[sdiy] Will LCD backlight PWM cause audio noise?

[Veronica Merryfield]

The PWM signal from the transistor, i.e. the high current PWM. With this filtered close to the transistor, the trace and/or cable to the LCD is not acting as an antenna since it now has DC on it rather than PWM.

Decouple the supply close to the transistor.

Flood filling the PCB with ground and the use of a 4 layer PCB plus being mindful of separation in space and power layout will see you ok.

One of my early projects many moons ago, somewhat naively, observed some of these but was a 2 layer no flood fill. I had all the SMPSUs in one place with a guard area. The audio out had a squeal at high-E. I flood filled the next revision but kept to two layer and it was fine. These days I use four layers.

I am assuming this is a DIY product.  

 
On 2012-08-19, at 5:58 AM, Tom Wiltshire wrote:

> Thanks Veronica (and others), but one thing I don't understand;
> 
> What do you mean with "post switch lowpass RC filtering"? Do you mean you filter the PWM output? Or you filter the power supply after the switch to prevent any noise it generated from getting elsewhere?
> 
> I'd thought that a separate power regulator for the backlight was a good idea, and I was going to try and keep its wiring away from everything else. As you say the PWM is at a constant frequency, and if I'm willing to accept less resolution in my backlight control (64 levels of backlight brightness is a lot, right?!) then I could push the PWM frequency very high (>300KHz). I don't know if that helps, but I suppose it makes filtering easier.
> 
> Tom
> 
> 
> On 17 Aug 2012, at 17:18, Veronica Merryfield wrote:
> 
>> Hi Tom
>> 
>> I have a couple of pro products out there with PWM control on a backlight. I use different rails for audio verse non-audio, I use local decoupling at the PWM switch with post switch LP RC filtering (any leads to the LCD do not therefore act as radiators), I use ground flooding on the PCB with plenty of flood round the switch and I try to keep the audio and non-audio separated on the PCB, often with a separation barrier if I can.
>> 
>> Straight PWM at an elevated frequency is reasonable to deal with since it is constant. SMPSUs chips that say they switch at high frequencies but then have long controlled off times under light load conditions play havoc since the actual frequency drops to audio. THe above techniques work however in these cases, so your constant 78KHz should be ok with the right care taken.
>> 
>> Veronica
>> 
>> On 2012-08-17, at 9:01 AM, Tom Wiltshire wrote:
>> 
>>> Hi All,
>>> 
>>> If I control the brightness of an LCD backlight using PWM, is it likely to cause a lot of audio noise?
>>> 
>>> I was planning to have a PIC PWM channel driving a transistor to switch the backlight on and off. This could be well above audio frequencies, say 78KHz. However, it struck me that switching a high current load (30-40 mA) on and off rapidly right next to a load of sensitive audio circuits wasn't the brightest idea I'd ever had.
>>> 
>>> What do you think? Will it cause me problems or will I get away with it?
>>> 
>>> Thanks,
>>> Tom
>>> 
>>> _______________________________________________
>>> Synth-diy mailing list
>>> Synth-diy at dropmix.xs4all.nl
>>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>>> 
>> 
> 
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>