Hi power,Three phase dimmer

[vertebro]

Yes,of course,Sir N I will definitely give it a try! this circuit will be able to dimm R,S,and T all at once,
with just one potentiometer?

[Sir N]

[...] this circuit will be able to dimm R,S,and T all at once,
with just one potentiometer?

Yes, either with "just a potentiometer" (plus resistors or trimmers) or by the attached circuit.

The max. and min. voltages can be set or VR1 and VR2 respectively...
Use either a plastic trimmer tool/"screwdriver", or, if you plan to change them regularly, a potentiometer with isolate shaft(!)
If you only plan to set it once, a resistor can be calculated to replce it.
As is, both levels can be adjusted around 32% (min. 0V..3.8V and max. 8.2V to 12V).

C1 and C2 helps against noisy potentiometers and short slips of the wiper contact - values can vary quite a bit without changing the behaviour radically (anything from say 47pF to eg. 4.7nF should do (although it will change the responce times, you probbly won't notice in use.

Next comes the buffers/voltage followers, which doesn't load the trimmers, but keep a low output impedance for driving POT1.

POT1 is the fader proper and it's driven by the min/max voltages you set.
R3protects the buffers, in case you had eg. the POT1 all the way up and the maximum level set to the full 12V when using the Kill switch.

R4/C3 holds the time constant for a slow(-ish) fade in. By using this, your (probably very expensive) lamps will have a longer life - the worst case scenario for a glow lamp is to be turned on at a dimmer setting of 50% with no slow start - like if somebody  forced you to do hard excersize the momet you wok up.
With the Soft Start switched in (by shorting X1-1 to X1-2), it will start from 0%light and reach the set level in a few hundred ms (with the values shown - more like if you get a nice breakfast in bed to wake you up.

Finally, a Kill Switch can be used on terminals X2-1 and X2-2. When they're shorted together, the light will turn off immediately and stay off as long as they're shorted.
Mind you, the circuit is still live, it's only the light you kill. I'm not sure if this would be needed or even useful in your application, but the terminals can be left out if not needed.

This is a bit more involved than the simple solution, but I think the streamlined setting warrants the op-amp.

To set minimum light (for all 3 phases):
Turn POT1 all the way down
Adjust VR2 to the level you need
Done

To set maximum light:
Turn POT1 all the way up
Adjust VR1 to the level you need
Done

The potentiometer- and trimmer values was chosen arbitrarily and it could be adapted to almost any other value (although a too low value will be a bad idea, due to the transformerless power supply, as there's not that much current to go around.

If you have to buy new components, it won't matter, but if you either have some existing potentiometers in a box or have some very nice ones in your collection (we all have "our precious ones ), tell me the values, and I can work it/them in.


I'll make the other changes I mentioneed (in a day or two, it's too late now) and when done, I'll recommend building it with just one power module first, as it will be simpler to comprehend - and the power modules can be made be 100% interchangeable, which means that a single spare module could be made for a swift repair (which won't be needed if you have it, to paraphrase on Murphys Law).

Check out the attached circuit and just ask if there's something you don't quite get.


Just reread your original post...

the 3 channels must be controled with just one potentiometer on control panel

Done

It needs to have a timer (with ci 555 would be nice) with a potentiometer on control panel too.a 0 to 15 seconds timer its fine.

Oops, didn't notice before - hav to add that.
Luckily, there's a vacant comparator that'll make a good timer, so no need to add more chips

We gotta have a serious talk about timing though...
a 0 second timer is no timer at all.
The timer will ,in effect, appear to work with a step size of 10ms (on 50Hz mains), due to the nature of the control method.
If you use a log(arithmic) taper potentiometer for the timer, it may be easier to set, as the distance between 1 and 2 seconds will be about the same as the distance between 5 and 10 seconds.
If precise adjustment is needed, it's easy to add an extra potentiometer to have coarse and fine adjustment. However, don't expect super precision from an analog timer of this simplicity.

A start input (and test button) from a magnetic sensor to activate the cycle.Its going to do hundreds of cycles per day.

I am assuming that the test button should replicate the magnetic sensor for a test run?
Do you have some info on this sensor?
Reed switch, hall effect (and if so, switch or linear) or what?
And what's the detection distance - or do you have some sort of switch/sensor in place already?

I'll implement the functions I overlooked and make the ramp generators adjustable in the next day or so - Until then...


Edit: Being too tired to check everything several times, so characters may be missing or double - my wireless keyboard is a joke with poor keys and apparently a will of its own.

[vertebro]

Sir you are the man! I dont have enough words to tank you all the effort you are doing!
Not too many people are like this,specialy in my country!

Well,you have a very special way to explain things! jaja 

I will buy all the components this week since here ,there arent too expensive,its a good idea start
with just one module,as you sugested,test it and learn everything about how it works,and then move on to the others two.
so,just tell me with pote values are you comfortabe with and ill buy them,
i loved the "soft start feature" you are absollutely right,the lamps are not cheap!
For the timer part ,im actually using analog timers on all older flashes we have working,so i dont worry about surgical precision! (from about 2 seconds to 12 seconds +/-)
By saying "magnetic switch" i meant reed switch,sorry,since they are cheaper to replace compared to an inductive one, I already set up a reed switch and a magnet on a moving part of the machine and trigger a 12V coil relay,since it must send signal to at least 4 flash units,if you have a better idea instead of the relay for handling the reed switch current,you are Welcome!

