The term Solid State Relay covers several semi-conductor devices that are used to switch a load electronically, thus a Triac, SCR, Thyristor, etc. even MOSFETS can be described as SSR. Indeed, I suppose that you could extend that to any transistor or any 'black box' that replaces a mechanical relay.

Buying an off the shelf device is a valid method (I've not tried it, but I would think that hitting it with a 590Hz PWM would be a bit hit and miss without tying it to the mains frequency, but then that depends whether it has additional circuitry to manage this). Using a Diac removes the need for other power supply.

I would imagine that since arduinohabib is on a microprocessor forum, and hasn't opted for a light dimmer from his local electrical factors, he is looking to learn how to do it himself. That is why I suggested a method that he can research and try, thus expanding his understanding of AC voltages, switching devices, timing etc. He can of course pick and choose the method to follow. It's all fun, and many have us have already been there many times over.

One lesson for him to learn, is to understand exactly what he wants to achieve, what the bounds are and then to explore ways of accomplishing this.

As an aside, arduinohabib, one might consider using an SSR with a mechanical relay (not in a dimming, but in a pure switching mode of course). The advantage of this is that the load begins to switch at zero volts and the SSR need only be quite small as the mechanical relay is in parallel and will take the full load within a cycle or two (relays tend to take about 20ms to pull in), and the relay suffers little or no arcing across the contacts, reducing pitting and extending life. (switching off gracefully is a little more complicated)

Again - Take care with mains

*Points at grumpy* Ha Haaaa!
Beat ya!  Before the TSR80, with a home made camera, a home designed/built Z80 'pc' and a neural net; I was able to identify if the "camera" was looking left, right or centre of a scene.
The camera used a loop of super 8 film (hand cranked, never motorized it) as a pixel mask to give a 16 and then a 64 pixel (8x8) device (just filmed a white card on a black background to create a moving window) and a single photo transistor

No forums (or web) back then  The point being, don't let them tell you it's impossible, keep plodding and use your imagination. There are many ways to the solution, but the fastest way there is to understand the problem.

I was talking to a guy a couple of years ago who was into neural nets, and he said he uses high end video cards, and hacks the multiple cores/processors as a very fast engine to "parallel process" his nets.

Now matter how fast the processor, we will always overload it with crud and make out it's slow! Like cupboard space, you always have 10% more rubbish than space!

I see. Thank you. I was just thinking, can I use MOSFETs to PWM it? Or can I convert it to DC then MOSFET it? Thanks.

Maybe you should define your needs and then find a solution?

SCR's & Triacs behave differently. Triacs are bidirectional! (so we can have higher resolution and less flicker)

We know the frequency of the AC being about 50/60Hhz depending where you are.

So we know the time between 1/2 cycles

We also know that if we apply a tiny gate voltage pulse the device starts to conduct, and will continue until the voltage across it's main terminals is zero. (zero crossing point - Mains is AC and the voltage will pass through a minimal voltage during each half cycle)

Thus by adjusting the time of the pulse relative to 0V crossing (after 0V occurred) we can vary how much of the (half) cycle that the Triac conducts. The further from the zero crossing detection, the shorter the conduction time (to the next zero crossing), so the less average voltage appears across our lamp.

Isolation from the mains is important so we are talking transformers, or optically isolated devices, hence opto zero-voltage detectors and opto-isolated diacs (diac NOT triac)

So, basically your code will be

read zero crossing device
delay for Tbrightnes (in uS to give you lots of resolution)
output high to diac
delay for XuSeconds (see data sheets)
output low to diac
repeat

BE SAFE!

Incandescents are being phased out in the developed world.

However, if you can still get them, you may want to look up Voltage Zero Crossing Detectors, This way you get your soft start, you then turn off at any point or at the next zero point.

You'll need a couple of  opto isolators and a triac etc.
Check out snubber circuits
 - be careful

Yes, many RC chargers are intelligent and have the ability to manage individual cells.

It looks like you are series charging your LiPos is that right? Do you have any problems? what C rate to you charge at? (I have man flu, and I can't trust anything I see with these meds  ) (my Bosch has LiPo already  - impressed compared to my NiCd drill)

Thanks, but I was referring to the *input* source, not the output. e.g. charging the single lithium from a wall supply instead of a solar panel. If it expects a certain voltage/amperage from the solar panel, what would be the problem in getting that instead from a wall adapter?

Ebay, buy a handful of these...

search for "Converter Step Down Module LM2596"  cheaper than chips! No really, you can't buy the parts for the money.

Obviously you will need your wallwart to be able to supply enough current, so you need to know the C rate (current) that you will charge at (always over specify as wallwarts are usually poorly regulated and over optimistic when quoting current, you don't want another source of fire or failure.

AC wallwarts may be better for you, but remember that after rectifying & smoothing, your voltage will increase by root 2 (about 1.4 times) So bare this in mind when specifying your voltage regulator.

