Specific electrical question [SOLVED]

Go To Last Post
44 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sorry for the double posts, the other topic was getting too conversationally crowded and off the main subject. I'll try to ask a direct question about a piece of circuitry that failed on me.

In one image you see the solenoid actuator. In the case of the failed relay, it was the relay that controlled the "ignition" in the car (when you turn your key all the way for about a second and the starter turns and your car starts).

Probing the circuit shows the transistor on the isolator failed. The diagrams are correct - so yes, the mosfets are connected to the voltage of the car.

I assume something is happening during the car start cycle, but I dont know what, and perhaps Im wrong.

OPTO: http://www.sparkfun.com/datasheets/Components/ps2501.pdf
MOSFET: http://www.sparkfun.com/datasheets/Components/General/FQP27P06.pdf

So the question is why did it fail, and how should I correct this circuitry to prevent failure again?
I read about hooking up a capacitor over the solenoid too... to reduce arcing. Thoughts?

Attachment(s): 

Last Edited: Wed. Oct 17, 2012 - 02:25 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Sorry for the double posts
I have locked the other thread.

John Samperi

Ampertronics Pty. Ltd.

https://www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The reason may not be in your schematics but rather in the actual layout.

Again, why do you have optos? The perform no useful service in your application. If you used logic level mosfets, you could drive them direct from the port pins. Of course, the layout of the circuitry is critical - you need to understand where your current is flowing and that it flows where you want it, not through another potentially damaging path.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Then please ignore the optos if theyre not important. I'd like to see if anyone can help me figure out why the presented configuration failed.

Here is an image of the layout. At the bottom there's a fuse... that leads to the schottky and regulator that feed the microcontroller. The thick red horizontal line (unregulated +12V from car) that goes through the relay to "YLW_WIRE" is where you get the largest current. That is the current that goes through your ignition harness. In my civic, this energizes another relay which in turn energizes the starter coil. You can see here again that the mosfet and top of the resistor bridge that goes to the opto output is connected to unregulated 12V from the car. I'm pointing that out, but I don't know if its significant.

***NOT shown*** flyback diodes exist over the relay coil: 1N4004

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You want us to divine an answer from next to nothing? My guess is your 0V track might be a problem, but since we can't see the whole picture, I can only guess. You want to keep you low current paths and high current paths separate.

Why did your opto die? As a one off, it might be bad luck, if you keep blowing them the same way, then that suggests a defect. Since the 4k7 resistor in series, I can't see how too much current could flow to damage it.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

FWIW, I would consider switching GROUND to the relay coil as opposed to the +12volts.
Less chance for reverse spikes to have to deal with on your transistor. Since you already have a PCB making this difficult to do, you could put a 10 ohm resistor in series with the relay coil and the MOSFET to cushion things a bit.

If you have an opportunity to look at an aftermarket remote car start, the relays are always activated with a switched ground signal as opposed to a switched V+. :)

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
the mosfets are connected to the voltage of the car....So the question is why did it fail...

I'll repeat what I said in the original Thread:

Quote:
You really need to post the FULL circuit, power supply, inputs, outputs, etc.

Are you familar with "load dump"?

Your circuit may work fine on the bench, but die an early death from the very noisy electrical system typical of vehicles. The micro needs to be protected from the noise, as does every input to the micro, and some outputs, if there is a path for the noise to get back into the micro's pin.

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I strongly believe that the circuit for the micro has anything to do with this. If it does, I'll be shocked. I'll post it when I get home.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Maybe so, but when you ask "what could it be?" and only give the part that is failing and nothing of what connects to it, the crystal balls get cloudy.

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Alright deal. Check it out. TVS diode, then schottky, then capacitors to hold power while engine is cranking, then regulator, then micro. output to the opto in the picture in my previous post is easily followed. nothing special.

note that before the opto failed, everything seemed to work fine. it failed over time. (car gets texts, sends back response, etc etc)...

