Hello,

I have a boat where I installed a number of homemade RGB LEDs to act as lights. To control the LED's I purchased the following controller:

http://www.superbrightleds.com/moreinfo/rgb-controllers/led-product/111/

The lights/controller work fine but after connecting it to the boat battery I started hearing a high pitched hum coming from the stereo speakers when the lights were on. If I shut the lights off at the controller, the hum went away. If I directly connect the controller to the battery then the lights work and I do not hear a hum. I have troubleshot the issue down to the controller and I think it is causing a ground loop or putting out a signal on the ground. I have attached a simplistic picture of how things are connected in the boat: stereo/amp, controller share the same ground from the battery.

If I remove the controller and connected the lights to the switch / ground block then the lights work and I do not hear any hum in the speakers. If I put the controller back in, the hum returns only when the lights are on (I.E the power is going in the controller and out on the RGB+ lines to the lights.

The question is what could be causing this? Is it possible for the controller to being sending something out on ground that is being amplified by the amp/stereo? If so, can you think of a way to filter it or stop it?

Thanks

Tim

 

Please clarify the difference between these two statements.

The lights/controller work fine but after connecting it to the boat battery I started hearing a high pitched hum coming from the stereo speakers when the lights were on.

If I directly connect the controller to the battery then the lights work and I do not hear a hum.

Are you bypassing the ground block in one scenario?

A 'hum' is usually considered low-frequency. Do you mean a loud low frequency hum or something more like a whining noise?

 

Thanks for the response. To clarify, if I have it connected like the diagram (controller ground going to ground block, positive going to switch then to positive block) I hear the hum. If I connected the controller directly to the battery (skipping the ground and positive blocks) I do not hear the hum.

By hum I mean a high pitched steady state noise similar to a computer beep but continuous. It only happens when there is load on the controller, meaning if the lights aren't connected to the controller then I do not hear it. If I connect a single light then the noise is there. The controller has a "flashing" setting where the lights will go on/off. When in this mode, the hum also goes on/off. The noise is coming from the speakers (if stereo/amp is off then I do not hear it).

Thanks

Tim

 

I'm thinking that when connected to the ground block, the ground is not solid enough and you are picking up switching noise from the controller due to poor grounding.

Is the battery ground connected only to the ground block ?

Make sure the connections are all good and tight. In a 12V system it doesn't take a lot of resistance in a bad connection to cause troubles.
What size/gauge are the various wires?

 

The power wiring from the controller to the switch / ground was existing as are all of the connections. It is 14gauge wiring that use to run to some incandescent fixtures which I replaced with the homemade led's. I pulled a new 4 conductor 18gauge cable for the led's and terminated it at the controller. So in short the power for the controller is using existing wiring/connections so everything should be tight with no problems. The boat has a large amount of systems on it (AC / DC, bonding blocks, etc). I trust the existing wiring/connections since it worked fine with the incandescents plus if I connect my lights (say blue + positive) to it then the LEDs light up and I do not hear the humming sound. It is only when the controller is connected that the humming occurs. I believe there is some sort of feedback being sent on ground from the controller and the amplifier is picking it up. Would a RF filter work or is there a way to test the ground from the controller? I am thinking I can connect the ground from the controller directly to the battery and use the existing positive connection and that will resolve the problem but I would prefer not to do that.

 

The controller is likely a pwm controller and/or has a switching supply of some kind which means it generates high frequency noise.

The fact that the noise is gone when the controller is connected directly to the battery suggests that there is some resistance between the battery Gnd connection and the ground bar. Any appreciable resistance lessens the effectiveness of any shielding and/or noise filtering circuitry in the controller.
This could be as simple as a corroded or slightly loose connection right at the battery or ground bar

Don't assume existing wiring is good because incandescent lights worked ok. If those were just light bulbs they wouldn't generate noise as would the new controller.

Edit: The LED's when directly connected also won't generate noise.

 

Ok, I can understand that. In case I can't isolate the problem after checking all of the ground bars, battery terminals, etc., can you think of anything else I can do such as build or buy some sort of filter that would reduce the noise being generated by the controller? Do you think running a new ground from the controller directly to the battery will resolve the problem?

Thanks again for all of you help, I realize this difficult to explain over email.

Tim

 

The ground bar should be connected to the battery with a heavy cable (as in your car).
Installing a filter might lessen the noise but it sounds like it went away when you bypassed the ground bar and connected direct to the battery, so it's very likely the problem is the battery to ground bar connection.

 

Hello,

Bringing this thread back up. After playing with this a little more, I do not think the issue is with the ground per say. I ran a new ground wire directly to the battery as well as connecting in various places and the issue remains the same. As previously mentioned there is a switch inline between the battery positive and the PWM controller. That switch is a typical marine rocker switch which is lighted. On the back of the switch there is a common ground for the LED light, and the two positive poles. I noticed that when I turned the circuit on with the radio off I can hear the same humm at the switch if I put my ear right up to it. If I rock the switch a little the humm get's a little louder/quieter. So now I am thinking the switch is actually the problem and possibly the ground that is connected to it for the switch LED. If I run both positive and ground back to the battery (not using the switch) then I no longer have the problem.

