IP-OUT-08 Propeller Mosfet Firing Circuit - REV-E - UPDATED!

DavidM

HI,

I have attached a schematic of a FIRING CIRCUIT for use with pyrotechnics. ( I am only showing the main part of the design at the moment)
I have also attached other relevant spec sheets.

My main aim is to utilise the I/O Pins of the PROPELLER ( or an I/O Expander IC ) so that the I/O pins can be set to OUTPUT for FIRING or set to INPUT for detecting CONTINUITY. The reason for this is to eliminate a lot of additional components and circuitry.

The whole project will be SMD.

I have chosen the ON-SEMI , NTD3055L104 Mosfet as it is LOGIC LEVEL at the GATE , its SMD and its relatively inexpensive.

The part of the circuit for the MOSFET's is repeated 8 times, ( this can be expanded to 32 mosfets)

The POLYSWITCH is an SMD fuse from RAYCHEM, it is only rated low ( trip at 0.34amps), I have used similar to these before and they seem to work. even though I will be requireing a much higher current ( maybe up to 2amps) to pass "through" the FUSE, it does not "TRIP" until after a few seconds, by this time the IGNITOR should have fired. This fuse is mainly to prevent SHORTS from the IGNITORS damaging the MOSFET, battery etc. ( I am not concerned with this device at the moment).



GENERAL OPERATION..

When the circuit is in TEST Mode ( continuity check) , all the I/O Pins are set to INPUT and the 12V TEST Power is activated. If an IGNITOR is connected properly then the corresponding INPUT of the I/O can be read as a HIGH and reported to the PROCESSOR. The corresponding GREEN LED will also be illuminated.

When the circuit is in ARM Mode ( firing mode) , all of the I/O Pins are set to OUTPUT and then can be set to HIGH as needed to fire an IGNITOR, the 48V circuit will be activated in this mode. The corresponding RED LED will be activated while the OUTPUT is HIGH, regardless of any IGNITOR connected. The OUTPUT to high will only be for no more that 500ms.

ANY or ALL of the outputs can be fired at anytime but only one time each.
When either "mode" is activated only one of the two power supplies ( 12V or 48) will be activated accordingly.
The 48 volts is derived from a 12-48V DC-DC Converter ( I already have this circuit and its working)
The whole device will be powered from a 12V SLA 2.2AH Battery.
I have a 100K resistor from the gate of the MOSFET to GND to prevent it floating at startup.
The MODE is selected via a KEY SWITCH and each power source is "ISOLATED" from each other to prevent any "FEEDBACK" loops.


Some Questions...

Q1) Can the I/O pins of the PROPELLER or an I/O Expander IC , be used as indicated?

Q2) If I need more I/O's than I have spare on the PROPELLER IC ( which is very much the case as my next design requires 32 outputs), Then I wish to use an I/O Expander Chip via I2C, One I have my eye on is the PCA9505 40Bit I/O from NXP, any thoughts on this type of device.

Q3) Will the ZENER DIODE work to limit the VOLTAGE to the I/O pins? I have chosen the MMSZ5225BT1 from ON-SEMI this is rated at 3.0V.

Q4) I need help with some of the resistor values, Mainly R1, R2 for current limiting to the LEDS. I have guessed the values shown, and I beleive these need to be of higher wattage than normal?

Q5) According to the specification of the MOSFET the Vgs ON is min 1V to max 2V , the current is only 250uA, So is the Prop I/O capable of driving the mosfet(s), and

Q6) is the 10K resistor the correct value for protecting the I/O pin from a failed MOSFET?


Any help is greatly appreciated!

Thanks

DaveM

Post Edited (DavidM) : 5/10/2009 4:56:58 AM GMT

Brian Fairchild

I'm not sure I like the fact that with the Prop in reset, or crashed, or uninitialised or just plain broke, the igniter might fire. Current flows from +48, through the igniter, through the Polyswitch, through the zener which will act as a normal diode in that direction, through the 10k to the FET which will turn on even if briefly. The On time will be a function of how long the 100k, the gate and any stray capacitances including that of the Zener take to turn it off.

In fact, as drawn, you're putting 48v on the Prop via the Zener.

DavidM

HI Brian,

Thanks for your comments.. ( I need to think about this for a while!)


The 48V supply is ONLY AVAILABLE if the KEY-SWITCH is turned onto "ARM" ( including its ground). So on start-up this should not be a problem. I forgot to mention that the key-switch has 3 positions ( not shown in schematic "OFF - TEST - ARM "

So if I think about this, My device will "turn on" or "startup" in TEST MODE ( 12V) , this will not set off the ignitors.

By the way , The IGNITORS need AMPS to fire, I have CURRENT LIMITING resistors inline with the LEDS, that should limit the AMPS.

re: the Zener Diode,

My understanding of a ZENER DIODE is that it CUTS OFF at a particular voltage in my case the zener I have chosen is rates at 3v, is this correct?


If I use the PCA9505 Expander IC ( instead of the PROP) the I/O default to INPUTS on start up, which will help.

One thing I have just noticed is that I may need a PULL-DOWN resistor from the ZENER ( cathode) to VSS so that the LOGIC Input is seen as 0 on startup.