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I should have been a little more specific and asked for the schematic. Looking at a pcb CAD is rather frustrating. No component values and labeled connections

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Upon reading the threads, unless the OPTO is still shorted when you unsolder it from the board and meter it I would say that the AVR power is going below it's reliable operation threshold long enough to cause the outputs to lock up in a funk. Can you put a meter on the AVR pin(s) to see what their level is at time of failure? If they are above 3 volts there is your answer. Why does the problem go away if you move the box could be any number of things small static charge causing the avr to wake up? I dunno. But even though you have big honker caps on the board does not mean they are going to be enough to keep your low voltage lines stable.

As I noted earlier about how other remote started systems I have seen work, I went out to the wifes car and took hers apart. After getting the evil eye, and a load of bunk from her, I took a good look. There are two ULN2003a drivers controlling the relays(switching ground).

the power supply is nothing special a 5 volt regulator, caps suppresion diode etc like yours. It also has two high current 30amp feeds coming into the unit. One a dedicated starter feed the other for ignition and lights.(the traces were followed to confirm.) The control circuit is powered by the ignition/lights feed. The sequence is, power the ignition on first, then after 2 seconds, activate the starter for 5 seconds or until the engine turns over. There is a wire that comes from the switched side of the #1 spark plug ignition coil to sense the engine running.

What I am getting at is from what I can see, your circuit is running off of one feed therefore more prone to sags and spikes as the wires resistance from heat of ignition and start will cause it to increase causing voltage sag. Without a schematic I am really shooting blind, but I would be curious to see what hapens to your regulators input voltage and output voltage at time of starter crank. do you have a DSO that could take a sample?

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

First of all, I really appreciate your input. I feel like you've got a good approach.

I'm gone for the weekend, so I'll have to tinker with things on Monday (desolder the opto). After it failed for the second time (10 minute delay before latching on my benchtop), it has been consistent. As soon as I plug it in, the relay is latched.

Probing the pins of the micro shows all outputs low - nothing is turned on. So the starter relay shouldnt be turned on either.

I have a scope, but I'd have to take it to a garage where I can plug it in. I've also never taken pictures with it to transfer to a computer... so if by chance I am able to get a garage, you might end up with some iphone pictures :)

As per your request, here is the schematic.
The order for my car starter:
1) accessory 1 on (radio)
2) accessory 2 on (hvac/key fob)
3) accessory 1 off, ignition on until engine turnover
4) accessory 1 on, ignition off

This is exactly the sequence that happens when I use my keys (the relays switch the same wires). So when you say I am "switching to power", I wouldnt use that terminology... I'm literally making the same connection that my ignition harness does... I technically dont care if it switches to ground, to negative, or something inbetween. Im just connecting the wires that connect when I use my key. Does that make sense?

Also, a note, please make sure to read the attachment file comments for more info.

Also note, the images are presented in pretty much a random order, but the naming convention is shared across all of them (wire names etc).

Also note, see next post for last pictures.

Attachment(s): 

Last Edited: Fri. Oct 12, 2012 - 04:58 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

last pictures:

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Two thoughts:

ARef gets a 0.1 uF cap to Ground.
You have it tied to +5V and that is not correct.

I do not see any By-Pass caps, e.g. 0.1 uF, across each V+/Ground pair of pins on the micro, as close to the micro as possible.

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

To be honest, the circuit is riddled with potential defects. As to what caused the opto to fail, I think I'd need more evidence. When you remove the opto, check to see if the optotransistor is actually short circuit. Why it would go short circuit is not clear. There doesn't seem to be an obvious source of over current or overvoltage. If it had been operated without diodes on the relays, then we could presume it was overvoltage that killed it. It wouldn't necessarily die immediately but with some time and heat it could die at a later time. Replace the opto and see if it happens again.

I would suggest getting rid of the opto and replacing with a charge pump circuit and a small mosfet (or transistor) to drive the main mosfet. The charge pump means you'll need to continuously toggle the port pin to keep it activated. If the micro dies or goes to sleep, the charge pump will discharge and the mosfet will turn off. Fail-safe.

