Is this circuit drivable by a Propeller?

Jay Kickliter

I found this circuit on the internet for cutting away a balloon. Is it appropriate for use with a Propeller? I don't know much about mosfets. Should I have a series resistor with the Prop's pin to limit current? Also, what is the 15k resistor for?

thanks,

Jay

T Chap

The 15k is to hold the gate low. If the Prop were to reset for example, there would be some time required for the pin to be reset as an output to hold it low, during that time the gate could float and allow the low side to conduct, causing a false trigger. The Prop's default pin state is an input until programmed as output, so driving anything that requires definite high or low requires pull up/pull down resisters.

The Prop would be suitable for this, assuming the gate will operate at 3v3.

Jay Kickliter

Thanks. Do you think I need a series resistor? I'm attaching a copy of my schematic. I just made a bunch of changes, so it's a bit messy right now. The cutdown section is at the top-center.

T Chap

I think you need a pull down on the camera as well, but obviously it is hard to tell what is on the other end of the connector, you may have it managed over there.

Someone else may know whether a series gate resistor is required for logic level fets, I would guess not since it is designed for a logic input, but some low value gate resistor (100-200) may be a good idea for short protection in case the fet gets fried.

Side note: You could always use labels instead of nets to tidy up drawings that start to get cluttered.

Make sure you don't actually need pullups on both SCL and SDA.

Post Edited (TChapman) : 11/21/2008 10:44:26 PM GMT

kjennejohn

I'm not Prop literate, but doesn't the Prop need a pullup resistor on the Reset line? I don't see one in the schematic.
kenjj

Jay Kickliter

Thanks for pointing out that the SCL doesn't have a pull-up. I looked up pull-up's on Wikipedia and it explicitly stated that I2C needs them on both SDA and SCL. But it has been working fine without it and the schematic I copied was made by Parallax and only had a pull-up on SDA. I wonder why they left it out, and why it works without it.

I haven't noticed the camera triggering by accident, but I will had a pull-down. Any idea on a good value, 5k?

T Chap

5k, 10k, should be fine for a one shot. You may not have a problem with false triggers in your design, but it is good practice to not let inputs float like that.

Jay Kickliter

kenjj, does it? I've found this basic schematic Parallax put out has what a lot of people have told me are a bunch of bad design practices. I wonder why they put it out?

T Chap

You could get the manual and read whether the reset is internal pulled up.

sylvie369

For the "igniter", be sure that you understand the difference between low-current "ematches" and high-current igniters.

blastzone.com/pml/components.asp?groupid=18

Those terms are sometimes confused, used interchangeably, and so on. If your circuit is designed for low-current devices, it's not going to light high-current igniters. Is your plan to use the igniter to light a small BP charge that severs a line from the balloon to the parachute/instruments? If so, either ematches or igniters will work. Ematches are far easier to light, but far harder to get your hands on.

Obviously ground testing is necessary.

Jay Kickliter

sylvie369, I ripped this circuit off of a web page of a similar project to mine. wiki.ukhas.org.uk/ideas:flight_support#pyrotechnic_cutdown_device. I'm not sure how you would qualify this igniter. But the system is elegant in its simpleness.

Post Edited (Jay Kickliter) : 11/22/2008 12:00:37 AM GMT

Peter Jakacki

Some notes:

1) Unless you are running Multi-master I2C (never ever seen it and the software gets complex) you do not need a pull-up on SCL as this is always a single output driven by a single master.

2) Pull-down R10 on the MOSFET is good but if you are going to use a series "safety" resistor R7 then put the pull-down on the logic side. This is a MOSFET and not a bipolar, you want the maximum gate drive possible but with the benefits of an guaranteed pull-down.

3) The BUZ11 is not logic level compatible unless you only want to switch 1ma. From your circuit the transistor only dumps the charge from C9 (which is?) at 5V so there are many small bipolar transistors far more suitable.

4) With BOE tied low the reset is a weak output with around 5K pull-up internally, no need for additional pull-ups.

Don't rely on what books and various sources say, all things need to be taken in context which requires digging a bit deeper and coming up with the answer to why.

*Peter*

Jay Kickliter

Thanks Peter. It is actually a P36NF06L transistor.

sylvie369

Jay Kickliter said...
sylvie369, I ripped this circuit off of a web page of a similar project to mine. wiki.ukhas.org.uk/ideas:flight_support#pyrotechnic_cutdown_device. I'm not sure how you would qualify this igniter. But the system is elegant in its simpleness.

