Constant lighting circuit _PLEASE!!!

[tramcar trev]

Ok so I'm going with the warm white 3.2V, 20mA Leds. I'd like a circuit that a) gives directional headlights while b)allows the internal lighting to remain on all the time. Analouge power, no DCC.... The net is full of them but most seem to be for incandescent globes. It appears to me to be simple enough to do a voltage regulator, a large cap to stop flicker and a rectifier.... BUT that may not give directional head lights, which I guess is not that important. A Vero or perf board solution would be positivley spiffing. One of you chaps will have done this already.......

 

[beavercreek]

Good link Ross! That one has been put straight into my 'useful' list.

PS have not got to post office yet but will post DVD today

 

[tramcar trev]

Even simpler could one not use a 1/2 watt 3.3v Zener diode in series with each led??? The led should come on as soon as track voltage reaches around 3v and the motor wont start till around 4 volts..... Just a thought....

 

[tramcar trev]

Yes Ross I did that.....

 

[Zerogee]

Sometimes it's still worth asking a question here, as Trev did, even if you COULD simply find the (or an) answer on Google - airing it here can result in all sorts of interesting additional feedback and ideas from various folks, some useful, some not, but all worth what you pay for it!
Besides, if everyone just googled everything and never asked anyone else, the forum would get pretty empty apart from people bragging about their latest purchases!

Just my opinion. Other opinions may be available.

Jon.

 

[Neil Robinson]

I'd be tempted to use two 7805 type voltage regulators. One fed via a diode bridge for the constant lighting and the other fed via a single diode for the directional light.

 

[tramcar trev]

Here's to you Mr Robinson!!!!
After a lot of googling I did come to the conclusion that I may have to do that and thank you for reading my question....
Most of the circuits that I found are for incandescent globes and from memory, the regulators need about 2v minimum across them so if you have 5v out, you will need 7v input as a minimum. For 3.3v out, need minimum 5.3v input. But it won?t give you full brilliance down to the voltage at which the tram stops moving.

 

[tramcar trev]

This seems to be a cheap and simple way to do it, the output is adjustable by tweaking the resistor values. This set up will turn on 3V leds at 4.5Volts on the track and will cost aproximatley AU$5 in bits bought via Ebay. One unit will supply around 1.5 amps which is 75 leds drawing 20mA ( you could run a bus through an entire train with this)......... so heatsinking the regulator ( lm317) is not needed for the 8 leds I will have on average.... It is not directional but I'll live with that....It could be made directional by replacing the rectifier (W04) with a single diode. The capacitor will stop flickering which with incandescents is not noticable as they take longer to turn on and off......
I'll make mine up on scraps of vero board but you could just solder the leads together and build that way if you had to hide this somewhere...
Ross just knew I'd answer my own question

 

[UKSGR]

You could put your 4.5v supply into a picaxe 08M2 and then drive 3 outputs to front, rear and interior. Use the power lines (suitably reduced voltage) to trigger either direction.sorted

 

[Neil Robinson]

Trevor, have you taken the voltage drop in the diode bridge into account in addition to that in the LM317?
An alternative for the directional lights may be to just use a diode in inverse parallel with an led with a 1k 1W resistor in one of the supply leads. You may find that high brightness leds give out enough light at around the motor starting voltage with this arrangement. If you've got the bits anyway you may wish to give it a try.

If non of the ideas work to your satisfaction you may need to increase the motor starting voltage. One way of doing this is in this topic.
http://www.gscalecentral.net/tm?high=&m=1230&mpage=1#1230 < Link To http://www.gscalecentral....=1230&mpage=1#1230

 

[tramcar trev]

Trevor, have you taken the voltage drop in the diode bridge into account in addition to that in the LM317? AHHH NO!!!
An alternative for the directional lights may be to just use a diode in inverse parallel with an led with a 1k 1W resistor in one of the supply leads. You may find that high brightness leds give out enough light at around the motor starting voltage with this arrangement. If you've got the bits anyway you may wish to give it a try. Yes I'll try anything. My circuuit is a fizzer - litteraly

If non of the ideas work to your satisfaction you may need to increase the motor starting voltage. One way of doing this is in this topic.
http://www.gscalecentral.net/tm?high=&m=1230&mpage=1#1230 < Link To http://www.gscalecentral....=1230&mpage=1#1230 Yep I have to agree... next thing on my adgenda.....

