LED PWM Dimmers: Unintended consequences....

[RayS]

I just finished a 2-week project in the sim, which now enables me to control all the cockpit lighting via dimmers using the various controls in the overhead panel for my Beech 1900D.

The panels are from Eric Tomlin and are perhaps the best backlit panels on the market.  http://www.flightlinesimulations.com/

After bolting everything back together I fired up the Beech, turned up all the lighting, put on my DC headset, and was immediately greeted with what I can only describe as 'Digital nails on a chalkboard'.

The lighting project uses 6 PWM LED dimmer controls, all powered by the on-board 12VDC Avionics Bus. Knowing I couldn't live with the noise for more than 10 seconds, I decided to isolate  the 12VDC Bus, and put the dimmers on their own 12v Bus.

No joy......

It was only after fully isolating both the DC bus and the ground bus from everything else, did the hum go away.

In all, I rewired the lighting system 3 times to get to a hum-free state.

Word to the wise: If you use, or are thinking of using a lot of cheap, 'Made in China' PWM LED dimmers in your project, you probably want to fully isolate the entire circuit from everything else.

[ETomlin]

Ray, thank you very much for your kind words.

I'm glad you posted on this subject of PWM lighting. I've never experienced what you have covered here, but I will agree that after talking with others that are far better electronics engineers than myself, PWM dimmers for LEDs are not too good sometimes.

Most of the LJ45 builders (including myself before selling that sim) use a PWM dimmer from ebay that is most likely like the one you used (and multiples of them). I had VATSIM and headsets in my sim and like I said above, never had this issue, but I do not know why.

I powered my own set of back-lit panels from a dedicated 12v power supply that had multiple dimmers and multiple panels attached. It was plugged into a power strip, then plugged into the wall. My headset was a cheaper one from Radio Shack that plugged into a long 1/8" extension that ran into the mic input on the sim PC. How I had no issues, I do not know.

This is what I used for PWMs and power supply:

PWM- http://www.ebay.com/itm/PWM-dimmer-switch-for-12V-DC-LED-and-panel-lights-6-amp-/131606600545?hash=item1ea45d5b61

Power- http://www.ebay.com/itm/AC-To-DC-12V-20A-240W-Switching-Power-Supply-Driver-Switch-for-LED-Strip-Light-/321752594336?hash=item4ae9f2f7a0

[RayS]

I originally started with one of these units:
http://www.ebay.com/itm/DC-12V-8A-LED-Light-Protect-Strip-Dimmer-Adjustable-Brightness-Controller-F5-/291548909081?hash=item43e1ab4e19


It is the only one still tied to the main power bus and I haven't had any issues with it. I guess over time, they changed the circuitry and introduced the hum coming from the PWM circuit. (Or that six is way too much hum and it was a cumulative effect)

Regardless, the PWM unit you are using looks like a better build quality and a more mature design. I cheaped-out and bought the bargain-basement builds. You'd think I'd learn my lesson after so many years in this hobby!

[727737Nut]

Love this line, "I fired up the Beech"   

The reason you has the problem is that most PWM dimmers/ speed controls PWM the Ground side of the circuit.  Low side PWM is cheaper and easier to build than High side dimmers.  So your PWM frequency was being shoved into the ground circuits which wrecked havoc on the mic and headphones and speakers that are ground critical.  Sucks having to buy high end dimmers but it is the only way unless you isolate every circuit having anything to do with your backlighting.   Ground Loops also come into play but most don't notice it do to minimal items in their pit.  The more complicated it gets the more important proper grounding and quality parts are in the build. You start mixing real parts and aftermarket parts look out! Trouble will come quick if you are not careful.

Always look for HIGH Side dimmers when shopping.  If you find any under 50.00 let me know. Please.

Rob

[ETomlin]

I originally started with one of these units:
http://www.ebay.com/itm/DC-12V-8A-LED-Light-Protect-Strip-Dimmer-Adjustable-Brightness-Controller-F5-/291548909081?hash=item43e1ab4e19


It is the only one still tied to the main power bus and I haven't had any issues with it. I guess over time, they changed the circuitry and introduced the hum coming from the PWM circuit. (Or that six is way too much hum and it was a cumulative effect)

Regardless, the PWM unit you are using looks like a better build quality and a more mature design. I cheaped-out and bought the bargain-basement builds. You'd think I'd learn my lesson after so many years in this hobby!

I do actually have 2 of the units you posted, but those are only on the wall for testing panels out before they get shipped. Because they're not connected to anything else, I'd never know if they produce hum.

However, as Rob pointed out, be careful about letting these that I use touch anything bare or each other because I did have the issue where I had ground problems because the dimmers are placed so close to each other behind the panel. I went back and put some black electrical tape between them to prevent them from touching each other.

[KyleH]

Eric, the reason you didn't have problems is because you used an isolated power supply for the PWM controlled lighting. This is what Ray learned the hard way.

The switching currents from the PWM (which I assume are in the audible frequency range or close to it) that are sharing a ground with other equipment will cause issues with that equipment. It is likely that there was enough power in the circuits that Ray's audio was picking up the ground currents, especially if the power for his headset was on the same bus. The multiple frequencies due to the several dimmers probably added to the issues and maybe caused some harmonics in the auditable range.

