Power supply buzzing with OctoWS2811 and Teensy 3.2

popcorndog

Member
New to using teensy and fairly new to LED WS2812B strips.

I have a pinball cabinet that I am adding LED rgb light effects to.

Issue: Connecting USB cable to computer results in power supply buzzing when orange data pair is connected to Backglass Lights. Removing the USB cable or the backglass data cable stops the power supply buzzing. Any suggestions are appreciated. I do not find any shorts. Is it possible the OctoWS2811 is bad?

Tested: Power supply connections (3) are each reading 4.94 volts at the power supply. The beginning, middle, and ends of the LED strips are reading 4.93 to 4.94 volts when USB cable is disconnected, backglass data cable connected. When USB cable is attached, the power supply begins to buzz and power supply connections are reading about 1.82 volts. Connected 46 leds from the backglass strips (last portion of light strip only, 3 strip sections) and still getting the voltage drop and buzz. Connected 163 lights from the backglass strips (1st continuous light strip only) and still getting the voltage drop and buzz. The buzzing occurs no matter what the teensy is programed to do, even the blink test.

Diagram:
Pinball LED.jpg

Hardware:

Teensy 3.2

OctoWS2811 Adaptor for Teensy 3.2

BTF-LIGHTING AC100-240V to DC5V60A Max300W Suit DC5V1A~DC5V60A Switching Power Supply Device Transformer Power Adapter Converter Jack for WS2812B WS2811 WS2813 SK6812 etc LED Strip/String/LED Matrix

Under Cabinet Lights:
BTF-LIGHTING WS2812B RGB 5050SMD Individual Addressable 16.4FT 60Pixels/m 300Pixels Flexible Black PCB Full Color LED Pixel Strip Dream Color IP30 Non-Waterproof Making LED Screen LED Wall Only DC5V [Using 108 LEDs]

Backglass Lights:
BTF-LIGHTING WS2812B RGB 5050SMD Individual Addressable 16.4FT 60Pixels/m 300Pixels Flexible White PCB Full Color LED Pixel Strip Dream Color IP30 Non-Waterproof Making LED Screen LED Wall Only DC5V [Using 222 LEDs]

Playfield Lights:
BTF-LIGHTING WS2812B RGB 5050SMD Individual Addressable 3.3FT 144(2X72) Pixels/m Flexible Black PCB Full Color LED Pixel Strip Dream Color IP30 Non-Waterproof Making LED Screen, LED Wall Only DC5V [Using 200 LEDs]

Amazon Basics RJ45 Cat-6 Ethernet Patch Internet Cable [Cut in half with portion stripped for use with OctoWS2811]

Connecting USB port to Windows 11 computer.

Software:
Windows 11
Arduino 1.8.16
Teensyduino v1.55
 
could it be a ground loop? Or a power path going thru the USB?

Is the laptop plugged into wall power? if so, disconnect laptop from wall power (run on battery) and disconnect any other cables (monitor, etc) so only the USB to your T3.2 is connected.
 
Well, I am getting an interesting result. I have a separate laptop connected to the teensy and I am able to control the teensy without a buzz on the power supply. I then plugged the laptop into the same outlet as my computer, still no buzz. Measured the power supply and led lights, both reading about 4.94 volts.

I then tried to run the OctoWS2811 basic test example, and nothing : ( Not a single light turns on.
 
I have tried to isolate the issue by using a small strip of LEDs, a different 5v power supply and connecting the teensy to a laptop as a separate test. I suspect I have a bad OctoWS2811.

I can get the teensy to run blink successfully.

For the test, I have connected the LED strip from the far end to the power supply and read 5.2v at the point nearest the power supply and at the opposite end of the strip. The LED strip has 33 RGB LEDs, which I used to set the LED count in the OctoWS2811 basic test. I have the LED strip type set to WS2811_RGB and frequency set to WS2811_800Hz. I uploaded the program to the Teensy and I get nothing. I then tried each of the LED types: WS2811_RBG, WS2811_GRB, & WS2811_GBR. All resulted in nothing happening on the LED strip. I also tried each type with the frequency set to WS2811_400Hz. Still nothing.

I noticed a point in the documentation that states the ground of the power supply should be as close as possible to the signal input. So I connected a second wire to the ground of the power supply and held it in contact to the strip ground at the signal end, then ran each of the types and frequencies again. All I got was nothing.

Has anyone used the specific BTF LEDs I purchased with the OctoWS2811 successfully? I know the LEDs are working, as I have used them with an Arduino Mini running them using FastLED for another project. They run great. I have even used the same power supply used in this test on the Arduino project.

Is there a way to tell if I have a bad OctoWS2811? Any other tests or suggestions?
 
I have the LED strip type set to WS2811_RGB and frequency set to WS2811_800Hz.
The OctoWS8211 Basic Test example has:-

Code:
const int config = WS2811_GRB | WS2811_800kHz;

A typo perhaps?
 
