next topic »

[Reply #45 on:]

A simple test would be to directly connect a relay coil to the output of the regulator chip and see if the regular chip can maintain the voltage near 12v. Do this without any of the other components in the mix (you may need to unsolder some components for testing). You could also see how much current the coils are using. Bottom line, you may need to start over from scratch and measure component performance at each step to verify expected performance.

[Reply #46 on:]

Going to make it all over, with new components.

[Reply #47 on:]

Don't have any software to make a real schematic, and not the skills to use one either.

yes you do. It's called a pen, use it to draw the schematic on a thing called paper. Then photograph the paper and post the photograph.

You can't possibly make anything without a schematic.

[Reply #48 on:]

Amazing, you know more about me than I do.

[Reply #49 on:]

You can't possibly make anything without a schematic.

You can make a mess.

[Reply #50 on:]

Right BLD, this has gone on long enough.  
We all want to help but running around in circles is doing no-one any favours.
Buy yourself a breadboard and do a build on that using components of servicable quality.  When you establish that works then do a rebuild on the perforated board using the same components.
jack

[Reply #51 on:]

Okay, for the sake of argument let's say it is 100microfarads,  Question is  :  what is it's voltage rating.   It needs to be at least 35 volts.  If it's any less than by now its capacitance may well be zero.  

However IF you have connected the two 300 microfarad units across the bridge output (raw DC) then it makes not two hoots what's happened to the 100 unit.

By the way, old capacitors have a habit of losing their capacitance if not used.

jack

[Reply #52 on:]

Something else to consider :

Firstly, what you are trying to do is not rocket science, it is a very basic set-up and should not be giving the problems you have unless something is seriously wrong.

27vac rectified and smoothed via a capacitor bank and providing no load produces around 38 volts DC.

The absolute maximum rating (vendor data sheets) for both the l7805cv and l7812cv is 35 volts.

It is therefore possible (probable) that you have already destroyed (partially ?) the two regulator chips. (and the 100microfarad capacitor)

The chips may still be capable of producing the desired output voltage with no load but the internal current sense circuit, that maintains output voltage under load may be damaged and hence not providing the required regulation under load.  This might help explain why the chips get exceedingly hot when you have relatively little load (1 relay) operating.

If this is the case then you need to reduce the voltage being fed to the regulator chips.  You could achieve this by using a single input diode, rather than the bridge rectifier you are using. The resultant waveform would be pretty noisy and you'd need a fair size of main capacitor - say 1000microfarads, rated at 50 volts minimum - to smooth out the ripple.  Once through the regulators any remaining ripple will be reduced to virtually nil.

As I previously said, new components, a breadboard trial and start again.

jack

[Reply #53 on:]

Firstly, what you are trying to do is not rocket science, it is a very basic set-up and should not be giving the problems you have unless something is seriously wrong.

I know, but as I mentioned earlier in this thread, it is the first time I do anything like this. I normally just purchase something that can give me what I need, but the reason I am not doing it this time, is because I don't have any other wall outlet there to supply it with power, it would be stupid to have two power supplies running when it is possible to use just one, and because the knowledge I get now can help me with further projects too.
Working on a little list of items I am going to purchase...

Should a 2200µF-100V capacitor for the input be able to do it, or maybe two, just in case?

And then a 0.1µF-50V for each of the outputs?

Going to see what the voltage is tomorrow, when I only add one diode instead of the bridge.

A DC REGULATOR 12V/2A POS. TO220 for the relays, and then a DC REGULATOR 9V/2A POS. TO220 to supply 9V to the Arduino through the vin pin. (http://www.st.com/stonline/products/literature/ds/2148/l78s18c.pdf)

They should have same specs as the others (35V max) but 2A instead.

And what about cooling?
Would something like this be about right?



Or should I go for this instead?

As I previously said, new components, a breadboard trial and start again.

Just ordered one.. didn't have one before.

[Reply #54 on:]

"Oh no... Not again..." hehe

Just going to update a bit on this... Ordered a hand full of new parts.




Rated 3A each.


TO220 12V & 9V


TIP102 instead of the ULN2803A



100x 10k resistors (always good to have some spare for later projects ;D)



Size to the voltage regulators... I think these will do...