Next time you do a through hole pcb, use the method of tracks on one side vertical, the other side horizontal. It makes the board a lot neater, also use a snap grid of 25thou and place all the components and tracks on the grid.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DocJC wrote:

ARef gets a 0.1 uF cap to Ground.
You have it tied to +5V and that is not correct.

Yes it is correct. Or at least, not wrong.

It is perfectly possible to feed external reference voltage to that pin, including supply voltage. You just cannot select any other voltage then.

DocJC wrote:

I do not see any By-Pass caps, e.g. 0.1 uF, across each V+/Ground pair of pins on the micro, as close to the micro as possible.

This is good suggestion for reliable MCU operation.
Anything without bypass caps may not work properly.
Every supply pin pair (VCC/GND pair) on a logic chip should have a bypass cap, so the AVR should have 2.

I also see the crystal capacitors (22pF) connect to AVR ground pin, but also elsewhere. Any noisy current flowing between caps and AVR GND pin that is caused by other devices on board will induce noise on crystal pins, and this may affect reliability.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The opto implementation in your circuit may lead to oscillation of the relay due to improper switching and related feedback of noise. I see no way to fix it, other than elimination.

My experience suggests relays are reliably switched from the low side using a logic level IRLL024NTRPBF. Use a 220 ohm resistor from gate to AVR output pin, and a 20K pull-down on the AVR pin. The relay coil is connected to +12V and the mosfet drain. The source goes to ground.

An important side note:
In may areas it is illegal to run an unattended vehicle. This helps protect lives from runaway vehicles.

It all starts with a mental vision.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I guess I'm going to have to take it apart. Your suggestions are good for general design and future projects, but in this case they are not the source of failure of the relay. As the current design stands, it has held up in the automotive environment (I mean the micro specifically).
What do you guys think about hooking the mosfets and optos to direct car voltage?

Edit. Sorry kit. Didn't see your post. I'll address it in my next reply. Gtg atm.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think your original question was about the relay getting stuck on? You'll have to examine the failed gadget. Is the device turning the relay on when you don't expect? Perhaps the relay has failed? Maybe you need a capacitor across the relay contacts.

If you don't know my whole story, keep your mouth shut.

If you know my whole story, you're an accomplice. Keep your mouth shut. 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Your suggestions are good for general design and future projects, but in this case they are not the source of failure of the relay.

Well then I guess you don't need any of us as you are saying you know what the problem is and are testing us?.

Quote:
As the current design stands, it has held up in the automotive environment (I mean the micro specifically).

And this is proven how? There is no way you can accurately say that until the problem has been solved...whic it has not. You can say you doubt the AVR is the fault, but thats all.

Before this goes off track, pull the opto and meter out the transistor side and see if it shorted and we can go from there.

Kitcarlson:

Quote:
An important side note:
In may areas it is illegal to run an unattended vehicle. This helps protect lives from runaway vehicles.

According to What I was told when I used to install car audio and remote starters is that it is illegal in all 50 states to install remote car starters in vehicles with Manual transmissions for the exact reason you mention. It is also required that remote car start systems have a sense line in the cars ignition and brake light wires. If the brake is stepped on, and the cars ignition is not turned on with the key first, the remote start shuts down as does the engine to prevent a child from accidentally causing the car to move, or prevent a thief from stealing the car.

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

:oops:
Yes, sorry. Jepael is keeping me honest.
The ARef can be tied to an external reference.
Just make sure you don't accidentally set the Mux for one of the internal references.

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Not at all! But my avr reads SMS texts, starts the car, then sends back SMS texts. So it's not like it's failing during the start job.. it's executing the program thoroughly. that's why I say the recommendations for the micro are good, but not the source in this case. Unless I'm missing something. Which I definitely could be. But if I am, please tell me what sneaky sources of failure they could b in regards to the above suggestions for the micro. What would a cap do around the relay contacts? The relay itself is not stuck, and the avr is not the source of switching.

Out of curiosity, under the assumption that the opto is in fact shorted, what would be your next steps or guesses?
Like I said, I'll have to take it apart after the weekend

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ps. Got an automatic car. The law for starters is that the car cannot be on for longer than 5 minutes

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What area of the worls are you in? Just curious.

