ESP8266 Community Forum

- Tue Jul 30, 2019 5:36 pm #83286

An easy workaround (although quite hackish) would be to place a few extra WS2812b's in between that 10m stretch.
This will reduce the wire's capacitance and each one will clean up and amplify the signal (of course you would need to account for those extra LEDs in your code).

Due to the layout of the display this is (almost) impossible as the existing cable runs behind about 100 leds of the display (in plastic conduit) I could (perhaps) change the layout which would bring the start/end points close to the NodeMcu but this would require a major amount of rewiring to the display. But TBH this might actually be worth doing anyway so perhaps reworking the layout "may" work better for future additions.

Hmm I need to seriously look at this option of moving the Start/end point...

Thanks for the idea!

Today it's LEDs, tomorrow IT'S MORE LEDS!

- Wed Jul 31, 2019 12:45 am #83287

An easy workaround (although quite hackish) would be to place a few extra WS2812b's in between that 10m stretch.
This will reduce the wire's capacitance and each one will clean up and amplify the signal (of course you would need to account for those extra LEDs in your code).

Due to the way the display is wired (in plastic conduit) the supply cabling (and signal wire) pass behind about 100 leds in the display to get to the 1st led on the strip. So adding extra leds before the main sequence is impossible.

That said there is an option of moving the sequence start/end to another location nearer the NodeMcu, but it requires a fair amount of rejigging of the power/signal connections and may make the display "unbalanced" to look at. There is however a possable advantage to this though so perhaps this may be an option... Its an interesting idea that (if all else fails) might work.

Thanks!

Today it's LEDs, tomorrow IT'S MORE LEDS!

- Wed Jul 31, 2019 12:52 am #83288

An easy workaround (although quite hackish) would be to place a few extra WS2812b's in between that 10m stretch.
This will reduce the wire's capacitance and each one will clean up and amplify the signal (of course you would need to account for those extra LEDs in your code).

If I've already posted a reply to this then forgive me. Lack of sleep (3hrs) and a phone that, this morning, has a mind of its own isn't helping...


Due to the way the display is wired (in plastic conduit) the supply cabling (and signal wire) pass behind about 100 leds in the display to get to the 1st led on the strip. So adding extra leds before the main sequence is impossible.

That said there is an option of moving the sequence start/end to another location nearer the NodeMcu, but it requires a fair amount of rejigging of the power/signal connections and may make the display "unbalanced" to look at. There is however a possable advantage to this though so perhaps this may be an option... Its an interesting idea that (if all else fails) might work.

Thanks!

Today it's LEDs, tomorrow IT'S MORE LEDS!

- Wed Jul 31, 2019 5:32 am #83296

An easy workaround (although quite hackish) would be to place a few extra WS2812b's in between that 10m stretch.
This will reduce the wire's capacitance and each one will clean up and amplify the signal (of course you would need to account for those extra LEDs in your code).

Further to my other (3) replies (sorry!). I've had a little play this morning with my whole design and as you (sort of) suggested i did try to move the LEDs about a bit. I managed to move the 1st led from about 10m to about 3.5m (that's the closest I can get to the strip without moving the processor), and....

I can get the last (now first) led strip to work fine but the next strip (formally the starting position) then starts to goof up. There's a gap of about 2m from one strip to the other over the same twin and earth as before. Very strange... High resistance perhaps at one end? I was unable to check one of the connections due to rain (and a flat roof full of water )

Further to this (I've been playing some more) I fitted a simple matrix (37x8) about a month ago (before I looked at the ESP8266) to scroll text and this seems to work fine over a distance of about 7m (nodeMCU or arduino). The main differences between the systems are 1] no breaks in the cable (matrix is on dedicated cable though share same supply) and 2] the matrix runs on AdaFruit librarys and not FastLED as the main display does.

while a useful experiment it still it doesn't actually help with my initial question which boils down to "Why does the arduino run the display yet the nodeMCU doesn't?"

Until I can get back on the roof I'm stuck (I've changed the display back to its original configuration, and no change with ESP8266 not "doing it" and arduino happy as normal )

Thanks

Today it's LEDs, tomorrow IT'S MORE LEDS!