Matronics Email Lists :: View topic

Matronics Email Lists
Web Forum Interface to the Matronics Email Lists

Get Email Distribution Too! FAQ Search Memberlist Usergroups Register

Profile Log in to check your private messages Log in

LED Ghosting
Goto page Previous 1, 2, 3, 4, 5 Next

Matronics Email Lists Forum Index -> AeroElectric-List

View previous topic :: View next topic

Author Message

donjohnston

Joined: 13 Dec 2009
Posts: 231

Posted: Sat Jun 21, 2014 1:49 pm Post subject: Re: LED Ghosting

Okay... The MOSFET approach isn't going to work.   



The LEDs and resistors are on a circuit board. This type of change isn't going to work given the constraints of the board. 



So I guess I'll have to fall back to my plan B and add resistors to dim all the other LEDs so they're about the same as that pesky engine analyzer. 



Maybe Bob will have an idea.

user9253

Joined: 28 Mar 2008
Posts: 1908
Location: Riley TWP Michigan

Posted: Sat Jun 21, 2014 1:55 pm Post subject: Re: LED Ghosting

 	  Quote: 	
		  Unfortunately, I am way outside my element. Not quite sure how to wire this sucker now. 	


Not to worry, use Eric's (edpav8r) last post as a guide and draw a new circuit.  Post it and others will offer suggestions.

  Diodes added to the bright LED circuits is starting to look easier, if it works.  I think that it is worth a try.  Two of 1N4148 in series will drop about 1.5 volts when conducting 20 ma and will drop 1.6 volts at 30 ma, according to the data sheet.  That will equal the voltage drop across the Engine Analyzer Alarm and hopefully make all LEDs the same brightness.  I have some extra diodes that I could mail to you if you wanted to try them.

  Of course you already have the P-Channel MOSFET.  Use this circuit

http://electronics.stackexchange.com/questions/3599/basic-p-type-mosfet-question and ignore the N-Channel circuit.  Substitute your dimmer for the power supply, use your 390 ohm for the load resistor.  Put your LED in series with the 390 ohm load resistor.  Connect the Engine Analyzer Alarm output and the pull-up resistor to the gate.

_________________
Joe Gores

user9253

Joined: 28 Mar 2008
Posts: 1908
Location: Riley TWP Michigan

Posted: Sat Jun 21, 2014 2:02 pm Post subject: Re: LED Ghosting

Dan,

 I would try adding 2 diodes in series instead of adding resistors.  The diodes are very small, maybe even smaller than the resistors that you have.  They should fit on the circuit board.

_________________
Joe Gores

raymondj(at)frontiernet.n
Guest

Posted: Sat Jun 21, 2014 2:09 pm Post subject: LED Ghosting

Instead of dimming all the other ones.  Can the problem one be replaced 

with a brighter one so that in it's lower than maximum light output 

mode, which the current one is apparently in, it is the same brightness 

as the others?



Raymond Julian

Kettle River, MN



The things we admire in men, kindness and generosity, openness, honesty, 

understanding and feeling are the concomitants of failure in our system. 

And those traits we detest, sharpness, greed, acquisitiveness, meanness, 

egotism and self-interest are the traits of success. And while men 

admire the quality of the first they love the produce of the second. 

-John Steinbeck, novelist, Nobel laureate (1902-1968)



On 06/21/2014 04:49 PM, donjohnston wrote:

 	  Quote: 	
		   



 Okay... The MOSFET approach isn't going to work.  



 The LEDs and resistors are on a circuit board. This type of change isn't going to work given the constraints of the board.



 So I guess I'll have to fall back to my plan B and add resistors to dim all the other LEDs so they're about the same as that pesky engine analyzer.



 Maybe Bob will have an idea.





 Read this topic online here:



 http://forums.matronics.com/viewtopic.php?p=425221#425221

edpav8r(at)yahoo.com
Guest

Posted: Sat Jun 21, 2014 8:11 pm Post subject: LED Ghosting

Laying in bed thinking about this and couldn't sleep...