High/Low side drive has nothing to do with it, the switching currents will be there regardless of where the switch is placed. All of our products (high quality LED drivers) are isolated low side drive and are designed to meet FCC standards.

[727737Nut]

I wasn't going to reply but I feel like the correct information needs to be here for everyone.  Kyle, no disrespect but you are not 100% correct. Not wrong either
When it comes to using 'real' parts, you will never find a low side dimmer. (and yes, high side vs low side makes a HUGE difference) Simple google search will give you 100000 pages of why and how. 

Real parts share a common ground, therefore any dimming must be done on the positive side of the circuit.  When properly designed there is no noise on either side.  I have 8 power supplies and 26vac and 115vac 400Hz power in my sim, all sharing the same ground within the sim and my house. If you take a meter and put it on any metal part in my sim you will read ac voltage. But, since there is one common ground plane shared with all power supplies, metal parts and house earth ground, no noise is present.  With a cheap PWM module that wants to PWM the gnd to your lighting, in my case, common gnd remember, it doesn't work. I have to run a high side or positive side dimmer.

It comes down  to using quality parts designed for the application you are using and matching to environment they are in. 

Believe me, i spent many hours in the beginning chasing ground loops and noisey and or inoperative circuits.  Humbles you real quick! LOL Lots of late night studying, reading and researching got me set straight.  I am by no means an electrical engineer but I have allot of training in aircraft specific wiring. 

Rob

[KyleH]

Rob I respect your experience, and I also only posted because I also want correct information to be available. I checked with our analog EE this morning to be sure I have my facts straight.

Suffice it to say we are both correct, we're really just talking about different applications.

My concern is with the generalization that only High side drivers should be used and are of high quality.
Low side drives can be of high quality, and deliver good results, they just have to be installed correctly.

It comes down  to using quality parts designed for the application you are using and matching to environment they are in. 

This is the important part. We all have different setups, and many use isolated power supplies that do not use the metal shell for all grounding.

In your setup as you've described, yes you need a properly designed high side drive. But this doesn't  necessarily follow for everyone's setup. My comments were with respect to Ray's application which I'm assuming is an isolated power supply to a bus that he has setup.

The switching currents from the load (not the driver) being switched will be in the circuit will exist in the ground path until it gets back to negative side of the source, the power supply. Any equipment that shares this bus could be susceptible to these currents, particularly the audio equipment.
In this particular case the fact the load was being switched and sharing power buses with the audio caused the issue. Changing the driver to be high side wouldn't have helped. Making a star-point ground at the power supply would help.

Of course I'm making assumptions on his setup here so I may stand corrected.



Ray, what I really want to know is; how's the project of turning the shaver into a linked yoke system coming along?

[RayS]

...way to fork the thread, Kyle! 

However, to answer your inquiry, and to update everyone on my effort to convert an electric shaver into a dual-control yoke system for the Beech, I ran into a slight (slight) problem.

It seems that I need a 2-head shaver, as opposed to the 4-head unit that was shipped from Alibaba. (<---hint)

Once all the kinks are worked out, I should have a working dual-linked system that will also double as a men's grooming station. I was going to install a simple mirror for ease of use, but... go big or go home as the saying goes.. I splurged and went with a nice sheet of 60/40 beam-splitter glass so it could pull double-duty as a HUD.

Now I will arrive well-rested AND clean-shaven.



 

[Trevor Hale]

It seems that I need a 2-head shaver, as opposed to the 4-head unit that was shipped from Alibaba. (<---hint)

Ok I thought I read that right when Kyle Said it..  Now I am seeing you Say it, and I may be a dumb Canadian here...  But WTF is a Shaver going to help you with on a Yoke, other than the "STICK SHAKER"

OH NEVERMIND...  We all know you like to use stick shakers

[727737Nut]

Kyle, very well put 

Ray, could you bring that to the WF2015? 

Also, off topic, I'm bringing a servo and Teensy and I would like you and I and Trevor and anyone else interested, to figure out the issue I have with jumpy servo movements. 

Rob

[Trevor Hale]

Lol we can try. I don't have any jitters except when I first start the sim. All my servos are noisy until they start getting information. With the exception of my fuel gauges which feel the need to go to empty before updating to the new value randomly.

I would be happy to help you providing we get time. Absolutely.

[ETomlin]

Ok I thought I read that right when Kyle Said it..  Now I am seeing you Say it, and I may be a dumb Canadian here...  But WTF is a Shaver going to help you with on a Yoke...

I believe he ordered some kind of hardware component that was to be used to linking his yokes together but somehow Alibaba sent him a shaver instead. LOL.

[daveaust]

If you switch the low side of the circuit, guess what, it isn't truley grounded during the "off" part of the duty cycle. I was thinking about how to place a noise bypass cap from the high side to ground but I think that would cause issue for 2 reasons. I am getting ready to connect my intercom and saw this thread so I may be in same boat. SOunds like I will need isolate via separate PS according to your experiences. Thank you in advance!