Agreed. However I think that the author of that page is using the word "igniter" to refer to what looks like an ematch, and the comment that "the igniter needs about 2amps @ 3Volts for just a few milli seconds to fire" supports that. You may want to contact the high power rocket guys in your area for assistance with this part. I may be wrong, but I think that circuit isn't going to fire the kind of igniter that you can just pick up in a local hobby store. I'm tempted to test it myself - if I put one together and try it, I'll let you know how it comes out.

kjennejohn

Jay, I found the time to try and answer my own question about the Reset pin on the Prop and discovered: I can't rightly say. Every reference Parallax offers on the chip says the BOEn (Brown Out Enable pin) determines how the Reset is handled. If BOEn is pulled low, the Resn pin has an approximate 5K pullup internally which allows external monitoring. Thus it puts Vcc out and keeps itself pulled high. This can be overcome with a simple logic low, or switch to ground. All the schematics they offer I can find show this is how they handle the two pins. You laid out the circuit this way so you are good to go.

If BOEn is pulled high, the RESn is a CMOS input with Schmitt trigger. Period. None of their schematics I've found are wired this way. This, to me, suggests an external pullup would be needed. This is just for future reference.

Later!
kenjj

Franklin

I haven't read all the posts so I'm sorry if this has been said already but Nuts and Volts has a running set of articles called "Near Space" that covers many of the issues of baloon data logging and such. They even had a way to take your data and create a flight path in Google Earth. See if you can get your hands on some of them.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- Stephen

Timothy D. Swieter

I believe with other balloon cut down designs I have seen they used sizzle cord and nichrome wire. Any reason to not use that design here? I think you turned on a voltage to the nichrome wire for it to heat up and burn through the cord.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com - Prop Blade, LED Painter, RGB LEDs, uOLED-IOC, eProto for SunSPOT, BitScope
www.tdswieter.com

sylvie369

Timothy D. Swieter said...
I believe with other balloon cut down designs I have seen they used sizzle cord and nichrome wire. Any reason to not use that design here? I think you turned on a voltage to the nichrome wire for it to heat up and burn through the cord.

Rocket guys sometimes use something to burn through a piece of fishing line, which is strong enough to hold things together but also more likely to burn through cleanly than some kind of braided (nylon?) cord. I've not heard of "sizzle cord", but I'll take a look at the forum search for references.

Jay Kickliter

Timothy, no particular reason. I did want keep everything on one board, and not have to introduce a new PCB or battery. As it is now I had enough on my current board for a large cap and a mosfet. If this doesn't work I can try the nichrome wire.

Chris Savage

Jay,

I don’t know if you ever got the answer you were looking for or not…on your circuit, whether or not the Propeller will drive it depends on the gate voltage of the MOSFET you’re using. The datasheet should provide that information in the form of V_GS. You don’t need a series resistor on the MOSFET.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Engineering

Phil Pilgrim (PhiPi)

The datasheet for the that transistor specs the transfer characteristics at V_GS=5V, but doesn't provide any graphs from which to discern performance at lower voltages. Without knowing what the actual current requirements are, but assuming the P36NF06L in the circuit shown is up to the job, the safest approach might be to drive it via a 74HCTxxx buffer powered from 5V.

-Phil

Jay Kickliter

You guys are right. It looks like the circuit I found on the Internet was meant to be used with 5v logic. Instead of adding another part to drive it, does anybody know of a MOSFET that can switch about 3 amps with 3.3 Vgs? Admittidly, I don't yet understand how mosfets work differently from say a 2n222 npn transistor.

Phil Pilgrim (PhiPi)

The IRF3708 is characterized for a V_GS of 2.8 volts (R_DSon = 0.029 ohms) and will handle the required current at that gate voltage.

-Phil

Peter Jakacki

In these low voltage applications it is easier to use a bipolar transistor. Even the ZTX689B will handle 3A continuously (8A peak) in it's tiny TO92 package because of it's hiigh gain and low Vce saturation. Bear in mind that the load is only momentary anyway. So this particular transistor has a gain of 400 at 2A loads so you can drive it very easily from the prop with nothing more than a low value resistor of around a couple of hundred ohms. KISS.

*Peter*

search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=ZTX689B-ND
(The spec summary at Digikey is incorrect, read the datasheet, mentions a 2A figure only for reference to typical gain)

P.S. There are other higher current variations of this transistor also.
search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=ZTX869-ND

Chris Savage

Phil,

Thanks for the find…this will be a nice one to use in Propeller applications.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Engineering

Jay Kickliter

Thanks everyone for the help.