I have no idea what a cap would do around the relay contacts. nothing to save your design is a solid guess.

From what I have posted so far I think you should switch ground and not V+ to the relays. This will reduce the voltage spikes the relay coils induce when de-energized. I have not looked at what the max the OPTO can handle across the Collector-emitter junction before failure is. Of course that should be irrelevant as you are driving a transistor with it IIRC.
I'll peek at the schematics a little better.

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Cap across contacts? Suppresses arcing if there is any.

Since he won't investigate the failed device or follow any suggestions, there's nothing to do but guess in the dark or ignore him. Tried guessing in the dark.

If you don't know my whole story, keep your mouth shut.

If you know my whole story, you're an accomplice. Keep your mouth shut. 

Last Edited: Fri. Oct 12, 2012 - 10:10 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

From what I can see, it looks like that the current to drive the starter solenoid(we call it the Bendix here), comes from the board supply, that’s a lot of current, that will give you some serious voltage spikes(load dump) when the relay switches off.

You need to separate those supplies, your board should only feed the onboard relays, nothing else outside of it(or with a separated and fused supply input, isolated from the rest of the board).

You could maybe try it by isolating(cutting) the big red pad that supplies the relay contacts with 12V from the rest, solder a heavy wire on it, fuse it and connect it directly to the battery(or in series with another relay, which supplies the car ECU and ignition coils, etc, as a safety) . Then solder wires on the opto’s and resistors(and anything else for that matter) that where connected to that pad(big red) with wire, and connect it to the boards +12V supply.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Kewl. Good to know all this. Well first things first - Monday I'll take out the opto and see what's happening there.

Torby - wtf?

Fleemy - thinking... Isn't the back emf related to the coil of the relay I'm switching? Or are you suggesting that the starter coil is inducing an even larger back emf, which my circuit is exposed to? If the latter, that intuitively makes sense... I will scope what's happening during crank. I wonder if simply connecting the positive side of the coils to the portion of the circuit after the tvs diode would suffice. That way the voltage spikes would never go above some 28ish volts. Does that make sense what I'm saying?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Which coils, the onboard relay coils? I don’t think that is the problem.

Just separate the common feed(+12V) to the relay contacts, from the rest of the board electronics supply. Use one big wire to feed all the common contacts, preferably fused, straight to the battery.

In the future, just switch to ground and forget about those opto’s, in general I use a ULN280X to switch relays, it can sink 500mA per output which should be enough for most relays.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

jgmdesign wrote:

Quote:
........It also has two high current 30amp feeds coming into the unit. One a dedicated starter feed the other for ignition and lights.(the traces were followed to confirm.) The control circuit is powered by the ignition/lights feed.

I would bet the 1.2kw starter in my wifes car sends a spike. :)

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Well youre not going to like this.

I took out the opto, and its working fine again. This is the second time it has come back to life. I want to re-iterate here that the input to the opt from the microcontroller WAS OFF.

Thoughts regarding this?

Fleemy: I was referring to the starter coil itself. I imagine a lot of current goes through there (fused at 50A) during start, so that could send back a pretty big voltage spike. No/yes?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If the AVR output to opto input was off, you have proven that is not the problem. Now what did you see on opto output or FET gate? Is the opto output near ground or 12V? That would pin-point the problem!

In my opinion, you drive the FET gate with very large resistance, which can cause a lot of issues. For instance, the FET turns on and off very slowly. In fact if you have a loose connection on the FET gate to source resistance, the FET won't turn off until some small leakage current has discharged the gate charge. Large gate resistance can also lead to all kinds of ringing problems, because of FET capacitance between gate and drain. And the gate can only handle +/- 25V with regards to source voltage.
Normally the FET gate is driven by quite stiff voltage source which can drive the base with substantial current to quickly charge or discharge the gate capacitance.

Is it possible to explain this by a latch-up condition? Do opto transistors or FETs have such a condition, when stressed with large spikes?