On Jun 21, 2014, at 4:49 PM, "donjohnston" <don(at)velocity-xl.com> wrote:

 	  Quote: 	
		   Okay... The MOSFET approach isn't going to work.   

 

 The LEDs and resistors are on a circuit board. This type of change isn't going to work given the constraints of the board.


	


Don, I presume this is because your board has a common trace that feeds dimmer output to all of the cells in the annunciator panel, with each cell separately switched to ground for activation.  If so, could you use a razor blade to cut the PCB trace for this "dimmer bus" on either side of the oil pressure cell, install a jumper wire to bypass this cell, then feed the oil pressure cell directly from the MOSFET drain?



Eric

donjohnston

Joined: 13 Dec 2009
Posts: 231

Posted: Sun Jun 22, 2014 4:22 am Post subject: Re: LED Ghosting

Here's (what I think is) the correct circuit. Still not sure I can integrate this into the existing PCB though.

LED circuit 3.pdf

Description:

Download

Filename:

LED circuit 3.pdf

Filesize:

29.1 KB

Downloaded:

479 Time(s)

user9253

Joined: 28 Mar 2008
Posts: 1908
Location: Riley TWP Michigan

Posted: Sun Jun 22, 2014 4:46 am Post subject: Re: LED Ghosting

Your LED circuit 3.pdf looks correct to me.

_________________
Joe Gores

donjohnston

Joined: 13 Dec 2009
Posts: 231

Posted: Sun Jun 22, 2014 5:57 am Post subject: Re: LED Ghosting

Definite improvement! Probably usable.  There's a small range where the MOSFET controlled LEDs go out before the others. But it's such a small range that I doubt it's an issue. 



Now I have to see if it's a doable mod to the PCB.

rickofudall

Joined: 19 Sep 2009
Posts: 1392
Location: Udall, KS, USA

Posted: Sun Jun 22, 2014 6:07 am Post subject: LED Ghosting

Paul, IMHO the variance in brightness would be a distraction the pilot would not need to deal with when any of the warning LED's started lighting. Even if you know that brightness is not an issue the difference would draw the eye and the thinking away from the real problem. For that reason I would want uniform brightness, but as I said, it's just my opinion. 



Rick Girard

do not archive



On Thu, Jun 19, 2014 at 3:15 PM, Fisher Paul A. <FisherPaulA(at)johndeere.com (FisherPaulA(at)johndeere.com)> wrote:

 [quote]--> AeroElectric-List message posted by: "Fisher Paul A." <FisherPaulA(at)johndeere.com (FisherPaulA(at)johndeere.com)>

  

 I'm confused by this thread.  Back to the original issue - why is it important that all of the LEDs in the annunciator panel have the same brightness?  Don't you just need to know if any particular one is on or off?  Aren't they used to indicate a problem? I would hope that during most flights they would all be off.  And if more than one of them was on, I certainly wouldn't be worried about one being brighter or dimmer than another - I'd be busy looking for a place to land!

  

 I do appreciate the electronics discussion about the physics of getting them to the same brightness.  I just don't understand why that is important in this use case.

 

 Paul A. Fisher

 RV7A N18PF

 

 --

_________________
The smallest miracle right in front of you is enough to make you happy....

edpav8r(at)yahoo.com
Guest

Posted: Sun Jun 22, 2014 6:28 am Post subject: LED Ghosting

Don,



New diagram looked perfect.  Glad to hear it works acceptably well.  Hope you can bodge it into your PCB for a permanent solution!



Eric

On Jun 22, 2014, at 8:57 AM, "donjohnston" <don(at)velocity-xl.com> wrote:

 	  Quote: 	
		   Definite improvement! Probably usable.  There's a small range where the MOSFET controlled LEDs go out before the others. But it's such a small range that I doubt it's an issue. 

