Led resistor question

[nookie]

So, just received my pack of 12v double Led platform lamps and about to wire them to my 12v bus, contacted the company and asked whether they are fitted with resistors or not, they couldn't say?? So, when connected to a 9v batt they light but not very bright so assumed they have resistors fitted?  (as they would have blown?) wired them up to the 12v bus, nice and bright but they are getting quite 'warm' so now assume they need resistors? ohms law not my strong point! 

If I wire them (6) in parallel any idea what resistors I should use to get the brightness but not the heat.

Cheers  

[johnb]

1. Are you sure that you are getting an accurate 12v?

2. Have you tried wiring them in series pairs. Although from what you say about them on a 9v battery that may be too dim.

3. Might be a bit of trial and error on getting the right resistor as if they are too dim on 9v it will be a bit of a fine judgement. As they haven't blown yet on 12v then that suggests that there is a resistor in there somewhere or that the LEDs on each lamp are wired in pairs as individual LEDs usually run on a max of 5v

[smokebox]

Are you absolutely certain they are LEDs?  There have been one or two cases of sellers describing lights as LEDs when they have been found to be filament bulbs.

 

The fact that they get hot at the higher voltage/current, without failing, is an indication of a filament lamp.  I would expect an LED to fail if it got too hot.

[DCB]

Sounds like your 12 volt Bus is a lot more than 12 volts.   

Model railway 12 volt is virtually anything, usually somewhere between 12 and 24 DC volts off load. LEDs draw so little current that the power unit is substantially off load. Car 12 volt is typically 13.8 to 14.2 volts. Computer 12 volt is wait for it, 12 volts,. Never assume 12 volts or 12 va is 12 volts  but always check the voltage with a test meter. 

 As the LEDs  work dimly on 9 volts they must have built in resistors,  even two Blue/ Clear LEDs in series would be bright on 9 volts and Reds on 4.5 volts each would usually glow brilliantly smoke and die.     If LEDs run hot their life can be measured in minutes, treated well they last well over 10 years used continuously. 

As yours haven't actually died in a brilliant flash I am guessing that while your power supply is more than 12 volts  it is probably about 16 volts DC so you need an additional resistor.  But you need to know the actual supply voltage not some nominal notional figure to be able to select the correct resistor.   Of course you could just get a selection box of resistors and experiment though you may ruin at least one of your LED lights in the process.   

Edited June 29, 2019 by DavidCBroad

[sharris]

50 minutes ago, DavidCBroad said:

  Of course you could just get a selection box of resistors and experiment though you may ruin at least one of your LED lights in the process.   

 if you're going to do it by trial and error, start high (a few k-ohms) which will probably be dim if you see anything at all, and gradually lower the value until they are bright enough for what you want. That way you're not likely to damage the lights.

[kevinlms]

1 hour ago, sharris said:

 if you're going to do it by trial and error, start high (a few k-ohms) which will probably be dim if you see anything at all, and gradually lower the value until they are bright enough for what you want. That way you're not likely to damage the lights.

Or buy a cheap multimeter and FIND OUT what the actual voltage is, then there is a known starting point.

The OP should be able to buy one from around 10 pounds.

[Paul80]

Something like this would do

 

https://rover.ebay.com/rover/0/0/0?mpre=https%3A%2F%2Fwww.ebay.co.uk%2Fulk%2Fitm%2F223468448427

[Nigelcliffe]

3 hours ago, Paul80 said:

Something like this would do

 

https://rover.ebay.com/rover/0/0/0?mpre=https%3A%2F%2Fwww.ebay.co.uk%2Fulk%2Fitm%2F223468448427

 

Yes, should be OK.   Don't expect any readings to be "absolutely precise", but will be well within the ball-park numbers needed for 99.9% of model railway applications.   Particularly, any attempt to measure DCC track voltage using the "AC" settings isn't likely to be accurate.   

