LED Query

[Ralphmp]

I'm adding LED signals to the station part of my layout. There will be maybe 8 or 9 signals in a relatively small area (2 metres by 3 metres) and all will use the same type of LEDs and resistors (i.e. all the green ones will be the same type and all the red ones will be the same type with appropriate resistors for each).

To simplify wiring, I wondered if I could create a kind of distribution centre which has a common negative feed and 2 positive feeds - one for green and one for red - that incorporate the appropriate resistors. The idea would then be to run cables direct to each signal LED from the distribution centre. This would avoid the need to incorporate resistors for each individual signal LED, saving me some soldering (not a strength of mine...) and some components.

The longest run from the distribution centre would be around 2 metres so given the small voltages and current needed to drive the LEDs, will I have an issue with "loss" of either of these between centre and signal?

Hopefully this makes sense.

Thanks in advance
Phil

 

[PhilP]

You can have one leg common to everything, but will need individual resistors.. As you light more LED's the current required will increase, Ohm's law will dictate the voltage across the resistor will increase, giving less volts across each LED. - They will dim a little, then not light.

If you *know* that only a few will be on at any one time, you could reduce the number of feeds (and hence resistors).
Another way wold be to use a constant current supply, but his involves more calculation and electronics.

 

[ntpntpntp]

Given that you'll only expect to have either red or green lit, you can use a single resistor common to both LEDs.

Personally I'd be inclined to fit a resistor to each signal, simply as a safety: you then know that each signal can handle your normal supply voltage and you're not relying on your distribution box to limit the voltage. You can remove and test a signal using the supply voltage without fear of blowing the LEDs.

 

[Ralphmp]

Thanks for the swift responses. I'll go with the "resistor per LED" approach as, although it means more soldering (which I just can't seem to master) I can see it will be a more reliable option.

Best wishes
Phil

 

[sparky230]

Buy a decent soldering Iron 25watt minimum, forget about lead free solder, get some decent proper 60/40 it's still sold

Clean your componets and wires, tin the wire before soldering to LED or resistor. Get a damp sponge to wipe your soldering bit on.

 

[Ralphmp]

Decent soldering iron - CHECK
Clean components - CHECK
Tin wire - CHECK (kind of... see below)
Damp Sponge - CHECK

Swear box - CHECK (will need a new one soon )

My problem is the dratted solder insists on staying on the tip of the soldering iron rather than going on to the wire I'm trying to solder. I'll try the 60/40 stuff.

 

[PhilP]

Decent soldering iron - CHECK
Clean components - CHECK
Tin wire - CHECK (kind of... see below)
Damp Sponge - CHECK

Swear box - CHECK (will need a new one soon )

My problem is the dratted solder insists on staying on the tip of the soldering iron rather than going on to the wire I'm trying to solder. I'll try the 60/40 stuff.

Tin the iron.. Place against wire, pause a second, and feed a little solder into the joint.. - The solder should melt, and the extra flux cause it to flow into the joint between the wire(s) and components..

 

[Ralphmp]

Re: LED Query - additional question

OK - I followed the advice on soldering and I've now assembled my first signal

I used 5mm LEDs and resistors from RS components, and checked the resistor type using a couple of LED resistor calculators. I bench tested the signal by hooking it up to my Massoth power supply and confirmed both LEDs worked as expected.

However, when I installed the signal on my layout and connected up, the Green LED works fine but the Red is dim (like me ). I've rechecked all the wiring, joints, etc. and I can't find a problem so I'm wondering if when I did my bench test I was misled into thinking the Red was brighter than it actually was (if you see what I mean).

Being a novice in this whole area, I wondered if everything is actually working OK and I've just bought the wrong "type" of Red LED. The ones I bought are Panasonic LN21RPX, Round Series Red LED, 2-pin 5mm (T-1 3/4) Through Hole Package (!) (RS Web Catalogue Ref - 7870051). The wavelength is defined as 700 nm (nanometres?) which, according to the colour chart on the web site, is towards the darker end of the red zone. However, it just doesn't look very bright even allowing for the colour.

Based on the LED spec (Fwd Volts 2.8 max, 2.2 typical; Fwd Current 25mA) both of the calculators I used advised an 820 Ohm, 1 watt resistor and this is what I've installed. However, one also suggested that a 750 Ohm, 1 Watt resistor could be used. If I did go with the lower resistor, presumably the LED would be brighter, but there would also be a risk of premature failure?

If any of you electronics gurus have any suggestions on things to double-check, LED type and resistor value I'd be very grateful as I'd like to try and fix this now before I install all of the other signals.

Many thanks
Phil

 

[ntpntpntp]

What's your operating voltage - 20v? Check it.

Try a different LED and see if it looks the same with the same 802R resistor? If that also seems too dark then I'd say try the 750R resistor instead and see how it looks. With a forward drop of 2.2 over the LED and an 820R resistor the theoretical current is around 22mA, but there will be variations due to component tolerances.

The difference between 820R and 750R is only a couple of mA in this situation. You could probably go down to 680R if really necessary, but then you are at or slightly over the rated current for the LED if you really are feeding 20V. Most online LED resistor calculators err on the safe side and only take the current up to around 60%-80%

 

[Neil Robinson]

Good news, IMHO your workmanship is fine.
Not so good news, I wouldn't have used your choice of red LED as I think it will be dim whatever you do. According to a quick glance the luminous intensity is a mere 4 mcd. Values in the thousands aren't uncommon these days. I didn't see the RS ref. for the green in your post so I couldn't check that ones intensity.
I suggest looking up the luminous intensity of the green and choosing a red led with a similar or brighter value to the green.

 

[ntpntpntp]

That's a good point Neil. I don't use RS much myself, but just looking at the cheapest 5mm LEDs at Maplin as a comparison, the red is typical 4mcd whilst the green is typical 20mcd at 20mA. Spend a bit more and you get into 100+mcd and double-ish to get 12000mcd

 

[Ralphmp]

Thanks for the replies. Ironically, I originally built a test LED setup using LEDs from a Maplins starter kit and they worked OK. However, as I wanted several "identical" LEDs for my signals I bought a pack from RS to ensure consistency - guess they'll all be consistently dim!

For info, the green LED data sheet (RS Ref 7870083) says the luminous intensity is 15 typical, 6 minimum (vs the red data of 4 typical, 1.5 minimum), so I guess I need to look for some similar red LEDs.

As they say, you learn by making mistakes and at least a packet of LEDs isn't mega-expensive. I thought things had gone too well!!

Will keep you all posted. Thanks again
Phil