 

 Now I have to see if it's a doable mod to the PCB.


	


do not archive

donjohnston

Joined: 13 Dec 2009
Posts: 231

Posted: Sun Jun 22, 2014 7:06 am Post subject: Re: LED Ghosting

 	  edpav8r(at)yahoo.com wrote: 	
		  Don,



New diagram looked perfect.  Glad to hear it works acceptably well.  Hope you can bodge it into your PCB for a permanent solution!



Eric 	




Thanks. Looks like I may be able to mod the board (I wish it didn't cost so much to get these small one-off boards made   ).



I was worried that I was going to run into the same problem with the LR3C voltage regulator low voltage warning.  But B&C apparently did it right.  A 1K resistor across the LEDs eliminates the ghosting but doesn't dim the LEDs when on. WooHoo!    



-Don

nuckolls.bob(at)aeroelect
Guest

Posted: Sun Jun 22, 2014 8:05 am Post subject: LED Ghosting

Your voltage measurements confirmed that the annunciator

   output from your engine analyzer is decidedly not capable

   of driving an incandescent bulb . . . in fact, it appears

   that this particular output is not well thought out. Have

   you had any contact with the factory about this? What

   brand and model analyzer are we talking about. Are the

   installation manuals available on line?



  	  Quote: 	
		  Thanks. Looks like I may be able to mod the board (I wish it didn't cost so much to get these small one-off boards made   ). 	


   [img]cid:.0[/img]



   This is the best alternative for 'buffering' the

   output port as it was handed to you. The size of

   the resistor still leaves some risk for 'ghosting';

   there might be sufficient leakage in their output

   to produce a source-gate voltage that is greater than

   the threshold voltage on the FET . . . typically

   on the order of 2-3 volts.



   A resistor on the order of 470 ohms is suggested.



 

  	  Quote: 	
		  I was worried that I was going to run into the same problem with the LR3C voltage regulator low voltage warning.  But B&C apparently did it right.  A 1.5K resistor across the LEDs eliminates the ghosting but doesn't dim the LEDs when on. WooHoo!    	


   That incandescent lamp driver was designed by yours

   truly about 30 years ago for some devices being 

   offered to Cessna and Beech . . . it got folded into

   the B&C products when I did their early designs.



   But nobody was driving LEDs back then. Turn-off

   resistors had to be added to accomodate the

   conversion.



   The FET buffer is the obvious 'fix' in this instance

   but should not have been necessary.



 

   Bob . . .

1f3471a0.jpg

Description:

Filesize:

16.55 KB

Viewed:

13300 Time(s)

nuckolls.bob(at)aeroelect
Guest

Posted: Sun Jun 22, 2014 8:14 am Post subject: LED Ghosting

At 08:57 AM 6/22/2014, you wrote:

 	  Quote: 	
		  



Definite improvement! Probably usable.  There's a small range where 

the MOSFET controlled LEDs go out before the others.


	


   The Perhelion dimmer is linear as opposed to duty-cycle

   switched so yes, at some lower voltage, the gate-turn-on voltage

   for the FET becomes a significant part of the supply

   voltage out of the dimmer and LEDs controlled by the FET

   will go out first.



   A refinement of the design could consider duty-cycle

   switched dimming or going to logic-level gate FET with

   a much lower threshold voltage.



   This has been an excellent study in the relative chaos

   that can arise from poorly tested products. We were

   recently discussing how poorly designed tachometer

   circuits for sampling ignition signals can add risk

   affecting engine performance.



   This "ghosting" problem is a case where less than passing

   grades on a manufacturer's homework caused a lot of unnecessary

   work-arounds that are STILL less than ideal.