The meter linked has the incredibly useful "continuity buzzer" option, for tracing whether wires are connected or not - don't buy a meter for model railways without a continuity buzzer.

 

 

[Tricky Dicky]

I am with Smokebox on this and as the OP had not replied we are not sure if we are dealing with LEDs or filament bulbs. The fact the supplier cannot say if resistors are included casts doubt on if they are even LEDs in the first place.

 

I think if the OP gets a multimeter the first thing I would measure is the current drawn by the lamps, grain of wheat/rice bulbs if I remember correctly should draw about 60-80mA. Anything less and it is likely LED.

 

If they turn out to be filament bulbs then before adding resistors to control brightness, wiring a couple in series will reduce their brightness by half only using resistors if the OP wants something between full brightness and half brightness.

 

If they are actually LEDs then having survived the OPs 9V & 12V tests suggests they already have resistors incorporated but maybe not enough so a little experimentation with additional resistors should achieve the correct brightness the 

OP wants.

 

Additionally checking the output of the power supply will help establish whether it is actually 12V or more. The power supply should be checked under load as many power supplies will give a higher voltage unloaded only dropping down to their rated voltage under typical load.

 

Richard

[kevinlms]

43 minutes ago, Tricky Dicky said:

I am with Smokebox on this and as the OP had not replied we are not sure if we are dealing with LEDs or filament bulbs. The fact the supplier cannot say if resistors are included casts doubt on if they are even LEDs in the first place.

 

I think if the OP gets a multimeter the first thing I would measure is the current drawn by the lamps, grain of wheat/rice bulbs if I remember correctly should draw about 60-80mA. Anything less and it is likely LED.

 

If they turn out to be filament bulbs then before adding resistors to control brightness, wiring a couple in series will reduce their brightness by half only using resistors if the OP wants something between full brightness and half brightness.

 

If they are actually LEDs then having survived the OPs 9V & 12V tests suggests they already have resistors incorporated but maybe not enough so a little experimentation with additional resistors should achieve the correct brightness the 

OP wants.

 

Additionally checking the output of the power supply will help establish whether it is actually 12V or more. The power supply should be checked under load as many power supplies will give a higher voltage unloaded only dropping down to their rated voltage under typical load.

 

Richard

A power supply should be checked on no load conditions, because that would be the case if all the lights (or LEDs) were off. If a PS at its rated load is 12 Volts, then unless a properly regulated one, is likely to be 16 volts or even more. That is the voltage that would be applied, at switching on the first switch to lights.

For a model railway, changing from a day scene to a night one, would mean turning on the lights over a shortish period of time, generally.

[cliff park]

A simple test for LED or filament is set the meter to a low resistance ( Ω symbol ), perhaps 200 Ω and measure across the lamp. Reverse the leads and measure again. The actual reading is not important, BUT if it reads the same both ways round it is definitely a filament. If no reading try a higher, perhaps 2K range. 

[DCB]

21 hours ago, cliff park said:

A simple test for LED or filament is set the meter to a low resistance ( Ω symbol ), perhaps 200 Ω and measure across the lamp. Reverse the leads and measure again. The actual reading is not important, BUT if it reads the same both ways round it is definitely a filament. If no reading try a higher, perhaps 2K range. 

It depends on the meter.  All the cheap meters I have seen have a single AA or AAA cell as power. These are around 1.6 volts and most LEDs don't pass any meaningful current until about 1.8 volts so the meter doesn't actually give a reading from the LED except infinity both ways.

The simple test is if the Lamp works on a power supply then reverse the polarity, the LED will go out while the filament lamp will continue to work exactly as before.

 

[PatB]

My own el-cheapo multimeter, when measuring resistance across diodes, shows infinite resistance one way, and a brief burst of conductivity, followed by infinite resistance 'tother way. Not sure why. It's enough to distinguish anode from cathode though.

