was flawless working before ;/
sinful oak
I don't know if it makes a big difference, but I would not twist the three signal wires together. Twisting is supposed to be good when one or more of the wires is GND. In this case, none is GND so they might create induced currents from one to the other.
Capacitors should be installed as close to the chip as possible. I see you soldered a ceramic capacitor on the outlet side of the board.
There is usually a diode on the boards that blocks any current in reverse coming from the outlet side.
trim igloo
tested with the capacitor on the power inlet and without twisting the cables and the problem persists, the brown 10uf capacitor isn't shown on the pictures bcuz i test with/without it
it was working just normal for several months... if the problem become the displays i'm wasted because it'll result of opening the entire unit
or may the reason it's because i'm not using a MAXIM chip...? but how could a parallel be THIS instable, even brand new
sinful oak
Most boards have a space with two soldering holes where a 10uf capacitor can be soldered.
Most boards also have the 100nf ceramic capacitor already installed as C1
sinful oak
Your board appears to have it too
Max7219 chips are known to be very sensitive to noise and power quality.
trim igloo
yeah i already tested with a 10uf capacitor there
i tested many kinds of power before, different 12v conversors, 5v phone charger, directly through the arduino 5V and like, not even with ONE digit minimum brightness it sustains
but my main concern is why it stopped working out of nowhere after months 😵💫
trim igloo
Max7219 chips are known to be very sensitive to noise and power quality.
but why are they? do mobiflight has support to another kind of chips that works better than this goofy chippy?
or any more reliable way to connect the chip to the displays idk
sinful oak
In my experience, some issues with noisy power but usually installing capacitors solves the problem.
trim igloo
yeah i think i bought fauty max boards... :/
sinful oak
There is support also for TM1637 but I have never used those.
Have you tested starting from a new profile and only recreating the output configs for the Max7219 displays?
With nothing else.
trim igloo
hmm I didn't try this
but I think i'll (also) try ANOTHER board on a stanby display to test too, i'm kinda running out if ideias hah
but i forgot to mention, that they sometimes bug even on test mode
slender elk
Are-you using the power supply off one board Max7219 to support others boards similars?
If you saw MF pdf it is recommanded to suply every boards with differents wires nt use the PC board VCC and GND of one to second.
In that case you can create a kind of ground layer and +5V Supply and use in front of each PCB capacitors as explained Jaimleon
In my opinion also when you decided to use wires it have to be perfect use in that case Dupont Wire Flat or if you do not have Twisted cable the same for supply twisted Black and White.
or use two copper PC board isolated one two one and connect supply on this in different point.
In that case you create an kind of PCB similar layers Ground and +5V
trim igloo
@slender elk i'm trying to power only one max, at minnimum brightness, at just 3 or two digits, and even with this low power consumption it fails
even throught it worked without before, I trying installing those capacitors that everyone mentions closest as possible to the vcc and grd but still failing
trim igloo
There is support also for TM1637 but I have never used those.
they seems less stable than the max... also the max itself should be enough as i see many projects using them
trim igloo
it's working with only 3 digits on the arduino without external power, but as soon i put more digits it dies
using external power even with the capacitors and jointing the ground the instability multiplies 100times, already tested mobile charger, 12v bulk inverter, car charger (who goes to 12v down to 5)
sinful oak
I am out of ideas. In my experience, up to two max7219 modules can be powered from one Arduino board.
slender elk
Do-you separate wiring of each group Speed heading for example all group and you can also use aluminium paper to rool the wires and put to ground to avoid effects.
did-you try only with one group and verify it works the same with the second but without the first one So
Speed verify
Heading verify without Speed
Altitude verify etc.
trim igloo
yeah I tested separed configs, but its crashing even on test mode, might do this alluminum test too
tested with 2 1K resistors on CS/CLK and now it seems to be stable... tested 3 digits on full brightness, now i'll solder the VS digits and test with all the digits at once
slender elk
sorry stupid question is it 7 segments digit?
slender elk
Did-you solve it? Did-you try withou wire and 7 segments digit directly on the PC board or test with a complete PC board with MAX7219 and Digit only wire for VCC CLK CS DIN
In my opinion your problem is a question of wires it would be much better to have all 7segments digit in front off your FCU .
Your cabling is too complicated when you can keep them simple.
For example, wiring 12 wires for 8 digits when you could use only **three **and Ttwo power supplies, thus separating each Speed, HDG, ALT, and VSS group.