[volts that jolts]

Your plan looks okay.
If using heat shrink, ensure that the individual strands of wire lay flat or they may poke through, and you may get erm, a shock!

Remember, it's the volts that jolts and the millies that killies!

0.5A is plenty to kill you....

Once you have your relay working, why not look at optical isolation and zero-voltage-switching, you'll be amazed how long your bulbs last too! (filament type)

I'm really not up on LiPo, but then nor is Boeing  

I understand that individual cells within a pack need to be charged individually, and that one should not charge cells that have dropped below a certain voltage, nor mix cells of differing voltages. But I read that a long time ago and it may not apply to current cells. (something else to do!)

Edit
This looks readable http://www.rcgroups.com/forums/showthread.php?t=209187

Edit, Edit
Mmmmm, quite old "Winter 2005"  But worth a read, Check out the lower end of the post if you want to be put off.

I've had a look around for home-brew chargers, but not found any I'd recommend, but there are many low price units out there (if you have a multi pack, get a balance charger). As with NiCd's it's the RC peeps who are pushing with LiPo so go read their forums.

Many years ago I used to make fast NiCd chargers for RC clubs, I always supplied them with metal charging cages due to the possibilities of explosion & fire!

It looks like the first part of your project will be a battery management system... Something for your Masters degree Then you can get a job with Boeing - even the big boys get it wrong.

Hello!

Have you read the getting started pages? http://arduino.cc/en/Guide/HomePage

You may find that setting the Port to use PORT 1 will work better.

Best of luck

Bonjour!

Avez-vous lu les pages de mise en route? http://arduino.cc/en/Guide/HomePage

Vous pouvez constater que la configuration du port pour utiliser le port 1 fonctionnera mieux.

Bonne chance

[rant]

Mortified!

This is indicative of the selfish, instant gratification, vague, rude, posts that seem to be forever increasing across the net.
Well done you guys that spoke out!

We all love to help, but, I at least, like to feel that the person receiving help will somehow meet me half way, do some research, and improve their lot and maybe pass on some of that new knowledge. Not, as I feel will happen here, take what they can, use it until bored and then demand more ways to fulfil their ill conceived plans.

Many people on this forum are keen to give up their time and freely spout code until the cows come home, personally I like to point in the right direction, try to get them to understand what it is that they are doing and why. That way they get more fulfilment.

I suppose that the Arduino is it's own worst enemy, it appears to de-skill to such a degree that looks like putting Lego together - which is cool - but belays the need for at least a meagre understanding.

But what really annoys me is the rudeness of these people. This guy should be ashamed of himself - he is basically mugging people and spitting in their face while he does it!

Everyone should take a stand against this kind of behaviour.

[/rant]

No Cerin, not a shield.

I have a microprocessor on board  - ideal for monitoring a battery. But as Chargrin posted, there are many chips to do the job! This allows you to make 'units' that you connect together and the operation of one does not impinge on another.

Do read up on batteries, e.g. if using lead acid, ensure you use it within it's recommended charge/discharge range, otherwise you will shorten the life. Don't forget about weight either!

Tricky!
Have you tried powering the servo after the main board?
I would suspect that it would have little effect since the setting time of the servo would be many times longer that the start-up time of the board, but may be worth a go.
It may also help to set the servo position on power-down and also initialise to the same position on power-up (maybe before applying servo power)

oh! I thought you were using a seperate supply for the motors

and charge both the netbook's battery and the additional 12V battery (likely lipo) simultaneously.

I'm making something similar ATM (no laptop), but using a solar panel to charge a 12V gel battery. Power requirements should be within daily charge capability, but I will be adding a sun-bathing routine if the bat gets low. I'll put the panel at optimum angle and allow the unit to rotate toward the sun. Bit dark for testing at the moment So waiting for the spring.  Continuous activity is not necessary as this is an exercise in navigation rather doing any useful work. For that I will probably aim to set up a self serve "fuelling point" as the area available will not allow a high enough current density from a PV panel.

Good luck

i dont understand y i should use a mosfet

please enlighten me

You don't have to (although I did leave a a huge clue why you might want to)

I would suggest some reading... DC42's posts are a good start,and re-reading the article you posted (it tells you what the diode is for!) and find some data sheets on your motor, your transistor and then maybe also compare your requirements against a mosfet and see what conclusions you come to.

Draw an H=Bridge and see if you can understand how it works (think of the transistors as switches). Hint - nothing to do with high currents as suggested above. patduino suggested reading about transistors as switches - great idea (although you didn't seem to think so)

The exercise will hold you in good stead for your next project. Asking questions is a good way to learn, but doing it off you own back with a few pointers to get you going is so much more rewarding than expecting someone to do it for you! It really is worth the time and effort.
Lets hope your next post is a video of you controlling your motor!

Kind regards

Thanks

The documentation is generally elegant and clean although not always consistent; but a great job!   (no clapping smilies)