led prototype wiring

Good morning all i am a little confused as to the outcome below, i wired up my breadboard connected it to my mega 2560 and to a strip of 3 rgb led's. I then got the program working and all was great, so i knocked myself up a prototype board using the same components. when i connected them the led stayed on dim, when it should have been off. i then dismantled the board and connected one color back to the breadboard, connected 1 color to the MOSFET hard wired and made a second prototype board for the third color.
when i tested it the one on the breadboard worked ok the hard wired one worked ok but the one on the prototype board left the led on dim. All 3 colors running at the same time.
here is what i am using.
12v power direct to led strip +12v terminal,
Groung from 12v to breadboard,
Arduino 5v grnd to same breadboard.
PWM pin from arduino to 10K resistor on to left hand side of STP16NF06 MOSFET,
Red of the RGB led strip of 3 to Centre pin MOSFET,
GND from breadboard to third pin of MOSFET.

pictures of the prototype board

the led with the prototype board on

Parts taken from board and soldered together the same way

Led's work fine none stay on if soldered like this or on breadboard.

does anyone know why the prototype board would keep the led on

more pictures here
http://www.telmal.co.uk/tel2.htm

diag i copied here
http://learn.adafruit.com/system/assets/assets/000/002/693/medium800/ledstripbjt.gif?1353964377

looks like there's a tiny short between the middle strip and the one below it

The mosfet gate is floating. You need a resistor between the gate and GND to ensure the mosfet stays off when the Arduino output is disconnected. Given that you have a 10K resistor between the Arduino pin and the gate now I would suggest a 100K resistor between the gate and GND.

thank you for the help i did check for shorts but there are none , i also tried with the 1m resistor between the gate and the GND but still the same problem. But you did get me thinking about the floating gate idea, so i put 1 mtr of lights on to increase the load and they went out, my worry is i only had a 200ma power supply on it so not sure if i will get the problem back when they hook up to the main power supply of 10A, i will check after xmas break as the wife wants to go shopping for the afternoon so i have my orders.
just seems strange how i do not get it on a bread board or if parts are soldered in a line?? only when i solder to a copper tracked prototype board
although with the lights turned off not sure if an increased power supply would make a difference,

Maybe it's some debris, but it sure looks like a short (bridge) to me.

bridge.JPG

i have cleaned up and look for it under a magnifying glass,
i have done a continuity test and there is no joint between the wiringl
i'm sure it is definitely not a short.

Attached a diagram.

IMG_1682.jpg

I plan to run 2A per MOSFET will the 100ohm resistor be enough, or should i use 220 ohm onPWM pin pleSe

normski001:
I plan to run 2A per MOSFET will the 100ohm resistor be enough, or should i use 220 ohm onPWM pin pleSe

It'll be OK either way, one or the other is not a deal-killer.
It won't "not work" because of that.
The more important part is the pull-down (10K) and, as always, getting the wiring right.

Thank you Runaway Pancake for posting the Drawing with values on.

here i have made a board to match your drawing will now fit parts and test, i will be fitting the 2nd resistor

if i only put 3 led strip on they light up faint when turned off, but if i up the load and put 1 mtr strip of 60 led's on they all go out and work correctly.
as i plan to run 5 strips at once this should be ok i hope.

normski001:
if i only put 3 led strip on they light up faint when turned off, but if i up the load and put 1 mtr strip of 60 led's on they all go out and work correctly.
as i plan to run 5 strips at once this should be ok i hope.

Distressing.

By "Off" you mean digitalWrite LOW or is that PWM 0?
It should be the same either way.
You might try a lower pulldown value (for 5K place two 10K in parallel.)

I see that you have only that one "Ground" lead (White) running off your board.
Is the Arduino running from the 12V, too?

no it comes off the board to a common ground terminal arduino 5v GND and the 12v GND.
it is not a problem once i put the 60 led strip on none come on not even the strip of three.
and all works well when i move the led up one percent at a time and back down.

normski001:
no it comes off the board to a common ground terminal arduino 5v GND and the 12v GND.

Please confirm whether it's two separate supplies (or battery packs, or whichever.)
The Arduino Vin/barrel_jack is connected to "DCPS A" and the 12V is "DCPS B"? Or what then?

The 12V supply's positive lead should go to the positive terminal of the load (LED strip) and the 12V's Gnd should go straight to the FET board, all on its own. [The load neg should go to its associated FET Drain terminal.]
The Arduino Gnd should lead out over to the FET board, too.
That's how my diagrams indicate.

normski001:
it is not a problem once i put the 60 led strip on none come on not even the strip of three.
and all works well when i move the led up one percent at a time and back down.

That's just not right because it really ought to go off without qualification.

same wiring on the breadboard does go off,
here is wiring i have.
12v + direct to led strip, 12v GND to connector block and on to MOSFET board,
arduino is powered by USB (will be on same 12v supply once in place)
arduino 5v GND to connector block with 12v GND
RED channel to 220ohm resistor to mosfet left hand pin , center pin out to led strip, right pin to ground, 10k resistor between left and right pin.
if i connect a strip of 3 led the draw is 20 milliamperes at full power and when if i get a dim light on all 3 colors, if i put a 1 mtr strip on they work great.
the green and blue are connected the same way via their own MOSFET

normski001:
same wiring on the breadboard does go off,

There there's something different.
If there's a lot of rosin between the strips then "dig" that out with a small screwdriver or similar.

If you want to look at it as humouring me then humour me - but I'm trying to get you to connect that 12V power supply ground lead (not via a jumper or a splice or anything else) right to the FET board and to run a lead from the FET board back to the Arduino ground.

ok will get on the ground lead and do as you ask,
i cleaned the board with techspray cloth from eltron electronics.
made up of 90% isotropy alcohol. so there is no residue on the board.

i have wired the 12v GND direct to MOSFET pin and i still get the same dim light when off, using only 3 led strip
i have checked under magnifying glass, and cleaned the gaps between solder joints,

Well, I'm sorry, normski, I'm baffled.
That there is this dim state when it should be off means that there is conduction through the FET in the Off-state (G = 0V).
Whether that's attributable to some Drain-Source degradation or a problem with the Gate is hard to say.
You have the 10K from Gate to Source (Gnd) so that should, so to say, "bleed off" any remnant charge on the Gate.
(I suggested trying a lower value, 5K, using two 10K resistors in parallel).

If I had it here, I'd get the voltmeter on it and look at the Gate voltage and a simple digitalWrite routine (HIGH, delay, LOW, delay,...)

Instead of connecting to an Arduino pin, what happens when you Ground the 220? resistor (the end that would go to the Arduino pin)?

Have a look at this simple fet led driver..........

i have it connected to 5 x one meter lengths now and they go out no problem, will keep my eye on it for a week or so and see how it goes. I have it dimming over a 3 hour period on each color, and starting up over a 3 hour period.