This brings me back to my original question, there must be some sort of isolator circuit that I can put inline between the switch and the PWM controller. I am thinking something like a cheap dc/dc power supply or building a isolator. Think this would work?

Thanks

Tim

 

Just try putting a large cap (2200uF) directly at the light between the + and - terminals. Most of those lights have very limited internal caps (to make them cheap) and draw power in high frequency pulses down the power wiring.

Adding a big cap right at the light will make it draw power as mainly a smooth DC down all the power wiring and will stop the speaker hum (ie makes it like running form a DC battery as in your earlier test).

 

Possibly a .1uf ceramic capacitor in parallel at the large capacitors terminals to catch any high frequency harmonics.

 

Thanks for the response. When you say put a cap at the light, do you mean at the PWM or at the internal LED in the switch (between the LED ground and positive coming from the battery)? It looks something like this:

battery---switch(lighted)----PWM controller-----RGB lights (6 total)

Thanks

Tim

 

The cap(s) need to be added accross the two power inputs to the PWM controller device.

If the switch is very close to the PWM controller you can put the cap + before the switch so the switch does not need to carry the cap inrush current when it is turned on.

 

Try adding a capacitor across your switch. Your switch is humming like a transformer can hum under load. Considering it is faint in volume and high in frequency, a 0.1 uF ceramic cap shoud do fine. Caps across switches are common.

 

Hello,

Bringing this thread back up. After playing with this a little more, I do not think the issue is with the ground per say. I ran a new ground wire directly to the battery as well as connecting in various places and the issue remains the same. As previously mentioned there is a switch inline between the battery positive and the PWM controller. That switch is a typical marine rocker switch which is lighted. On the back of the switch there is a common ground for the LED light, and the two positive poles. I noticed that when I turned the circuit on with the radio off I can hear the same humm at the switch if I put my ear right up to it. If I rock the switch a little the humm get's a little louder/quieter. So now I am thinking the switch is actually the problem and possibly the ground that is connected to it for the switch LED. If I run both positive and ground back to the battery (not using the switch) then I no longer have the problem.

This brings me back to my original question, there must be some sort of isolator circuit that I can put inline between the switch and the PWM controller. I am thinking something like a cheap dc/dc power supply or building a isolator. Think this would work?

Thanks

Tim

If removing the switch fixes the problem, or rocking the switch changes the noise I think you may have corrosion/excess resistance in the switch.
Are you drawing more current than the switch rating???

I would try a new switch.

 

Spent some more time playing with this. I tried adding a .1uf cap across the switch and it didn't help. I then put a 2200uf cap on the + and - of the controller (the -/+ of the incoming power from the battery/switch). This also did not help with them hum, and after a few minutes the capacitor blew. Just to reiterate some of the things I have done:

- Removed the switch by connecting the switched + and hot + together. The hum was still present, so not a problem with the switch.
- Removed the ground from the switch, hum still present
- Tried the capacitors, hum still present
- Removed all of the lights from the controller, no hum
- Added the lights back one at a time, with a single light the hum is barely noticable but it is there. As more are added, it gets louder. Total AMP draw of all the lights is 3.4amp's. The controller is rated for 4amps.
- Ran a new ground directly to battery, but used the existing + wire that goes to the switch, hum still present

Any other ideas?

Thanks

Tim

 

Why did the cap blow? What voltage was it, and/or did you connect it in reverse?

If the cap was a 25v type on a 12v DC supply, and connected the right way around it should not blow. The only thing that would blow it then is if the light is drawing massive peak current in pulses. Which is also possible and just means you need a better cap.

 

Inline some toroid coils used for SMPS.

Try both the stereo and the LED circuit.

Common mode chokes are also worth trying these have 5mH.

Normal storage choke could be 330uH or something.

 

If the LED controller is drawing all its power in high current pulses adding choke will cause a big reduction in average current and the lighting performance will suffer.

The first thing is to find why the cap blew, and then do a test with a cap which works. Once that is sorted, a choke might be added BEFORE the big cap, but may not be needed.

 

Easiest way: Run the controller's power and ground all the way to the battery.

If not supplied by a properly-placed capacitor, the fast-changing currents will induce spiky voltages across the inductances (and resistances) of the wires themselves, even if the wires and connections are all perfect. The wire from the ground bar to the battery will have those voltages induced across it, and those voltages will then bounce the ground voltages of every other system that is connected to the ground bar.

Installing decoupling capacitance at the controller (between power and ground) "should" work, as long as the distances from the power and ground wires to the capacitors are very, very short. I would try a small-lead-spacing 1uF film capacitor in parallel with a hundred uF electrolytic and see what happens. If it's not enough, you could add high-current chokes in series with both wires, on the battery side of the capacitors, and another set of two capacitors on the battery side of the chokes. You could also increase the values of the capacitors.