Edited July 1, 2019 by PatB

[sharris]

1 hour ago, PatB said:

My own el-cheapo multimeter, when measuring resistance across diodes, shows infinite resistance one way, and a brief burst of conductivity, followed by infinite resistance 'tother way. Not sure why. It's enough to distinguish anode from cathode though.

 

Obvious question, but have you tried changing the battery? 

[PatB]

5 minutes ago, sharris said:

 

Obvious question, but have you tried changing the battery? 

I put a new battery in not so long ago. It doesn't do anything odd with simple resistances, just with diodes. As I'm not checking the resistance of semiconductors all that frequently I'm not terribly concerned, just intrigued.

[kevinlms]

2 hours ago, PatB said:

My own el-cheapo multimeter, when measuring resistance across diodes, shows infinite resistance one way, and a brief burst of conductivity, followed by infinite resistance 'tother way. Not sure why. It's enough to distinguish anode from cathode though.

Has to be a case of just enough voltage to cause breakdown on initial contact, then dies below. As long as it works for you, it doesn't matter.

Many multimeters do have a 'diode test' setting, which presumably boosts the voltage - multiplier circuit?

[Crosland]

5 hours ago, DavidCBroad said:

It depends on the meter.  All the cheap meters I have seen have a single AA or AAA cell as power.

 

All the cheap digital multimeters I've ever used have a 9V battery. You need more than an AA to run the electronics and display. They generally have a diode test setting too.

[298]

A filament bulb would change colour as the voltage increases, LED would get brighter but stay the same hue. You could test this by wiring one between the rails with a train running to act as a bigger current draw on the controller.

 

Or take one apart to see what's inside, and use it on the layout with a broken globe...

Edited July 1, 2019 by 298

[Amand]

Which manufacturer / supplier did you get them from? Maybe someone else on here has the same lights so could advise whether a resistor is needed, and the value. Swapping the power supply feed wires around will tell you instantly is it's an LED or filament. LEDs are polarity sensitive, so you have a 50% chance of getting the power feeds the right way round. If they remain lit when the power feeds are reversed then it's a filament bulb.

 

I've found that some lights from Trainsave are 12 volts, others 6 volts, though a basic wiring diagram is usually given.

Edited July 1, 2019 by Amand

[DCB]

6 hours ago, Crosland said:

 

All the cheap digital multimeters I've ever used have a 9V battery. You need more than an AA to run the electronics and display. They generally have a diode test setting too.

Yes, you're right.  My analogue(?) meters have AAs but the digital multimeter has a 9 volt.   That said the digital still does no't show any ohm reading on several new and used LEDs on any setting  20 / 200 / 2K /20K 200K or 2000K.   Works fine otherwise.  Which is why I thought it had an AA cell.

[sharris]

13 hours ago, DavidCBroad said:

Yes, you're right.  My analogue(?) meters have AAs but the digital multimeter has a 9 volt.   That said the digital still does no't show any ohm reading on several new and used LEDs on any setting  20 / 200 / 2K /20K 200K or 2000K.   Works fine otherwise.  Which is why I thought it had an AA cell.

 

The 'diode' setting on many digital multimeters gives a measure of the forward voltage of the diode under test - often it maxes out at about 2V, so some LEDs might show as functioning and some will not pass the test (e.g. Vf is around 3.5V for white) - I suspect the test voltage for the resistance measurement settings is of a similar magnitude and may not be high enough for testing LEDs with a higher Vf. 

 

Since you say you have more than one meter, you could set one to 'resistance' and measure the voltage across the probes with another to see whether it would be high enough to cause an LED to conduct. 

 

 

Edited July 2, 2019 by sharris

[sharris]

On 01/07/2019 at 17:51, 298 said:

A filament bulb would change colour as the voltage increases, LED would get brighter but stay the same hue

 

 

Apparently white leds can change their colour a bit as you change the current 

 

https://www.maximintegrated.com/en/app-notes/index.mvp/id/3070

 

I only found out about this a few days ago.