[Sir N]

Didn't get anything major done today (was on an all day family trip).

Yesterday, I banged together a timer, only to just now discower that I inverted the function (so that it kills the light during the timing period) - I'll redo it tomorrow.

I might have to add the optocouplers to separate the circuit ground from mains zero, but I'll have a look with fresh eyes on that as well - after a generous amount of shut-eye

i loved the "soft start feature" you are absollutely right,the lamps are not cheap!

Another thing to add is what we used to call "zero light", where the lamps get a very small amount of power (a few percent of the full period) to keep the filament wire on a very faint glow (this won't heat the glass bulb to an "ouch level" in itself) and hence its resistance up (a cold filament is extremely low impedance compared to it's power rating, and cold filament starts has been the death of the majority of incandescent lamps used in households as well as industry.
I'll find a way to include this as well,

For the timer part ,im actually using analog timers on all older flashes we have working,so i dont worry about surgical precision! (from about 2 seconds to 12 seconds +/-)

OK, best to keep it to the needed range, to use the full travel of the pot, as this gives you an easier setting.

If you want/need better precision, another method would be to count the ZX pulses with eg. CMOS-timers.
If you need a granularity of, say, half a second, one counter can be set up to divide with 50, and will thus produce 2 pulses a second (on 50Hz mains), while another counter and a 24 position rotary switch selects how many half-seconds the outpu should last,
This will be extremely precise. compared to a basic analog approach,m as the mains frequency is usually very steady (although I don't have any experience with Arentinian power grids).
It will also be a somewhat larger timer board and slightly more expensive, although the most expensive part is probably the rotary switch.

You could use a rotary switch (or even two) to set up the comparator based timer as well. The benefit of this is to have well defined times easily picked out.
If you usually only work with a few (eg. <12) different time periods, it might be worth considering a switch over a pot.

Not knowing the purpose of it all, I'm pretty sure that it will adhere to the laws of physics, so remember that the difference between 2s and 4s is double up, while the difference between 10s and 12s is only 20% up.
With a switch, it is better to use a constant percentage increase each step, like...
12 steps, 2s to 12s =>  2.00, 2.35, 2.77, 3.26, 3.84, 4.52, 5.31, 6.25, 7.36, 8.66, 10.20 and 12.00 [seconds]

By saying "magnetic switch" i meant reed switch,sorry,since they are cheaper to replace compared to an inductive one, I already set up a reed switch and a magnet on a moving part of the machine and trigger a 12V coil relay,since it must send signal to at least 4 flash units,

With a reed, you need careful consideration of how your parts move and whehe they'll always be within range, as the field strength of a magnet in-/decrease with the square of the distance (double the distance and the field strength is down to 1/4).

If this gives you trouble, an IR or LASER gate/reflection detector might be better (don't worry over the IR - your lamps are far infrared, while you'd be using near- or mid infrared for the position detector.

if you have a better idea instead of the relay for handling the reed switch current,you are Welcome!

The lamp controller we're glueing together here can be tailored to whatever... Like 1mA, 100µA or what you'd like, as long as a few things are taken into consideration: Electric environment (noise), wire length and if some part of the wires have to run close to heavy polluters like transformers etc.
Depending on your exact reed switch, it probably have a current handling of between 50mA and 10A.

A relay in itself can be a heavy burden on the reed contacts, unless flyback from the relay coil is properly snubbed.

But your first task on the triggering is to see if a reed will work in all possible scenarious, if not, there's an endless supply of solutions


Are you using this for rapid baking of powder paint coatings or somehing like that (if it's not a secret)?

[vertebro]

 I WANT TO SHOW YOU HOW THIS FLASH CURE WORKS,
INSTALLED ON THE MACHINE SO YOU CAN SEE IT IN ACTION..

this flash is the one i ve just build using the circuit dimmer posted in this web page
heres the link http://www.pcbheaven.com/circuitpages/Voltage_Controlled_AC_Light_Dimmer/

[vertebro]

another video with different point of view

[vertebro]

and some pictures of the prototype!

[cheerio]

looks nice for a prototype dude.

[vertebro]

THANKS A LOT CHEERIO!

[kam]

Oh dear, am i happy to see this work!!! Very good job mate! I'm really happy to see my circuits on heavy duty work!

Am i allowed to upload the videos on youtube and add them into the content of the circuit? Can you send me more details about the machine and the operation? I'd be happy to show how well this works (frankly better than i expected).

[vertebro]

OF COURSE KAM !!! YOU WERE MY PARTNER IN THIS ONE!!!!

