LED Failure? Partly solved but not quite!

[Bealman]

I'm  getting a  bit annoyed with all this and I tend to think Peter might be too.

LEDs are lights, for hecks sake. They are diodes that just happen to turn electrical energy into light, with DC current  going in the correct direction.

We can always  go back to incandescent light indicators, with no polarity problems but needing a bank of relays to switch 'em.

George

[austinbob]

I'm  getting a  bit annoyed with all this and I tend to think Peter might be too.

LEDs are lights, for hecks sake. They are diodes that just happen to turn electrical energy into light, with DC current  going in the correct direction.

You're right George, this ought to be a very simple problem to solve. Unfortunately we (the posters in this thread) are not in a position where Peter can take some basic measurements so we can confirm what is going on. Many of us here with a few minutes and a multimeter could solve this problem quite easily.
We're all trying to help - but we are really only guessing at the moment.

[PaulCheffus]

It seems that your LEDs are blowing because you are trying to pass too much current through them. T

I don't think that is the problem Webbo. The PPI4 instructions show the LED's connected with no resistor added. So the current must be limited on the PPI.
http://www.blocksignalling.co.uk/index.php/points-position-indicator-ppi4-dc
If the LED's were running at too high a current they would be extremely bright until they blow. I'm sure Peter would have noticed that.
The most likely problem, in my opinion, is that the LED's have been connected the wrong way round. There seems to have been some confusion in previous posts about which lead the resistor is connected in - I'm not sure if that has been resolved.
I still think the best solution, if Peter thinks he is capable of it, is to use plain LED's as described in the PPI4 instructions.
If this is practical for Peter then he could just wire up one set of LED's and see if they work reliably. If not then we have to think again.
Regards

Hi

Why does it matter which lead the resistor is in it will work whether it is connected to the positive or negative side. What does matter is which way around the led is connected to the supply.

If you mean that the convention is to put the resistor in the positive side and this is how they have been tested then a quick reversal of the supply should show this. I have never killed an LED by reversing the supply through it provided there is a resistor in series of an adequate size. I tend to use 1K for 12v

Cheers

Paul

[austinbob]

It seems that your LEDs are blowing because you are trying to pass too much current through them. T

I don't think that is the problem Webbo. The PPI4 instructions show the LED's connected with no resistor added. So the current must be limited on the PPI.
http://www.blocksignalling.co.uk/index.php/points-position-indicator-ppi4-dc
If the LED's were running at too high a current they would be extremely bright until they blow. I'm sure Peter would have noticed that.
The most likely problem, in my opinion, is that the LED's have been connected the wrong way round. There seems to have been some confusion in previous posts about which lead the resistor is connected in - I'm not sure if that has been resolved.
I still think the best solution, if Peter thinks he is capable of it, is to use plain LED's as described in the PPI4 instructions.
If this is practical for Peter then he could just wire up one set of LED's and see if they work reliably. If not then we have to think again.
Regards

Hi

Why does it matter which lead the resistor is in it will work whether it is connected to the positive or negative side. What does matter is which way around the led is connected to the supply.

If you mean that the convention is to put the resistor in the positive side and this is how they have been tested then a quick reversal of the supply should show this. I have never killed an LED by reversing the supply through it provided there is a resistor in series of an adequate size. I tend to use 1K for 12v

Cheers

Paul

It doesn't matter Paul. It's just they were trying to work out the polarity of the led (the connections being hidden by sleeving) by the location of the resistor. Not my sugestion- just an observation.

[Steve.T]

Peter,

I would agree with what has already been said above.

Unless there is fault with the China made power supply it should be fine.
Nothing against China in anyway as I would think 90% are made there anyway - but sometimes you do get faulty cheap items from Ebay - I actually had one of this type of supply that was faulty but in that it just did not work at all.
If you do have a multimeter it would be useful to check it has somewhere about 12V DC output (according to label +ive is the inner part of the plug). Sorry but in all the messages I cannot remember if you said you had a multimeter or not.