The OctoWS8211 Basic Test example has:-

Code:
const int config = WS2811_GRB | WS2811_800kHz;

A typo perhaps?

Thanks for the thought, tried all configuration options by just modifying the basic test code. Not a typo as everything compiles successfully. Any unknown value for configuration results in compilation errors.
 
Well, ordered a new OctoWS2811 board. Should be in later this week or early next week. Hopefully that solves the problem, as I can not find any other reason for the issue. Will post what I find.
 
Ok, it was a bad OctoWS2811 to start with. Received the new OctoWS2811 board and resolved the issue with the basic test not working. At least now I know my lights can work properly, at least when connected to a separate laptop. Although that has not resolved my issues completely.

I am using a teensy to drive LED lights via direct output framework via a connection to a computer that is used to play pinball games on a pinball cabinet.

The issue is that I am getting a short any time I plug the pinball computer USB to the teensy controller. When plugging in the teensy, the 5v power supply buzzes and the power output drops down to ~1.2 volts.

When the teensy is connected to the pinball computer without connecting to the LEDs, the teensy operates perfectly and no short detected.

When the teensy is connected to a separate laptop (not the pinball computer) while connected to the LEDs, the teensy operates perfectly and no short detected. I am able to run the basic light test successfully and all lights are working, cycling through the different color wipes.

When the teensy is not powered, the 5v power supply is stable putting out 4.98 volts.

So I am able to power the teensy controller using another computer successfully. It only happens with the computer which drives the pinball machine, which is the computer I need connected to operate the teensy using a direct output framework.

The computer which drives the pinball table (not the laptop) and the 5v power supply are plugged into the same surge protector, so I believe they share the same ground with all devices in the pinball table. In fact, all power supplies are plugged into the same power strip. That said, it appears that I still have a ground loop.

Connected to the power strip are the following:

AtGames Legends Pinball pinball table transformer
12 v power supply, large wattage, using a 3 prong plug connected to neutral, hot, and ground
12 v power supply, small wattage, using a 3 prong plug connected to neutral, hot, and ground
5 v power supply, large wattage, using a 3 prong plug connected to neutral, hot, and ground
2x audio amplifiers for SSF each using a non-polarized 2 prong plug
network hub using a non-polarized 2 prong plug
USB hub using a non-polarized 2 prong plug

The computer is connected to the USB port of the AtGames Legends Pinball table, a USB port for the hub, and a "Sainsmart" board.

I was able to test a couple things. I disconnected all other USB devices from the computer other than the teensy with LEDs connected, I still shorted.

I turned one of the audio amplifiers and network hub plugs to change the position of the prongs in the power strip, still shorted. I will have to open up the pinball table to rearrange the other two 2-prong power cables, so unable to test those quickly.

Does the teensy OctoWS2811 need to be grounded to the power supply? Not sure what else to try.
 
Update, I was able to unplug all the non-polarized plugs and test the teensy while connected to the LEDs, still shorted.
 
Another test.

I unplugged the 5v transformer and the computer from the power strip. I hooked both up to an extension cord that you can connect up to three plugs into. Hence, eliminated the all other pinball devices from the common ground and isolated the power transformer and the teensy to one ground. The pinball table was turned off, so no other devices received power. Only the computer and the 5v transformer were powered. Attempted to connect the USB cable and the 5v transformer started to short out.

Clearly not a ground loop, at least I don't think it is, but something with the teensy connection and this computer, but only when connected to the power supply. I am able to control the teensy fine as long as the OctoWS2811 is not connected to the LEDs. Any ideas on how to resolve this?
 
Yeah, I googled the power supply and saw it had a barrier strip for hookup, so a potential area where a user could mess up. But I think if wrong you would be blowing fuses.

From the low voltage and buzzing, I would guess the power supply is shorted and in foldback current limiting. It would still be producing a lot of amps, maybe not 60.
I find it surprising that a usb cable could handle say 40 amps and not get hot. I am not clear how things are hooked up ( is this just the usb cable to the Teensy? ) , but maybe the 5 volt supply for the lights and the 5 volt supply in your computer are fighting each other. Perhaps some experiments with a usb cable with the 5 volt line cut would shed some light on the situation.
 
The 5v power from the transformer is connected to the led strips at the points where the green lines meet the black lines (led light strips) in my diagram. The red lines in my diagram represent the data/neutral only connection between the Octows2811 and the led strips via a cat 6 cable. The computer is directly connected to the Teensy via USB.

I have tried two different USB cables, did not resolve the issue. I read that they recommend injecting power as close to the Teensy as possible, not sure why, as the Teensy is powered by usb. What would happen if I connected a ground between the Octo and the 5v transformer? Would it fry the Teensy? I have to maintain the data feed via usb for the direct output framework, so can’t power it externally.

