Maker Pro
Maker Pro

Is this mosfet good for my project?

Xenobius

May 15, 2012
125
Joined
May 15, 2012
Messages
125
By any chance, could you look at the oscilloscope settings? are they correct?
Is it because it was set to AC? Im using DC!
 

Xenobius

May 15, 2012
125
Joined
May 15, 2012
Messages
125
Oh I am using a power supply and the ground is connected to this power supply. I have common ground troughout the circuit.

How can I use a ground compatible with the oscilloscope? Hmmph this is a computer oscilloscope and I dont have other options... but I do have a reference pin on the DSO2090 ... maybe that can be used?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
By any chance, could you look at the oscilloscope settings? are they correct?
Is it because it was set to AC? Im using DC!

Hahaha, yeah, that would do it. Set it for DC.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Oh I am using a power supply and the ground is connected to this power supply. I have common ground troughout the circuit.

How can I use a ground compatible with the oscilloscope? Hmmph this is a computer oscilloscope and I dont have other options... but I do have a reference pin on the DSO2090 ... maybe that can be used?

if you don't see sparks fly when you connect the grounds together, you're OK,

You can start by measuring the potential difference between the grounds. It should be very small.
 

Xenobius

May 15, 2012
125
Joined
May 15, 2012
Messages
125
Hi all again,

I think things have gotten more complex now. I have several issues which are putting me off this project. I am kind of lost here!

So the attached circuit schematic is not alone in the circuit. It is replicated for 8 times and so I have it identically in the same circuit connected together only by 12V Test, 12V Live and Ground.

1. When the circuit is on 12V Test (no ignitors) the green LEDs all light up dimmly
2. Same with 12V Live and no ignitors (the red LEDs all light up dimmly.

The above are understood but have no solution HOWEVER the following is a mystery:
3. When the system is on 12V Test and I insert JUST ONE ignitor, ALL 8 green LEDs light up bright! What is going on? I mean the only thing which is common to all 8 circuits is supply... how come one ignitor is affecting all 8 outputs?

Thanks for en-lighting me! :)
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
So the attached circuit schematic is not alone in the circuit. It is replicated for 8 times and so I have it identically in the same circuit connected together only by 12V Test, 12V Live and Ground.

OK

1. When the circuit is on 12V Test (no ignitors) the green LEDs all light up dimmly
That is correct, you will see some slight conductance via the resistors as I pointed out earlier.

2. Same with 12V Live and no ignitors (the red LEDs all light up dimmly.
Again, as expected.

The above are understood but have no solution
Think of it as a dim LED indicating power but no ignitor. (it's a feature :))

HOWEVER the following is a mystery:
3. When the system is on 12V Test and I insert JUST ONE ignitor, ALL 8 green LEDs light up bright! What is going on? I mean the only thing which is common to all 8 circuits is supply... how come one ignitor is affecting all 8 outputs?
The answer is, it shouldn't, and I would suspect a wiring error.

How have you constructed these?

The first test, with all power removed (and no ignotors) is to measure the resistance between a pair of connections for ignitors (the non-ground ends). The resistance should be in the order of 60k
 
Last edited:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
What is the state of your O1 input? I presume either open or pulled low.

There is a potential problem with the mosfets turning on when the real 12V signal is not active. They could turn on.

This is not a problem for a single ignitor, but may be the issue with multiple ignitor circuits.

My recommendation is to place a diode in series with the source of the mosfet. This will prevent any reverse current flowing.

edit: a simple test would be to insert a (say) 47 ohm resistor in place of the ignitor (again the LEDs should light up) but then change over to the 12V supply for just this channel. If I am right, the red LED will light up for this channel, and the other green LEDs will all turn off (well, back to dim).
 
Last edited:

Xenobius

May 15, 2012
125
Joined
May 15, 2012
Messages
125
Hi Steve,

I attached the layout relevant to our discussion and I marked out several components to make it easier to look at. I will try the 47ohms soon.

Thanks once again
X
 

Attachments

  • layout.jpg
    layout.jpg
    158.8 KB · Views: 104

Xenobius

May 15, 2012
125
Joined
May 15, 2012
Messages
125
The first test, with all power removed (and no ignotors) is to measure the resistance between a pair of connections for ignitors (the non-ground ends). The resistance should be in the order of 60k

Ok I checked all 8 ends. (lets number them from 1 to 8 for simplicity)

Test1: Resistance between the non ground ends of each connector
Between 1 and 2 I had 55k resistance (these 2 pins I was testing with
Between the rest of the pins I had 70K resistance

Test2: Resistance between ground and each individual connector
While I was there, I also checked each end against ground this time.
I found 1 to have 20k resistance to ground while all the rest had 35k resistance.


The difference in resistance bothers me to be honest. Could it be that something is deteriorating on use? I mean obviously because of a bad design or something... I did check if something is overheating but I was meticulously careful when choose power ratings and I even chose a more powerful rating just to be on the safe side.
 

Xenobius

May 15, 2012
125
Joined
May 15, 2012
Messages
125
edit: a simple test would be to insert a (say) 47 ohm resistor in place of the ignitor (again the LEDs should light up) but then change over to the 12V supply for just this channel. If I am right, the red LED will light up for this channel, and the other green LEDs will all turn off (well, back to dim).

