Hi All. I just delivered a project to a client and he’s seeing some noise that never occurred when I tested the project home. The project is 4 panels of 144/m APA102 strips. Each panel has 4x 1m strips that are wired in parallel for a total of 576 channels.
The setup is controlled by a Teensy with the OctoW2811 adapter board, I’m using pins 7 and 14 which are hardware SPI.
Here’s a video. The chase is normal but the random noise is not…
What is the power supply? Seems like here is a signal disturbance, but shouldn’t be caused by an irregular power supply though. Might need a digital filter on the TTL side.
He’s using the same power supplies I was using when I tested, 2x 60A 5V power supplies.
How could I filter the signal ?
Hmmm.
Not sure… Should be a simple LRC circuit though. Never seen this on an LED setup though.
Kinda scary,
I asked him to test on another electrical circuit, we’ll see if it helps.
Otherwise I’ll try playing around with the SPI speeds to see if it could help.
What controller are you using?
Really, LRC filters aren’t all that scary… They’re just a bit confusing math wise.
Hehe, Signals and Systems is the only class that’s keeping me from my Bachelors in Mechanical Engineering. ;o)
It’s a Teensy 3.1 with the OctoWS2811 board so that takes care of the level shifting…
@Nathanael_Lecaude But there has to be a difference with the setup that was working at home ??
That’s what’s I’m wondering… I’ll try to dig further…
With regards to the LRC suggestion, it makes no sense to consider that unless you are able to see the actual noise or crosstalk with a scope !
I do have a scope (an Analog Discovery USB) which should allow me to see what’s going on…
Please follow up with the scope results @Nathanael_Lecaude , I’m interested to see what’s going on there.
Will do, going there tonight !
Well, went to the venue and indeed, lots of digital noise. The noise only seems to happen when the leds are updated, anytime it’s static the noise stops moving (but is still there). The more strips I connect in parallel, the worse the problem is. If I connect only a single strip there’s no noise. I checked on the scope and didn’t see anything weird although I’m a newbie and wasn’t sure what to check exactly.
Now the very strange thing, brought the panels home, no noise… I also checked on the scope and the signals looked similar. The one strange thing I noticed is that at the connector the data line is at 5v and on the strip it drops at 4v. My data cable is very short (6 inches). But I noticed that home too and there’s no noise. Perhaps the LEDs do some reshaping like WS2812 ?
Could it be the AC at the venue that is not stable ? My friend tells me he measured 140v on the AC line which seems quite high…
Here are some scope screen grabs of the data line:
You should not have a 1V drop on that data line.
There is practically no current draw there and it should be 5V.
In my opinion, the AC is ‘unlikely’ to be the problem. PSUs are typically designed to work with a wide range of AC input.
perhaps I should power the teensy from the same supply as the led instead of USB ?
Yes, I would power the Teensy from the same PSU myself. There is a trace to cut on the Teensy to disconnect the USB power.
Did you also have a look at your 5V power distribution for DC drop and also for AC noise ?
Also have a look at the clock line to see what that looks like ?
Just noticed the scope you are using is a USB based scope and it only has about 5Mhz bandwith !
You will not be able to see much detail with that unfortunately.
Last suggestions…
Check the Teensy clock and remove overclocking to 96Mhz if it is set.
Reduce the SPI clock to some lower setting.
Thanks for the tips !
The scope is actually a bit more, around 20mhz, but the -3db point is at 5mhz.
I’ll try powering the Teensy from the main supplies to see if it helps.
I also tried lowering the rate to 1mhz but it did not help.
I still find it strange I have no issues home…