Automatic Light Dimmer On A Timer

[keenonfish]

In short I do not know how the pro dimmers work, I just know how to use them. If I ever came across one that was bust, then, being twin phase mains, I would likely have a professional electrician in to look at it. Twin phase will kill instantly in almost every case if you come on the reciving end of a jolt of it, so I would not touch the inside of a dimmer unit unless I knew what I was doing in there. Any transformer can in practice store a charge for a while after use, and is one of many potential hasords within the unit that I would be concerned about... Lanterns and the rest of the rig where you are basically wiring a plug, plug socket and extension lead without a fuse (They are in the dimmer unit either as standord fuses or as RCD's) I can cope with more than happily, but gubbins within a dimmer is a bit beyond where I am comfortable to go with electricals.

A healthy person can take a jolt from normal mains, so long as it isn't directly across the heart (jolt revieved from the right hand/arm and grounded through the right leg is relatively safe in most cases, but painful. Been there, done that, wrote the book and collected the T-shirt ), but twin phase mains isn't as forgiving apparently (I haven't been shocked by it yet, so couldn't tell you )...

All the best
Rabbut

I assume you were using two phases of a 3 phase system then... A lot of shops which require lots of power often have all phases present amonst the sockets so as not to overload the supply / consumer unit!

I have worked with three phase power (ex catering engineer) yes if you touch two phases at once you are absolute toast, 415V across two points of a human not good. But, you may be interested to know that every household is connected to one phase of a three phase system so touching one terminal of a three phase supply is no different than touching the live at home - it still bloody hurts like hell and you should consider yourself lucky to walk away alive

Anyway I might look into getting creative with lighting when I get my larger tank so will keep the forum updated

 

[rabbut]

I've done more thinking on this one, and can't see now why I got a different fade time on and off...

The capasitor takes time to charge and discharge, due to the resistor after it. It take time to set-up and disipate a potential across any capasitor, but with low resistance, discharge and charge should be fast. Charge and dis-charge should obey T=RC, timeconstant is equal to the total capasitance of the circut, multiplyed by the total resistance of the circuit, after the capasitor. Because the same equation applies to the charge and discharge, T remains constant for both, so shurely the charge and discharge times would be the same I suspect I must have been leaching charge into the circuit, or somehow getting an increased resistance in the old loop after switch off Can't think how though...

All the best
Rabbut

 

[ichatfilipina]

I've done more thinking on this one, and can't see now why I got a different fade time on and off...

The capasitor takes time to charge and discharge, due to the resistor after it. It take time to set-up and disipate a potential across any capasitor, but with low resistance, discharge and charge should be fast. Charge and dis-charge should obey T=RC, timeconstant is equal to the total capasitance of the circut, multiplyed by the total resistance of the circuit, after the capasitor. Because the same equation applies to the charge and discharge, T remains constant for both, so shurely the charge and discharge times would be the same I suspect I must have been leaching charge into the circuit, or somehow getting an increased resistance in the old loop after switch off Can't think how though...

All the best
Rabbut

NICE. I want to try this one. can you post the photo or pics?

 

[keenonfish]

I've done more thinking on this one, and can't see now why I got a different fade time on and off...

The capasitor takes time to charge and discharge, due to the resistor after it. It take time to set-up and disipate a potential across any capasitor, but with low resistance, discharge and charge should be fast. Charge and dis-charge should obey T=RC, timeconstant is equal to the total capasitance of the circut, multiplyed by the total resistance of the circuit, after the capasitor. Because the same equation applies to the charge and discharge, T remains constant for both, so shurely the charge and discharge times would be the same I suspect I must have been leaching charge into the circuit, or somehow getting an increased resistance in the old loop after switch off Can't think how though...

All the best
Rabbut

If the resistor is after the capacitor (as in your example above) then charging will be near instantaneous (of course this would depend on the supply itself). However the resistor (and LED's) would limit discharge time which would be good. If the resistor is on the supply side then it would obey the T=RC rule (capacitor charging through resistor) but then discharge would be at the mercy of the LED's which would depend on how many and how they were connected (series or parallel).

All I am arguing is that you are never going to get a satisfying on / off fade with such a simple circuit.

 

[pastabake]

It can't be that hard to dim LED's ... I've got one of those torches that you shake to make power ... the LED gets brighter the more you shake it ... so it's basically a case of limiting the power so a simple variable resistor should work.

I've also just bought a CFL that claims to last 8 years and is truely dimable (no messing about with on/off) just turn the light dimmer and away you go. Having installed the light switch I don't believe that it anything more complex than a variable resisitor.