   Bob . . .

donjohnston

Joined: 13 Dec 2009
Posts: 231

Posted: Sun Jun 22, 2014 11:08 am Post subject: Re: LED Ghosting

 	  nuckolls.bob(at)aeroelect wrote: 	
		  Your voltage measurements confirmed that the annunciator

   output from your engine analyzer is decidedly not capable

   of driving an incandescent bulb . . . in fact, it appears

   that this particular output is not well thought out. Have

   you had any contact with the factory about this? What

   brand and model analyzer are we talking about. Are the

   installation manuals available on line? 	




But... It does work with an incandescent bulb...  As long as it's wired to 12v and not a dimmer.



The Engine Analyzer is a Grand Rapids EIS 6000.

Doc's are here:

http://grtavionics.com/manualseis.html

nuckolls.bob(at)aeroelect
Guest

Posted: Sun Jun 22, 2014 2:43 pm Post subject: LED Ghosting

At 02:08 PM 6/22/2014, you wrote:

  	  Quote: 	
		  --> AeroElectric-List message posted by: "donjohnston" <don(at)velocity-xl.com>



 

 nuckolls.bob(at)aeroelect wrote:

 > Your voltage measurements confirmed that the annunciator

 >    output from your engine analyzer is decidedly not capable

 >    of driving an incandescent bulb . . . in fact, it appears

 >    that this particular output is not well thought out. Have

 >    you had any contact with the factory about this? What

 >    brand and model analyzer are we talking about. Are the

 >    installation manuals available on line? 



 

 But... It does work with an incandescent bulb...  As long as it's wired to 12v and not a dimmer. 	


    ??? What size of bulb . . . how many milliamps

    does the bulb draw?



    The voltages you observed in circuit 2 say

    that with JUST the two LEDS and their companion

    resistor attached to the engine analyzer, all

    measured voltage drops add up to the supply

    voltage. This means that for that current level,

    the engine analyzer will pull the lamp port 

    nearly to ground.  I believe the resistor was

    390 ohms . . . so load on the engine analyzer

    in case #1 was 8.07/390 = 20.6 milliamps.



    You added a resistor across the array in case

    #3 which says that voltage across the 390 ohm

    resistor dropped to 6.55 volts . . . so indeed,

    current through the lamps went down to 6.55/390

    equals 7.2 millimaps . . . approximately 1/3

    that of case #1. Further, you cite 10.7 volts

    across the 1K resistor for an additional

    10.7 milliamps of load on the engine analyzer

    lamp port . . . for a total of 17.9 milliamps

    which is less than case #1. Yet you LOST some

    voltage in the engine analyzer. Total voltage

    drop in the lamps and resistor is 10.65 volts

    with a system voltage of 12.37 . . . or 1.72

    volts went missing in the engine analyzer's 

    lamp driver. In case #4 we see the same thing

    happening . . . with the disappearance of about

    1.5 volts.



    The output switching device should not have that

    much loss . . . I would expect tens of millvolts

    tops. The voltages say yes, the LEDs are indeed

    biased up significantly lighter when the anti-

    ghosting resistor is added. But it's not clear

    why. 



 

  	  Quote: 	
		  The Engine Analyzer is a Grand Rapids EIS 6000.

 Doc's are here:

  http://grtavionics.com/manualseis.html



 	


   Okay, the warning light output is described thusly:



 [img]cid:.0[/img]



   The fact that it can 'shut down' automatically implies

   some kind of protective 'smarts' on this pin. At the same

   time, the rated switching current is more than

   adequate to the task of handling either LED or incandescent

   lamps . . . although the 'ghosting' phenomenon is not

   understood.



   Have you talked with GR about this? This still not

   making sense.



 

 

   Bob . . .

209dcdfc.jpg

Description:

Filesize:

66.63 KB

Viewed:

13285 Time(s)

donjohnston

Joined: 13 Dec 2009
Posts: 231

Posted: Sun Jun 22, 2014 3:46 pm Post subject: Re: LED Ghosting

 	  nuckolls.bob(at)aeroelect wrote: 	
		  ??? What size of bulb . . . how many milliamps does the bulb draw?