OK so I done this. I inserted a 47ohm resistor in place of one ignitor. What happened is this.

FIRST BOOT: All green leds were dim. then I switched the relay so the greens were off and the reds were dim. As I inserted the resistor, that specific led lit up properly (as I wanted it to be)

I removed resistor again

ON RESET OF ARDUINO but not shutdown power: All green LEDS were dim. As I flipped the relay, the red LEDs were on as always EXCEPT for the one used with the resistor! This was absolutely OFF...

I removed power, waited 10 seconds and started up again AND back to FIRST BOOT...

I think in the code I should be clearing the shift registers... I thought that once powered off they will loose their state... I will try this and let you know soon.
 

Xenobius

May 15, 2012
125
Joined
May 15, 2012
Messages
125
Hi again,

Ok so no reply = too complex.

Maybe I've complicated life too much. How would you fix the leaking current in the LEDs?

Should I make a transistor which will allow the LEDs to light and solve the leaking current issue? I could make a pullup resistor which is pulle down when an ignitor is inserted switching on a PNP transistor which will allow the LEDs to light up.

Secondly why do you think ALL outputs are being effected? Could it be something related to the shift registers? Could one output be effecting all the others due to a wrong configuration of the registers? Im using 74HC595 and 74HC165

Thanks all
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
no reply because you said you were off to try stuff.

I recommend you get the circuit operational with NO additional stuff on the inputs. Clearly you have issues with this too, and it's hard to get several things fixed at once.

the easiest way to fix the LEDs is (probably) to place a diode in series with the anodes of the LEDs and then place a 2k2 resistor across the LEDs

The diodes will prevent current flowing back into the open 12V supply (may not be required). The resistors across them keep the voltage across them lower than their turn-on voltage when they have 35k to ground.

Breadboard it first.
 

Xenobius

May 15, 2012
125
Joined
May 15, 2012
Messages
125
Great I will use a bread board to test the LEDs first then.

Thanks Steve :)
 

Xenobius

May 15, 2012
125
Joined
May 15, 2012
Messages
125
the easiest way to fix the LEDs is (probably) to place a diode in series with the anodes of the LEDs and then place a 2k2 resistor across the LEDs

The diodes will prevent current flowing back into the open 12V supply (may not be required). The resistors across them keep the voltage across them lower than their turn-on voltage when they have 35k to ground.

Breadboard it first.

EDIT ALL TEXT:

Ok so I tried this on a bread board and it worked perfectly. I had 1x BI-Led (as I am in fact using in my circuit). I used a 100 Ohm as an Ignitor (even if the real ignitor is just 4ohms).
I had NO 5v1 Zener and No Capacitor.

When I tried the 2k2 resistor in the original circuit, the LEDs were ALL LESS dimmer than usual (meaning one LED is still effecting the rest) and the led which had the 2k2, was flickering abruptly. I am assuming its until the capacitor is charged? In fact after around 1 or 1.5 seconds, the resistors regain the original dim brightness and the led with the 2k2 is back to dim.

So I said, if I can discharge the capacitor quicker, it wouldn't be visible so I replaced the 2k2 with 1k resistor and it looks like its working.
I will connect an oscilloscope and see what's really happening but visibly to the eye, its gone and is how I want it to be.

Is this correct? Is this safe?
 
Last edited:

Xenobius

May 15, 2012
125
Joined
May 15, 2012
Messages
125
Ok I'm impressed. I just fitted ALL 8 outputs with 1k resistors on both LEDs (8x outputs * 2 LEDs = 16 resistors) and its working flawless! They do not flicker, they do not light up if no Ignitor (100 Ohms for now) is inserted, they do not interfear with each other either. If I insert an ignitor and turn the system ON from cold only one LED lights up as expected (unlike as before).

What do you think of this? What I'm asking is for a possible answer as to WHY this was happening because really and truley even tought you and I were thinking it was a wiring problem, it wasn't. I still would like to understand why so I learn but I am a but clueless because the only thing that interconnects between the 8 circuits is 12V test, 12V Live and Ground.

Thanks once again all :)
X
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Removing D4 and C6 shouldn't have had *that* much effect.

Are you sure you has D4 around the right way? and c6 was really only 22nF (not 22uF?)

Anyway, good result.

It should be safe as it is. Any spikes that get to your output will be current limited by the voltage divider so should not cause damage to the inputs of your microcontroller (assuming that's what you're doing)
 

Xenobius

May 15, 2012
125
Joined
May 15, 2012
Messages
125
I can assure you the capacitor is 22nF however the diode I will have to check. I know this is silly but the catode is the one with the LINE right? In that case, it should be correct but will check once again as I can never be too sure.

The INPUT is going to an input shift register which is then connected to the microcontroller.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Yeah, the band is on the cathode end. Anode to ground in this case..
 

Xenobius

May 15, 2012
125
Joined
May 15, 2012
Messages
125
Yep, so it is then.. confirmed that diode is correctly soldered.
 
Top