The LEDs also look fine.
But I would still test them on a 9v battery first and then you are happy the leds work. They should easily be bright enough to see they are lit but dimmer than normal (try in a darkish room is you find it difficult to see) and that if they consequently blow (I do hope not!) then it was not the fault of the LED but whatever you are connecting it to.

Steve

[Steve.T]

Peter,

Sorry if you have already answered this in other threads.

On the power supply have you cut the plug off the end and stripped back the wires or have you bought a socket that fits the plug?

The reason for asking is that these plugs and sockets have a huge range of diameter of the inner socket and also in the actual length of the outer spindles. It is real easy to have a plug and socket that appears to fit together fine but there is little, or no, contact between the inner part and can give intermittant contact. The result is a plug and socket that gets hot due to a high resistance contact and delivers very little power.

Steve

[Malc]

Peter,

Sorry if you have already answered this in other threads.

On the power supply have you cut the plug off the end and stripped back the wires or have you bought a socket that fits the plug?

The reason for asking is that these plugs and sockets have a huge range of diameter of the inner socket and also in the actual length of the outer spindles. It is real easy to have a plug and socket that appears to fit together fine but there is little, or no, contact between the inner part and can give intermittant contact. The result is a plug and socket that gets hot due to a high resistance contact and delivers very little power.

Steve

The problem is, as Peter stated, not that the LEDs don't glow, they stop working and are dead.
Peter, if the only thing you are doing is hot gluing them in place, stop doing it. They obviously don't like it up 'em. The best approach to mounting I found, was to but some bezels. The heat from the glue must be killing your LEDs. @petercharlesfagg

[petercharlesfagg]

The problem is, as Peter stated, not that the LEDs don't glow, they stop working and are dead.
Peter, if the only thing you are doing is hot gluing them in place, stop doing it. They obviously don't like it up 'em. The best approach to mounting I found, was to but some bezels. The heat from the glue must be killing your LEDs. @petercharlesfagg

Malcolm, this is something that I too have begun wondering about!

I have bezels but the LED's do not stay in them when I fit them!  They do not snap into place and I cannot retro-fit the little plastic grommets that lock them in place!

The only recourse I have is to use tape or Blu-tack but it would have been better to have purchased LED's with the Chrome bezels already fitted but of course the costs rear their ugly heads again!

As it is, I am ditching the LED's that were cheap from China and I shall purchase new from the company in Yorkshire.  The LED's from there may also be from China but I can go back to the supplier to complain if anything goes wrong!

Warmest regards Peter.

[MalcolmInN]

(4) Retained the 20 Cm cable length that came with the LED's and soldered the (C1 negative and the C2 Positive) together as suggested by our mutual friend Malc.

Have we all got the stamina to revisit this one from Peter in his very first post ?

If I remember right ( bearing in mind I dont have experience of this PPI but I did 'do' electronics for many decades after the dinosaurs lol! ) so I may have forgot a detail and I have not the stamina to go back and track it down, some of you youngsters can do that   :-
I think C1 and C2 were, respectively, the common returns for each of the two groups of LEDs ( the green ones and the red ones, but in Peters case what should be the "Group 1 Blue" and "Group 2 Blue", route set and route not set or summat like that )
In which case negative and positive for C1 and C2 have no relevance.
In view of the other misunners on the meaning of 'in-line'
is it worth re-exmining if what Peters implementation of what he thought Malc meant is the same as what malc infact meant,, if you see what I mean ?

I think there is room for a boobo somewhere there, it has been troubling me for quite some days now !

[Malc]

At this stage, Peter hasn't connected anything to the PPI. He has just connected the common legs together and hot glued the LEDs into place. So we suspect the glueing has stuffed the LEDs. They just don't like getting warm.

[MalcolmInN]

At this stage, Peter hasn't connected anything to the PPI. He has just connected the common legs together and hot glued the LEDs into place. So we suspect the glueing has stuffed the LEDs. They just don't like getting warm.