So If I add a resistor from the zener to VSS say 10K, and increase the resistor to the gate to 100K, then would this be naturally pulled low?

Thanks

Dave M

Leon

You might need a driver for the MOSFET, to get it to supply enough current. 3.3V from the Propeller might not be enough.

leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

DavidM

HI leon,

From reading the Specs on the mosfet I don't beleive it needs mush current, I beleive the on current is only 250uA! , But is the resistor dropping the current then?

Also, I might actually used the PCA9505/06 device it has more current and can actually run at 5V as well.

thanks

Dave M

kwinn

DaveM, you may want to take a look at this thread " http://forums.parallax.com/forums/default.aspx?f=15&m=324351&p=8 ". We are doing something similar to what you are doing here.

A quick look at your schematic brings up the following:

R2 10K seems a bit high. At 12V the led will only have about 1mA of current. I would have expected 5 to 10mA for reasonable intensity.

The prop is not well isolated from the 12/48V. If either the zener or transistor goes it looks like the prop would be toast. I would use opto isolators, or diodes to limit the voltage on the prop lines.

evanh

- Answer to question 1) Yes, since both the currents and voltages you are dealing with are trivial, there is no real need for isolation. However, to avoid glitches, you still have to be careful with board layout. And noise suppression components are always a wise addition.

- Both the I/O Expander and the Prop are fine for driving a logic level mosfet as a basic derated switch like that.

- Regarding startup issues. You have external safeties already. Use them. Maybe add a little extra watchdog circuit that ensures the ARMed mode can only engage after the Prop is alive.

- R1 and R2 don't need to exist. The LED resistors do the job already. All the current from the 12 volt test supply flows through the green LEDs. LEDs+resistors drop the full 12 volts. That's a nice 15-20 mA with the existing 510R resistors.

- No, zenor diodes don't cut off at a certain voltage, it's the opposite, they conduct at a certain voltage instead. You will destroy the Prop with the way they are atm. I'm not sure what you are wanting them to achieve in that position anyway.

- The gate resistors are too high in value at 10k. Try some more 510R instead. And put those zenors in across the mosfet's gate and source, pointing the other way round.


Evan

evanh

Ah, you will need an extra rectifier diode inline with each and every LED.

DavidM

Hi Kwinn

Thanks for your comments..

I will remove the two resistors R1 & R2, as these are not needed.

Hi evan,

Thanks for your comments.

Ok, I have figured out that I am using the ZENER DIODE incorrectly. So I will rework/scrap this idea. Instead how about just a simple voltage divider be enough?

I will add the other noise suppression components later, my main concern is to get the "LOGIC" right first.

I will reduce the gate resistor value,


I will post a new schematic soon with my next lot of changes..

regards

Dave M

DavidM

OK,

I have updated to REV - B

I have discarded the diode idea and i am now using an ANALOG SWITCH instead. This should provide 5V to the I/O INPUT pin when continuity is detected.

I will be using the PCA9505 I/O Expander as I need 32 outputs all up.
Remember that not all the components are shown, ( i.e capacitors etc)

The firing mosfets are always pulled to ground, until the I/O OUTPUT is set high.
I have also added 4 MOSFETS ( the same ones) so that the 12V for Continuity and the 48V for FIRING can be activated by the PROPELLER.
In the startup state these mosfets are pulled low to disable them.

The line thickness is indicating higher currents.

Am I getting closer??

regards

Dave M

evanh

I think I see what you want now - the test circuit to turn on the mosfet and put a full 12 volts onto the ignitors. Do you really want to do that? I don't see a real advantage over just the 20 mA circuit.

The new circuit as it stands looks out of control. The mosfet is pulling low at the same time as the test negative is pulling low which will turn off the green LED. There is a lack of ground referencing amongst other problems like glitches coming in from the test circuit.

And the supply enable mosfets are just wrong. You can't level-shift like that. Not to mention you've thrown your safety away letting the Prop control ARMing!


Evan

evanh

Uh-oh, if you are disconnecting the 48 volt GND from the rest of the circuit then you will kill all the mosfets without zenors across the gate and source pins.

Even with zenors added I don't recommend disconnecting the 48 volt GND. It removes the key path for suppressors to prevent spikes punching back into sensitive components. Mosfet gates need protection against high voltages. That goes for the IC's too. I/O expanders are all too often failing to spikes coming back up the signal path.

Optocouplers might be what you need after all.

Socketed relays are looking a whole lot easier to use ...

kwinn

If you only want to test for shorted transistors and open ignitors the attached circuit may be suitable. It only requires one extra pin on the prop to test for a shorted transistor and to verify all the ignitors are ok.

On power up and with the 48V TEST on and none of the transistors enabled by the prop the output from the optoisolator will be high unless a transistor is shorted.

By enabling each transistor in turn and checking the state of the input pin from the optoisolator the status of each ignitor can be checked.

I would recommend using an optoisolator with a darlington output. You may need to change the resistor values for optimum result, but the ones provided should be a good starting point if the ignitors can handle 10mA without igniting.