NOW IM TRYING TO BUILD THE NEXT GENERATION  ,WITH SOME MODS TO CORRECT SOME
LITTLE ASPECTS OF BEHAVIOR AND USING JUST ONE POTE AND NO TRAFOS HAVE YOU SEEN THE OTHER
CIRCUIT POSTED ABOVE? IT USES OPERATIONAL AMPLIFIERS

THANKS A LOT AGAIN KAM!

THE BLUE MACHINE IS AN M&R SCREEN PRINTING MACHINE MODEL SPORTSMAN,HERES THE LINK
to the specs  http://www.mrprint.com/uploaded/pdf/SportsmanEX_EN_LTR8.pdf

[Sir N]

I WANT TO SHOW YOU HOW THIS FLASH CURE WORKS,
INSTALLED ON THE MACHINE SO YOU CAN SEE IT IN ACTION..

Thanks for the view. I've loved automation since when I had to make it with sticks and elastic bands and it's alwways exciting to see all the cogs fall in synch
(Please fix your caps lock key, shouting is not needed).

Judging from your metal work, I guess you've made lots of similar work - Love the red box!

I've attached a schematic for a power module. Please forget the first one, it's flawed, but I didn't realize, as it behaved nicely it seemed, but then it dawned on me... it drops every second halfwave

The new one is just a power module, the transformerless supply and the control (timer + levels) will follow - just thought you'd like to start somewhere.
C3 may or may not be needed. C4 and C5 are mutually exclusive. They're both on the schematic to get thm on a PCB - one is a through hole component and the other is a 1206 or 1210 SMD and I made it that way because it should be an NP0 cap and I don't know in what form you can get it.
All the jumpers is likewise related to a PCB - just think of them as traces.

Just to repeat myself - don't touch when on.

Are you able to make a PCB from a layout?

I'll finish the rest (and write a tut on it), as soon as I get the needed time (probably posting the bits as they're in), but if you want to test the power board ahead, let me know and I'll put a temporary controller together.

How well is the reed and it's wires isolated?
Will they need to be run on a safe (galvanically isolated) voltage, or is it OK to just use the 12V supply.

Remember to check the national rules on mains - If anyone besides yourself got hurt by it, you are better off, if you kept to the rules of the national grid - Isolation, creep distance to other mains traces and to low voltage etc.

Not knowing your skills at soldering, I'll advice you to practise until you produce really smooth joints...
A bad joint is a bad joint whoever gets smoked (man or joint )

[Sir N]

You still alive?

Please answer the questions below, as I have a slightly changed power module and the controller and PSU ready (PCB's as well), but I won't spend further time on documenting it, if it is of no benefit to anybody.

Are you able to make a PCB from a layout?

How well is the reed and it's wires isolated?
Will they need to be run on a safe (galvanically isolated) voltage, or is it OK to just use the 12V supply.

[vertebro]

Sorry I had to travel to a couple of factories to fix some broken machinery

Are you able to make a PCB from a layout?
Sure no problem!

How well is the reed and it's wires isolated?
Their contacts from factory will handle 50 mA on 220V ac,but in this proect,it must work with  24 v cc supply.(dont need to be isolated)

[Sir N]

Are you able to make a PCB from a layout?
Sure no problem!

OK, but pay attention to some of the narrow tracks on the controller board. They're a little close for comfort, but since I probably have to change it here and there, according to what components (sizes) you can find, I didn't wanna spend oceans of time on it - use it for testing and I can change the final version.
I have used a handful of jumpers, rather than using a double sided PCB, as this is usually cheaper, but if you have facilities for making dual side PCB, I can get rid of the jumpers.

How well is the reed and it's wires isolated?
Their contacts from factory will handle 50 mA on 220V ac,but in this proect,it must work with  24 v cc supply.(dont need to be isolated)

The reed switch should connect directly to he controller, so I'd stuff it into a length of flex-tubing.
You might wanna get a new (unused) reed switch, since the one you have already used for switching a relay will probably be slightly pitted at the contacts, unless it's a mercury wetted type (hard to come by these daays, due to the ill informed paranoia of the western world - mercury in its metallic form is not poisonous, but trying to convince a tree hugger about that is a true Sisyphus job )  and the switch currrent, as is, is very low, so is best served by a clean switch. The switch will carry around 12V open circuit and will have to switch a max current of ~120µA as long as nothing goes wrong. However, the 12V is directly off the mains and shit happens, so isolate to avoid stinky fingers

As the text says, install bleeders on the 3 caps in the supply, or you may have to consider direct punishment, as they can hold a charge for a short while without them. They're not needed for the correct functioning of the circuit, but may keep you from the occasional nasty bite.
They can be ether stuffed under the caps, under the PCB, or lying on top of the caps (double sided foam tape may help) with the wires going down the ends.
Electrolytes should be 25V or 35V nominal and if the box get hot, you may wanna consider 105°C types (over the regular 85°C types).

If something is unclear or you find errors, please shout - I have shuffled a bit of files around and while I have given the final .pdf a read through, one can never be too sure when proofreading ones own work.
Anybody else have comments... Shoot.

Edit: Testing the file downloaded correctly, I discovered that I have swapped the PCB overlays around. I'll fix it in the next version, juat live with it for now