Peter, I like the idea of using the smaller transistor. I didn't know such small transistors could handle that current. But, I think I'll go with the MOSFET. Insead of locking myself into one cutdown design. I'm going to add a jumper and the abilty of firing an emach from the capacitor, or just use the MOSFET as a switch that I can connect in series with a Lith 9v and some nichrome wire.

Jay Kickliter

Ok, I think I have a final design. I'm going to test it of course when the parts get here and before I send the board files off to be made, but a quick review buy you guys would be great. I cleaned up the schematic a bit so it's easier to read now.

Peter Jakacki

I'm not a fan of bulky components so I found this part in a small 8-pin soic pack. Check out the specs on this one!

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=IRF7463PBFCT-ND

You should allow for pullups on the SD card, some cards need them. It mightn't be a bad idea if you allow for at least a 10uf tantalum across the SD card supply as they can draw quite a bit of current at times and this might normally cause the VDD to dip badly.

R3 can easily be increased to 10K as it looks like an input from a +5V signal. You will find that there might be too much current that flows into VDD during low-power modes like RESET and the VDD gets pushed up to the chips breakdown voltage. The same goes for any inputs from a higher voltage, the only disadvantage of a higher value would be a limitation to the pulse response but I'm talking more like megahertz it's not a worry here.

I like to keep VDD clean and not too heavily loaded just to minimize problems as it all depends upon the prop having a clean supply.

Also, I favor sinking LEDs to +5V rather than sourcing current from the VDD supply. Admittedly my +5V supplies are usually switch-mode (why not consider this) so they don't have a problem with supplying the current. Because the minimum voltage drop of an LED is at least 1.6V the prop doesn't have a problem with them connected to +5V. Plus this also allows blue and white leds to be operated as well.

That remark about the series resistor not being required for MOSFETs is correct but it is good practice to allow for a resistor for safety reasons (the prop's), not because it needs it. This is especially true if there is enough inductive kick back from the drain via the internal drain-gate capacitance or if the device fails.

Other than these few remarks the only other thing to watch is the pcb layout.

*Peter*

Jay Kickliter

Somebody said...
Also, I favor sinking LEDs to +5V rather than sourcing current from the VDD supply. Admittedly my +5V supplies are usually switch-mode (why not consider this) so they don't have a problem with supplying the current. Because the minimum voltage drop of an LED is at least 1.6V the prop doesn't have a problem with them connected to +5V. Plus this also allows blue and white leds to be operated as well.

Peter, could you expand on this a little? What if I need to switch LED's on and off in code, I have no choice but to hook them up to the prop. And I was planning on using blue LEDs, what makes their application different?

Phil Pilgrim (PhiPi)

Jay,

With the Propeller, there's absolutely no sound reason to choose current-sinking (anode to Vdd) over current-sourcing (cathode to ground) for driving your LEDs. The reason is that the Propeller's outputs are symmetrical: they can source and sink current with equal facility. Not all microcontrollers are designed this way, however. Some controllers can sink more current than they source, in which case a current-sinking output would be preferrable to your current-sourcing drive.

What Peter may be referring to with blue LEDs is the fact that, unlike their red and green brethren, a blue LED's forward voltage is often more than 3V. (Be sure to check the specs.) This would typically make a 3.3V source unsuitable for driving them, since there's so little headroom for the current-limiting resistor. Under such circumstances, it may be better to connect their anodes to +5V and pull them low with the Propeller pin via a current-limiting resistor.

But this applies only if you intend to drive the LEDs at their nominal operating current. At lesser currents, a blue LED's forward voltage will drop to well below its nominal value. For example, an LED rated at a 3.25V forward voltage when drawing a nominal 20mA, may have a forward voltage of 3.0V when drawing 3mA. Under such circumstances, driving it with a 3.3V output through a 100-ohm limiting resistor will work just fine.

I assume you're planning to use them as indicators anyway, not to light up a room. Right? OTOH, if you're using them as a strobe beacon, you'll want to drive them through an open-collector transistor driver, in order to sink more peak current than the Propeller can provide.

-Phil

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Just a few PropSTICK Kit bare PCBs left!

Post Edited (Phil Pilgrim (PhiPi)) : 11/29/2008 2:03:00 AM GMT

Jay Kickliter

Phil, thanks for explaining. I have to admit I am more lost. I read the Wikipedia article on current sourcing/sinking. It didn't help much. I'm attaching a schematic of what I'm guessing you are suggesting. Tell me if I'm completely off. Or better yet, if anyone can point me to a simple schematic that explains this graphically, that would be wonderful.

Thanks everyone for the help. I'm amazed every day at how much I don't know. I was an electronics tech in the Navy for many years. But now that I'm starting to design stuff I realize how little I actually knew.