DavidM

HI kwinn,

I only need to test for open ignitors, and I need to test all of them at once, not individually, that's what my design spec calls for. but thanks anyhow.


regards

Dave M

DavidM

HI Evenh,

The test circuit is no different to before, except that I wish to turn on the 12v test via some mosfets INSTEAD of a KEYSWITCH. maybe my MOSFET Circuit is wrong, Some advice on how I can switch 12v on/off via preoppelr would be handy) but the rest of the test circuit is similar to my first version.

The ignitors are still protected from hi current because of the 510R resistor and led in SERIES. ( although this resistor may need to be say 560R and at least 1/2 watt.)

thanks

Dave M

kwinn

Is there a reason they all have to be tested simultaneously?

With the circuit shown the prop could enable each transistor in turn and verify that the ignitor is not open. The total time required to test even a hundred ignitors would be less than a second.

DavidM

Hi kwinn,

The reason is that the device will have LEDS on the front panel showing the user in realtime/live, ALL of the ignitors continuity status's,

But the continuity status can also be sent back to a remote "HOST" , that's the reason for interfacing the continuity check back to the I/O expander.

regards

Dave M

DavidM

Hi evenh,

you said..

Uh-oh, if you are disconnecting the 48 volt GND from the rest of the circuit then you will kill all the mosfets said...


how is this so?
please explain


thanks

Dave m

Timothy D. Swieter

I have worked on various pyro and flame systems and therefore I get nervous when I see threads like this. I won't comment on specifics of the circuit because I don't want to get involved in the design. However I would like to suggest that you really, really make sure you system is safe by shunting the match and disconnecting both sides of the match. You can read an interesting e-mail, research from our company here: www.birket.com/safety/tipyrostory.htm

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com - Prop Blade, LED Painter, RGB LEDs, 3.0" LCD Composite video display, eProto for SunSPOT
www.tdswieter.com

evanh

"We also use a isolated and pre-charged capacitor as the firing source for each match circuit. No possibility of cross-talk between circuits. And, we keep every match shunted until the very moment of intended ignition."

Cool! True isolation. [noparse]:)[/noparse]

evanh

David. This thread could go on for a very long time.

You need to get a better understanding of the circuits you are designing and a keener mind for guessing how it interacts with the real world. Tim has posted a wonderful example of someone truly leaning about what they've built.

DavidM

HI,

Thanks for the replies.

How about this circuit (ATTACHED IS REV-E)

1) The 48V supply to the RED LEDS is now running from 12V when the 48V power supply is activated, as this was looping back to the common positive of the ignitors. I change the resistor value accordingly.

2) I am using a ANALOG SWITCH SPDT to manage the I/O pin of the I/O expander, The analog switch will default to NC on start-up ( this is TEST MODE), and they will all be enabled at the same time via a pin from the PROP when in arm mode. So during continuity testing the GATE of the MOSFET will still be pulled low ( OFF ). The two resistors R20 and R21 provide a voltage divider to enable 5v to pass to the I/O pin if continuity is TRUE.

We will discuss the 12V / 48V Switching circuit later! assume for now that it works.

regards

Dave M

kwinn

David, as Timothy pointed out earlier in this thread working with pyrotechnics is hazardous so safety is the first concern. As the complexity of a design increases so does the chance of circuit failure and accidental firing of matches, so the firing circuit should be kept as simple as possible.

For safety reasons the circuit should be powered up in stages, preferably with a high reliability mechanical key switch. Both power and ground should be disconnected from the match firing circuit in the off position.

The arming sequence should be similar to the following steps.

1. Turn key from “OFF” to “SHORT TEST” to verify that none of the transistors or other circuit connected to the matches is shorted. A short should be indicated by a red led, and optionally a non-short by a green led. This test should be done with the same voltage applied as used to arm the system.
2. Turn the key from “SHORT TEST” to “CONTINUITY TEST” to verify match continuity. Continuity should be indicated by a green led, and optionally non-continuity by a red led.
3. Turn the key from to “CONTINUITY TEST” to “ARMED” if no shorts or open matches were found. The +48V should be turned on through a transistor to the matches only during the time ( ~ 500mS ) a match is being lit.

In my opinion the approach you are taking is overly complicated for the functionality you seek, and creates unnecessary potential hazards. By taking advantage of the power of the microcontroller you could have a simpler more reliable system.

DavidM

Hi kwinn,

well then,

please post an ACTUAL CIRCUIT illustrating your SIMPLICITY vs my COMPLEXITY.

Thanks

Dave M

kwinn

OK, will do.

DavidM

HI kwinn,

Thanks,

How about using something like this..

IPS7091

INTERNATIONAL RECTIFIER INTELLIGENT POWER HIGH SIDE SWITCH

http://www.irf.com/product-info/datasheets/data/ips7091.pdf

It has all the features I need

PROTECTED GROUNDS
SHORT CIRCUIT PROTECT/DETECTION
OVERLOAD PROTECTION
OPEN CIRCUIT DETECTION

etc

What do you think?

It would mean the LEDs are driven from another device ( LED DRIVER etc) rather than directly I think?

regards

Dave M

Post Edited (DavidM) : 5/11/2009 6:57:18 AM GMT