Sorry for my conclusion but in my opinion you have to modify your product which in fact will be nice👍
viscid wren
I had so much issues with my 7 segment display for my overhead. It is the most frustrating component of a sim build in my opinion. As what @slender elk already mentioned. My experience is the wire gauge is one of the most important component of a 7 segment display. The display goes all whacky if not connected properly.
I had external power to my display before and I could not get it to work. Decided to just power up from arduino and it works as long as I don't touch the wiring from the display. Very temperamental component.
slender elk
I agree with @viscid wren is not easy as component.
But **Are-you using FSUIPC7 inside SimVariable offset?**some plane need to use these variable to work?
in my case I use directly these board for my B737 Land altitude and FLT different but same principle ==> issue i have I had a problem with +5V and ground as explained and separate it. found explanation inside MobillFlight pdf
for example Warning
MobilFlight advice WARNING
MAX7219 modules are sensitive to poor electrical connections and low voltage.
Always use high-quality cables and an external +5V power source when connecting these modules to a board. When chaining multiple MAX7219 modules do not daisy-chain the +5V power.
https://github.com/MobiFlight/MobiFlight-Connector/wiki/Example-7-segment-display
About your photo where we saw a lot of wires connected in my opinion uses these wires Dupont Wire Flat or similar
So you can consider that it is a complicated circuit.
trim igloo
i'm still working out on a fix and i guess now i'm getting somewhere
MobiBOT plz don't close the thead since when I get everything done i'll share the solutions here and my considerations
slender elk
seems very good we try to help you as we can you use FSUIPC7 and variable offset
trim igloo
(ignore the wires all messd around hahahashjhw)
trim igloo
seems very good we try to help you as we can you use FSUIPC7 and variable offset
i'm using fsuipc offset for VS only
slender elk
I am surprised I think that you have to use for eevery Speed HDG and Alt but personaly i made an overhead see what said other guy or try it to add in Mobilflight program
trim igloo
the main issue wasn't the config itself, as it used to work normally past years, was more like the displays brakedown out of nowhere
slender elk
but i saw on your video that you are using B737 suppose Pmdg and sure it is necessary to use all FSUIPC7 offset
in fact you create a nice product
trim igloo
I had so much issues with my 7 segment display for my overhead. It is the most f...
man i hate my lack of PCB knowledge and how instable are these chips without a proper pcb
But it is what it is
slender elk
YEs perhaps but when you box will be closed it will be perfect
I think that you are right about Fsuipc offset you need only** VS control** in fact we learnt every day yourself. 😉 How do-you do for others varibles Speed Hdg and Altitude?
trim igloo
regarding the MAX7219:
-Sacrifice a few extra Arduino pins instead of using a daisy chain if you want better stability.
-It is recommended to use 10 µF capacitors (as well as ceramic ones), although in my case they didn’t make any difference.
-Wires should be as short as possible, with a minimum thickness of AWG 26.
-use a decent 5V power supply and wire all grounds together
**
-What made a huge difference was adding 100–330Ω resistors (i tested with 1k and worked alright) to the arduino pins connected to CS and CLK, since those are the most unstable pins.
**
my wiring isn't the cleanest or most electrically clean, but that actually proves the module’s stability so far
about the MCP, it’s a project I made some time ago (before this whole winwing slop) that ended up becoming obsolete
It's fully operational, including the disengage switch, which was modeled from scratch and entirely made on the small Ender 3 V2.
The Mobiflight configuration was done entirely via PMDG SimConnect, with only the VS using FSUIPC.
my wiring is not the cleanest or most electrically ideal, but that actually proves the module’s stability so far.
viscid wren
Are you powering up the 7 segment with external voltage? I couldn't get mine to work but when I just use arduino 5v everything is stable.
I went through around 10 iterations of the Batt display and maybe around 5 for the flt alt and land alt display before I got it right.
Currently none of my displays are using any external powers. so far so good.
trim igloo
It seems that using an external 5V power supply can introduce a bit more noise into the project/arduino, since there are two power sources sharing the same ground. In practice, it’s noticeable (not in terms of power supply, but in stability) that a single MAX is more stable when powered directly from the arduino.
In my case, what solved it was using resistors.
viscid wren
it worked fine when I tested it on the bench with just the display, external 5v but when I introduce it to the overhead, it won't light up or strange characters appear. I didn't want to troubleshoot so I went without the external 5V. Probably as you said about the noise.