I'd be tempted to probe things with an oscilloscope what kind of spikes happen, but I'd be very afraid the oscilloscope dies because of these spikes.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hmm. Does anyone know of any latch-up conditions could exist here?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Could be, and this is a longshot that an oscillation is being created whre the spike from the starter is enough to get it going. It may not happen every time but only when conditions are right. One way is to use a scope probe whi;e the latchup has happened and poke around. Another is to place your finger on some of the connection points around the area in question while grounded to the car chassis and see if the probelm goes away. Not guaranteeing anything but worth a shot.

tkurowski,
After reading through the last 30+ posts in this thread I am guessing that you this project is on a PCBoard and not a protoboard hence why you have either ignored the idea of redesigning parts of it, or replied that it's not the issue. While I can certainly understand the logic behind wanting to find out the cause of the issue, there comes a point where rather than banging your head against the wall, responding to a post by saying that an idea is wrong when you have nothing to prove otherwise, and lastly coming out and finding a dead battery you might want to instead incorporate the suggestions on how to build your mousetrap better to save your sanity.

Besides, I don't know where you are located, but it's gonna get cold soon and it would suck to be looking for jumper cables on a cold night. ;)

EDIT:
I did not mean to place your finger on some of the spots while the engine was running or every time you try to remote start your car. I meant to do the finger press when the relay sticks and see if the latchup releases.

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I am guessing you have some sort of GSM modem thingie to recieve messages from your phone to start the car. Where is that on your schemataic, and where is it located on your PCB?

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I have another really good clue!

I honestly cant tell if its the mosfet or the board, because I don't think I am even flexing the board. What do you guys suggest I check here?

It only happens to the first relay position!

And in this state, the resistor bridge is open on the opto output side.. that is it SHOULD be high... but it drops!

http://www.youtube.com/watch?v=Dbj9YJfcAQU&feature=plcp

also... i think it must be the board?

http://www.youtube.com/watch?v=D90vCQmCx_I&feature=plcp

if you listen with sound you can hear when the relay makes or not.

im probing the gate, the schematic of which is:

Vcar----22K---o-----4.7k---OPEN
              |
              |
             GATE
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

jgm:
true. i could just get some transistors and hack the board so that I switch to ground and get it over with :) believe me, if the clues from the above post dont further help (they are good clues!!!!) thats the next step!

but alas, i do really want to know why :(
I appreciate the help youve put forward so far. you have good debugging.

GSM module: yup. its connected to the 12V after the schottky. its a demo board, so i give it 12V, it handles the rest. the only connection to it is Gnd Tx and Rx to the micro, both operating @5V.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

wow i think i found it.

the 4.7k resistor was touching a solder lump. not the lead, the resistor. this happened once already a long time ago on this same board. i tried making sure the resistors touched nothing! i moved the resistor away from the solder spot (which by the way was GND). i flexed the board a lot and couldnt get it to switch the relay. then, since i separated the resistor from the solder by less than the width of a solid wire, i pressed a wire between the resistor and the solder. when i pressed hard enough, it made the connection and the relay switched.

lets hope that was the final answer. due to the fact that my car has been drained twice because of my brave experimentation, I bought a battery back up :D
so I can give'r as much as I want hahahaha!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Why do you have a resistor (4k7) going nowhere(open)?

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

It was open after I desoldered the opto! The 4.7k goes to opto to pull the gate voltage down

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ok, well ass long as it's fixed, and your happy then great.
Please change the titlr of the thread by putting the word SOLVED! in it so others know there was a resolution.

Congrats.

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

and I'd personally like to thank YOU. great ownership! and thanks for taking apart your wife's car starter hahaha. I look forward to chatting with you again soon!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Over and out! ;)

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You are running your mosfets on the auto supply line. That's very very noisy. Any stray capacitance on the gate of your power mosfet is going to be bad. Take a look at the equavalent circuit of a n-channel MOSFET device and you will see that parasitic capacitance on the gate could induce HUGE current flows. This could be why you are frying your opto-isolators. Is the coil current on your relays so large that you need a MOSFET to control it or would it be better with a power transistor or even UJT / SCR combo? A TIP140 will source up to 10A continuously at 125W, how much coil current do you need?