What baffles me most is that it works fine when connected to my laptop vs the pinball computer. I agree, there is some kind of strange power interaction between the computer power and 5v power supply. Wondering if I put a non powered usb hub in between the Teensy and computer???
 
>>> as the Teensy is powered by usb.

I have never played with these LED strips, but over here https://www.pjrc.com/teensy/td_libs_OctoWS2811.html it says that the Teensy should be powered by the large 5 volt supply and the etch connecting Vin and Vusb should be cut on the Teensy ( and be careful not to cut too deep ).

One other idea I had was to power it all with a GFI outlet to test for any ground issues.

I hope you get it running and post a video to youtube.
 
I agree, there is some kind of strange power interaction between the computer power and 5v power supply.

I have no specific experience with the OctoWS2811, but maybe you could check one thing: have you cut the trace between Vin & Vusb on the Teensy ?? When you connect your USB cable to the Teensy 3.2, you are supplying 5VDC from the laptop/computer to the Teensy thru the USB cable on the Vusb pin. If/when you connect the OctoWS2811 to the LEDs, you may also be supplying the 5VDC from the LED power supply to the Teensy on the Vin pin. This is effectively shorting the two sources of 5VDC together & they are almost certainly *not* putting out the same identical voltage. Most likely, neither the laptop's/computer's USB port nor the external 5VDC power supply care too much for having an external source of 5VDC back-driven into their output. The trace between Vin & Vusb on the Teensy is made to be cut to avoid this exact problem. If you decide to try this, make sure to search the forum for warnings about not cutting too deep when you sever this trace !!

Hope that helps . . .

Mark J Culross
KD5RXT
 
Okay, cut the trace and tested on the laptop, the Teensy no longer has power when connected to usb.

I did a quick check on the pinball computer without connecting power to the Octo/Teensy, interestingly the 5v transformer buzzed and dropped voltage with the power led flickering on the transformer. I was expecting nothing to happen.

I’ll work on connecting the Octo to the transformer next.

The OctoWS2811 has connection point for 5v+ and ground. I am going to hook up the 5v- from the power supply to the ground of the Octo. I hope that is correct???
 
After connecting power from the 5v transformer, the Teensy powers up and is running the program. I then went to connect the USB cable, and the 5v transformer began to buzz and power light dim and flickering.

Any other thoughts?
 
https://m.media-amazon.com/images/I/71R-kx6e5qL._AC_SL1100_.jpg

Hope the image shows. I think this is the supply you are using.
The first 3 terminals are for the power cord, line neutral and safety ground. The safety ground does not connect to anything in your circuit.
The next two are the negative side of the 5 volts. This is connected to your electrical ground on the LED's and on the Teensy ground.
The next two are the positive side of the 5 volts supply. It connects to the Teensy Vin and to the 5 volt line on the LED strips.

Please don't be offended if this is too simple of a description. I am thinking you may have your LED's connected to the AC safety ground of the power supply.
 
With the USB cable near the Teensy using a DMM:

> what is the Voltage between the Teensy USB connector hood and the Cable metal end GND?
> what is the Current between the Teensy USB connector hood and the Cable metal end GND?

With 5V trace cut the cable is providing a rich GND to the Host computer through Teensy.

Measured over an AMP to USB Host once connecting an RGB matrix doing this because the GND to PCB supply from the matrix was not as nice as through the Teensy. Resolved adding a second GND wire from the matrix connector, then the Teensy started programming normally without Button.
 
With the USB cable near the Teensy using a DMM:

> what is the Voltage between the Teensy USB connector hood and the Cable metal end GND?
> what is the Current between the Teensy USB connector hood and the Cable metal end GND?

With 5V trace cut the cable is providing a rich GND to the Host computer through Teensy.

Measured over an AMP to USB Host once connecting an RGB matrix doing this because the GND to PCB supply from the matrix was not as nice as through the Teensy. Resolved adding a second GND wire from the matrix connector, then the Teensy started programming normally without Button.

Voltage measured approximately 40 volts when first contact of usb metal outer to Teensy port metal outer and quickly dropped to 0.

The current read 0.05 amps when touching the usb metal outer with red and the Teensy outer with black at 10 amp DMM setting.

When I contacted the two or completed the circuit between the two, I see a little spark as well.
 
Voltage measured approximately 40 volts when first contact of usb metal outer to Teensy port metal outer and quickly dropped to 0.

The current read 0.05 amps when touching the usb metal outer with red and the Teensy outer with black at 10 amp DMM setting.

When I contacted the two or completed the circuit between the two, I see a little spark as well.

Not an EE here - but, 40 Volts is bad potential - and if connected, bad actual. Perhaps that adds to understanding the nature of the buzzing problem.
 
Back
Top