 

[raptorrex]

It can't be that hard to dim LED's ... I've got one of those torches that you shake to make power ... the LED gets brighter the more you shake it ... so it's basically a case of limiting the power so a simple variable resistor should work.

I've also just bought a CFL that claims to last 8 years and is truely dimable (no messing about with on/off) just turn the light dimmer and away you go. Having installed the light switch I don't believe that it anything more complex than a variable resisitor.

I would be really interested in finding out more about dimable metal halide because everything I've ever read about them makes me think that it's not possible.

well it is!!!!!!!! limiting power will not do the job, not for long anyway. these PWM were not made to cause you extra cost and complication. they are there because they are needed. and its as simple as that!
you will need to change the ballast units for both halide and fluorescent lighting if you wish to dim them! all of which has been said before, in this thread, by more than one poster.

 

[rabbut]

I've done more thinking on this one, and can't see now why I got a different fade time on and off...

The capasitor takes time to charge and discharge, due to the resistor after it. It take time to set-up and disipate a potential across any capasitor, but with low resistance, discharge and charge should be fast. Charge and dis-charge should obey T=RC, timeconstant is equal to the total capasitance of the circut, multiplyed by the total resistance of the circuit, after the capasitor. Because the same equation applies to the charge and discharge, T remains constant for both, so shurely the charge and discharge times would be the same I suspect I must have been leaching charge into the circuit, or somehow getting an increased resistance in the old loop after switch off Can't think how though...

All the best
Rabbut

If the resistor is after the capacitor (as in your example above) then charging will be near instantaneous (of course this would depend on the supply itself). However the resistor (and LED's) would limit discharge time which would be good. If the resistor is on the supply side then it would obey the T=RC rule (capacitor charging through resistor) but then discharge would be at the mercy of the LED's which would depend on how many and how they were connected (series or parallel).

All I am arguing is that you are never going to get a satisfying on / off fade with such a simple circuit.

The charge time would depend on supply as you say. If large current is created at start-up, the charge would be almost instantanious, but if current remains constant, as most supplys will, you will theoretically get equal charge and discharge times. A potential across the capasitor takes time to set up, as electrons have to be pushed onto one side of the capasitor and off the other, through the time control resistor. This limits the current from the capasitor while charging, and the current from the capasitor increases with the PD stored on the capasitor, as the time control resistors resstance will not change. The PD storage comes from an imballence of electrons across the capasitor plates. When they discharge, elecrons have to be pulled to the posively charged side, and pushed off the negatively charged one. The resistor after the capasitor should limit the charge time and discharge time equaly as each other, as electrons have to pass through it in both cases, unless your supply unit can push lots of current in at start-up Batteries and laboritory power supplys will give equal charge times IME, but some plug-in transformers will not, depending on how they are transforming and converting to DC current.

I will sketch the circuit diagram for the poster that requested pics, but this unit moved with the person I built it for a while ago now

I intend to re-build one at some point for myself, as the heat from my halides in the closed cabinet of my marine tank will cook my fish in summer unless I cange the light I am using, or put enough PC fans in the hood to drown out the noise of a jet engine

The unit I built was an all blue LED night lamp, but the one from my marine tank will need 14K white lamps and I would like some LED's that mimic Atinic Florescent tubes. These will take time for me to source, and will likely be more expencive to buy than normal blue LED's. The cost from my friends unit came to about £10 will all 50 LEDs, circuitary, mountings and transformer fittings...

The circuit you describe in your post could not give a slow fade on or off time. A resistor before the capasitor will serve to limit the current through the circut, not from the capasitor. You have almost neglagable resistances from LED's, so the time to fade the LED's on and off will be less than a second in each case Some voltage will be disipated over your pre-capasitor resistor, so it would serve to make the LED's glow more dimly, but not make them fade on or off

All the best
Rabbut

 

[pastabake]

Boboboy - Technology has moved on mate ...

Go take a look here, they also sell ballasts for those wanting to dim LFLs using a standard dimmer switch ... h**p/www.varilight.co.uk/Pages/page%20digiflux.htm

btw - no amount of wrong makes a right.

 

[rabbut]

Just in case anyone is still following this thread and is interested in circuit diagrams, I'm make a DIY "TMC Aqua Ray" unit, that I'm documenting in THIS THREAD. Though I have not posted a diagram, one will be going up just as soon as I work out how to use the new scanner....

All the best
Rabbut