First little setup, not tested yet as I am rewiring something in the garage atm

[Reply #55 on:]

bld, I am just lurking on this thread, reading it to see how things work out - hope they do for you.

I noticed your choice of breadboard; good choice. I have one of these and it has lasted me almost 20 years now. However, in the future (near future, if you are doing a lot of playing) I would suggest spending the money to get the "larger" breadboard that has something like 3-4 of these smaller boards, plus power bus strips along the top with power binding posts, all mounted to a metal plate. Having the larger area to work on is a treat, and well worth the money spent. You could, of course, build your own version of this from the smaller boards if you need something larger (but expect to pay a bit more).

Also - you will likely find that the relays, while sized 0.1 inch to fit the PCB board you have, probably won't make contact with the breadboard's pins; the relay leads/pins may be too short. Even if they are long enough, they may be only "just so long" - so you put them on the breadboard, and it looks like it is working OK, then it stops because you nudged one or something. IE - it can lead to headaches trying to determine what is an intermittent fault due to the contacts barely touching each other.

So - do yourself a favor and when you get to that point, solder the relays (and shunt diodes) to the PCB, and run temporary connecting wires back to the breadboard for testing.

Something else which is useful during testing and later after installation is to hook LEDs up to the input side of the drivers (in your old system that would be between the Arduino and the ULN2803; in your new setup, beween the Arduino and driver transistors), with an appropriate current limit resistor - if you size it right you can use a single resistor on the base of the transistor to drive it and the LED at the same time without causing harm to the Arduino. You just want to use a resistor to give you enough current for driving both, without exceeding the output capability of the Arduino (keep it somewhere around 20-30 ma).

Having such an LED in place will give you visual verification of the circuit working, even if you can't hear the relay clicking or are too far away - or if the relay or transistor fails. It would be a little piece of hardware debugging tool.

Good luck with your rebuild - I think you were really close last time; it would have been better had you started on a breadboard instead of a PCB. Oh, one other thing for the future as you build a project - if you use DIP ICs on a PCB, put them in sockets (and add 0.1uf caps across the power supply inputs for filtering). Sockets make it easier to repair the circuit in the future should a DIP IC fail; after you have isolated and fixed the reason for failure first, of course...

 

[Reply #56 on:]

reading it to see how things work out - hope they do for you

I hope they do too, and soon working on all the wires in the garage, so I can control everything, plus working on the positioning project for the door too.

I noticed your choice of breadboard; good choice. I have one of these and it has lasted me almost 20 years now. However, in the future (near future, if you are doing a lot of playing) I would suggest spending the money to get the "larger" breadboard that has something like 3-4 of these smaller boards, plus power bus strips along the top with power binding posts, all mounted to a metal plate. Having the larger area to work on is a treat, and well worth the money spent. You could, of course, build your own version of this from the smaller boards if you need something larger (but expect to pay a bit more).

Yes, wanted that too, but this was the biggest they had where I ordered a bunch of other parts, and didn't want to pay shipping for just one board (shipping would almost be the same as the board itself)

Also - you will likely find that the relays, while sized 0.1 inch to fit the PCB board you have, probably won't make contact with the breadboard's pins; the relay leads/pins may be too short. Even if they are long enough, they may be only "just so long" - so you put them on the breadboard, and it looks like it is working OK, then it stops because you nudged one or something. IE - it can lead to headaches trying to determine what is an intermittent fault due to the contacts barely touching each other.

So - do yourself a favor and when you get to that point, solder the relays (and shunt diodes) to the PCB, and run temporary connecting wires back to the breadboard for testing.

Yep, already thought of that. That is the smallest problems too

Also got the right caps for the filtering this time.

Next problem will be to place it all correct on the board after all testing is done, and make it look good and all.

Currently trying ExpressPCB but can't find the right parts in it. Also going to make a version after this (close to 100% sure). One that will have the ATMega 328 in a socket on a homemade print. So it will be a more professional (i hope) looking product. Plus then make a socket for the wiShield to fit in, so it can be reused for it, and maybe later projects.