The voltages you observed in circuit 2 say that with JUST the two LEDS and their companion resistor attached to the engine analyzer, all measured voltage drops add up to the supply voltage. This means that for that current level, the engine analyzer will pull the lamp port  nearly to ground.  I believe the resistor was 390 ohms . . . so load on the engine analyzer in case #1 was 8.07/390 = 20.6 milliamps.



You added a resistor across the array in case #3 which says that voltage across the 390 ohm resistor dropped to 6.55 volts . . . so indeed, current through the lamps went down to 6.55/390 equals 7.2 millimaps . . . approximately 1/3 that of case #1. Further, you cite 10.7 volts across the 1K resistor for an additional 10.7 milliamps of load on the engine analyzer lamp port . . . for a total of 17.9 milliamps which is less than case #1. Yet you LOST some voltage in the engine analyzer. Total voltage drop in the lamps and resistor is 10.65 volts with a system voltage of 12.37 . . . or 1.72 volts went missing in the engine analyzer's  lamp driver. In case #4 we see the same thing happening . . . with the disappearance of about 1.5 volts.



The output switching device should not have that much loss . . . I would expect tens of millvolts tops. The voltages say yes, the LEDs are indeed biased up significantly lighter when the anti-ghosting resistor is added. But it's not clear why. 



Have you talked with GR about this? This still not making sense.

   Bob . . . 	




The bulb I used is a bulb that I got from B&C with the LR3C.  I hooked it up just to see if it would behave properly (and it did).  If you want, I can check the specs on it. 



I can't speak to or dispute the numbers or calculations that you present as that's not my area of expertise.  I can only say with certainty what happens when the switch is thrown. 



I have spoken to GRT as I mentioned in the original post. They said the alarm output is designed for an incandescent bulb and that to make it work with an LED I would have to put a 1.5K resistor across the LED.  Now I didn't explain complex (elaborate, convoluted, bizarre, whatever term you wish to use to describe my annunicator panel) nature of the warning light I had. Only that I was using an LED.  They would have assumed that it was a single LED powered by a constant power source. 



I appreciate all the help and suggestions from everyone. I wish that this would have been a simple fix but it would seem that something is going on inside the engine analyzer that defies explanation.  



-Don

nuckolls.bob(at)aeroelect
Guest

Posted: Sun Jun 22, 2014 5:49 pm Post subject: LED Ghosting

 	  Quote: 	
		  

I appreciate all the help and suggestions from everyone. I wish that 

this would have been a simple fix but it would seem that something 

is going on inside the engine analyzer that defies explanation.


	


   Actually, it's very explainable . . . given all

   the facts. Can you give me a name/number at GRT?



   Bob . . .

donjohnston

Joined: 13 Dec 2009
Posts: 231

Posted: Sun Jun 22, 2014 6:47 pm Post subject: Re: LED Ghosting

Phone number for GRT is on their contact page.



http://www.grtavionics.com/contact.html



I don't recall who I spoke to about this.

user9253

Joined: 28 Mar 2008
Posts: 1908
Location: Riley TWP Michigan

Posted: Mon Jun 23, 2014 5:27 am Post subject: Re: LED Ghosting

Don,

Good going getting your project working to your satisfaction.  

The readers of this forum would love to see a picture of your finished circuit board and display.

Joe

_________________
Joe Gores

donjohnston

Joined: 13 Dec 2009
Posts: 231

Posted: Mon Jun 23, 2014 5:59 am Post subject: Re: LED Ghosting

Details on the annunicator panel are here:



http://www.velocity-xl.com/blog/2013/08/03/13-99-annunicator-panel/



http://www.velocity-xl.com/blog/2013/08/15/13-99-annunicator-panel-3/

Display posts from previous:

	Matronics Email Lists Forum Index -> AeroElectric-List	All times are GMT - 8 Hours Goto page Previous 1, 2, 3, 4, 5 Next
Page 4 of 5

Jump to:

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum