RF interference is REAL!


More of a PSA: RF interference is real! I had a successful setup of Octoprint on RPi4 with an Ender 3 Pro. Then I introduced a Gosund LED Light Strip which uses the Tuya style of wireless connection. I started getting connection problems and the printer would randomly stop mid-print. This can be really frustrating when you have 10 hour print. I Googled the potential issues, and it always came back to "check the cable", "check the signal noise." The smart LED strip was the only net new addition to the environment. But given that there was a wireless webcam very close by, in the environment, plus other Tuya-based smart devices all around my garage, I refused to believe that the LED strip had anything to do with it.

Instead, I ordered a brand new gold-tipped USB cable, I ran Octoprint in safe mode, I reinstalled Octoprint several times, using several different versions. Out of options, I disconnected the LED light strip, not just turned it off, but disconnected it.

Voila! 10 hour print that I started and tried, and retried, a week ago - success first try. I'll never ignore the experience of others on the forums!!! :slight_smile:



I had the same problem, it turned out to be the web cam I was using. Change it for a pi cam sorted it out.
Web cam must have been drawing to much current.

Question: Why does wireless interference cause a print to fail? Should OctoPrint keep chugging along regardless of what is happening on the WiFi side of things? If this is the case, wouldn't this be a bug? I thought that the network connection on OP during a print only has the sole purpose of serving data to the user. Is that not the case?

It's not the wireless interference. It's the interference form power devices like washing machines, hand drills, etc. Those can produce spikes on the power lines that can induct to your printing system: A bad power supply, a bad USB cable, a bad wiring.

Any cable or track on a PCB is an arial, even a piece of plastic with a metal coating can act as an arial. This is why an oscilloscope that measures high frequencies has probes that cost more than your house, otherwise the probe gives false signals. And why on modern multilayer pcbs you have a whole layer that is a ground plane, to reduce interference.

Modern appliances use switch mode power supplies, which can be noisy, especially if cheap. It is why audiophiles etc still insist on very heavy linear power supplies which are cleaner but inefficiently does heavy.

Also ironically something that is very noisy is motors, e.g. stepper motors, this is why you get things like Z banding on the cheap motor drivers.

In printing common symptoms of noise are poor WiFi, corrupt serial, z banding, probes or endstops randomly not working or miss firing.

1 Like

Just to make this clear, this IS the case, and as pointed out the interference issue lies with the serial connection (and/or the printer itself), NOT the wifi connection.

And no, OctoPrint can't detect that. The only thing it can detect is a serial connection that constantly breaks off and sucks in general. That can have a myriad of issues which all need to be fixed by the user and outside of OctoPrint (or any host for that matter), so when stuff breaks it links to the FAQ item in question.