I know I have been watching, you say "all the common legs", he is talking all the C1negs and C2pos, are you sure you(we) are talking the same language ? we need diagrams but he cant draw them

I just invited consideration of what Peter may have been on about in C1 negative and C2 positive you see. It dont connect with what I understood you woz on about. (in case the hot of his thingumy isnt as hot as yours ?! )

There are, all through these threads, a disconnect between what those 'in the know' mean by a phrase and what Peter understands the terms to mean , , u c ? (see the other "in-line" thread, oh forget that I already said it ! )

As I said, have we the stamina for,
obviously not,,
so I'll butt out again.

[NinOz]

So we suspect the glueing has stuffed the LEDs. They just don't like getting warm.

Hot glue won't get any way near heating from soldering so not a consideration.
CFJ

[NinOz]

Peter,

1.  Supply Voltage
Unregulated 12V DC 1A should supply about 12V at 1A, depending on construction the power supply may be up to 18V or more without a load.  Always feed electronics through a voltage regulator if the circuit doesn't have one.

2.  Incorrect resistor to limit current through LED.
If you have a multimeter ( of course you do) you can measure the resistance by connecting one probe to the bare wire and one to the leg of the led, obviously the one with the resistor (easy if you have probes with points) but not connected to the power supply.  You don't have to strip the whole piece of insulation.

3.  Zzzzap
Don't let the hot-glue gun nozzle contact the leads on the led, probably some nice stray voltages from 240V line.  Super glue (gap filling kind) works ok for mounting also.

4.  Reverse voltage
Have only lost a few to wiring up in reverse with 12V to 15V.

Test a couple of them with the 9V then the power supply before installing.  Will tell you if voltage is a problem.

CFJ

[petercharlesfagg]

Peter,

1.  Supply Voltage
Unregulated 12V DC 1A should supply about 12V at 1A, depending on construction the power supply may be up to 18V or more without a load.  Always feed electronics through a voltage regulator if the circuit doesn't have one.

2.  Incorrect resistor to limit current through LED.
If you have a multimeter ( of course you do) you can measure the resistance by connecting one probe to the bare wire and one to the leg of the led, obviously the one with the resistor (easy if you have probes with points) but not connected to the power supply.  You don't have to strip the whole piece of insulation.

3.  Zzzzap
Don't let the hot-glue gun nozzle contact the leads on the led, probably some nice stray voltages from 240V line.  Super glue (gap filling kind) works ok for mounting also.

4.  Reverse voltage
Have only lost a few to wiring up in reverse with 12V to 15V.

Test a couple of them with the 9V then the power supply before installing.  Will tell you if voltage is a problem.

CFJ

Thank you for your reply to my queries.

(1) I had assumed, being a lay person, that a transformer reducing 240 volt AC to 12 volt DC would be a constant flow of electricity at that rate?  This "voltage regulator" confuses matters yet again!  I studied the results from a Google search and there are thousands to choose from and I haven't the foggiest idea which one, if any, I should buy!  (Back to square one!)

(2)  Yes I have a multi-meter but I cannot understand all the jargon in the instructions!  I ONLY use it to test if I am getting a connection by the needle moving, the rest is a mystery! As for removing the shrouding on an LED, I gave up and chucked it in the bin!  I trust that the suppliers of LED's who state that their products can be used with 12 volts DC and have resistors and wires fitted, because I know no different!  I cannot solder those pieces myself because I shake so violently when trying to concentrate on small areas, my soldering nickname is Mr Blobby!

(3) With reference to the last paragraph, I HAD to rest the glue gun on something, probably the LED to stop me spreading glue over a wider area due to my shaking hands!  It is EXTREMELY likely that I DID rest on the LED's , cables, resistors etc!  I have ordered a a small tube of silicone sealant instead.

(4) Reverse voltage I do not understand!  If I connect up the LED the wrong way it doesn't work, if it is the correct way around it does, job done!  I WILL be testing EVERY LED BEFORE fitting, just as I did the first, second and third time! 

I REALLY appreciate your helpful comments, regards, Peter.

[Bealman]

LEDs are cheap throw away devices like ballpoint pens, unfortunately (not getting into environment discussion here), so I concur with earlier posts that the cause needs to be addressed - LEDs are cheap and easily replaced, but there's no point if the power supply issue is not sorted.