 

[tramcar trev]

Wellllllll ,its looking good. I used the LM317 circus above slightly modified. I fiddled with an L7805 as I thought it would be simpler and it is but it does not turn on until around 7 volts on the track but you can see form the pic it runs the 8 LEDs in parrallel quite nicely....

I went with the LM713 as its both adjustable and needs only 1.5 volts above the output to work so this setup turns on at around 5 volts on the track, the extra 1/2 volt is due to the drop over the diode the real purpose of which is to prevent the "fast battery" ie the electro capacitor from discharging back through the motor. I upped the anti on the Cap to 2200uf 16 volt ( I had a supply of these) and use a 12 ohm current limiting resistor in series with the output so I only need 1 current limiting resistor for the 8 leds. Yes these are very bright leds so I was able to turn them on at 3 volts instead of 3.2 and supply them with 8mA each instead of the 20mA in the documentation and get a nice "scale glow"... I used edge connectors simply because its easier to screw the wires into place. The boards are 12mm x 50mm x 23mm high but you could use radial caps or lay the bits over if hieght is a problem....
What I thought was a difficult problem was in the end not really hard to solve.... Now if I could lay my hands on a circuit for a HF track cleaner.......

 

[tramcar trev]

High Drama on the tramway today..... Car No1 was proceeding under way to test the rack to ensure that there was enough flangeway etc and there was a loud pop and a puff of smoke from under the car.... then nothing....
Inspection revealed that the 16v 2200uF cap on the constant lighting board had disintegrated.... I was able to replace it with a 25 volt cap and all is going swimmingly now... Maybe the 16V cap was faulty or did not like being "pumped up" by the PWM controller who knows. My wife is in control of the enquiry as the bang scared her while she was watching the momentous event.....

 

[Neil Robinson]

Inspection revealed that the 16v 2200uF cap on the constant lighting board had disintegrated.... I was able to replace it with a 25 volt cap and all is going swimmingly now... Maybe the 16V cap was faulty or did not like being "pumped up" by the PWM controller who knows.

Ah, electrolytics are notorious for intolerance of overvoltage conditions.
When connected to the output of G scale controllers I use a minimum of 50V, often 63V, working to ensure reasonable life expectancy. A reasonable rule of thumb is to use capacitors rated at roughly double the expected voltage.
Edit,
I've just looked closely at the photos of your circuits. All three appear to be different, one has a 63V capacitor and two have 16V ones in different positions on what appear to be otherwise very similar circuits. are both 16V ones on the input to the regulator or is one on the low voltage output?
I think you could get away with less than 2200uF, using a bit less space and reducing the cost when replacing with higher voltage ones.

 

[tramcar trev]

Inspection revealed that the 16v 2200uF cap on the constant lighting board had disintegrated.... I was able to replace it with a 25 volt cap and all is going swimmingly now... Maybe the 16V cap was faulty or did not like being "pumped up" by the PWM controller who knows.

Ah, electrolytics are notorious for intolerance of overvoltage conditions.
When connected to the output of G scale controllers I use a minimum of 50V, often 63V, working to ensure reasonable life expectancy. A reasonable rule of thumb is to use capacitors rated at roughly double the expected voltage.
Edit,
I've just looked closely at the photos of your circuits. All three appear to be different, one has a 63V capacitor and two have 16V ones in different positions on what appear to be otherwise very similar circuits. are both 16V ones on the input to the regulator or is one on the low voltage output?
I think you could get away with less than 2200uF, using a bit less space and reducing the cost when replacing with higher voltage ones.

Neil yes different is the word... all sort of experimental... the replacement is a 1000uF 63 v one... I use what I have to hand...