Arduino Mega problem... FTDI chip overheats

Hi guys,

Well let me start by saying I have two 1280 Mega boards, both have the same problem meaning it's unlikely it's a defect in the board.

I have 3 servos connected to 3 digital pins, 2 motors connected to the board, a force sensitive resistor, and a Wii Nunchuck to control the servos and the motors - all connected to the Mega.

When the Wii Nunchuck is connected by itself on the Mega it works perfectly fine. As does everything else individually.

When I connect everything together excluding the Wii Nunchuck, everything also works fine. It's not until I attach the Nunchuck in the 3V3 pin that things go odd.

As soon as I attach the Nunchuck when the Mega is connected to my computer it disconnects itself from my computer (you hear the 'beep' noise that Windows makes when you disconnect a device) and the FTDI FT232RS chip gets extremely hot, it is NOT meant to get this hot at all.

I'm not sure where the problem lies... I have the idea that the Nunchuck is drawing too much current away from the board?...

Any help and advice how to fix this would be AMAZING.

Thanks a lot, Laura.

I have the idea that the Nunchuck is drawing too much current away from the board?...

That is most likely correct, the (1280) Mega boards draw all of their 3.3v power from the FTDI chip and this can only supply about 40-50mA. Not sure how much the nunchuck requires but I wouldn't be surprised if it was more than this.

Mowcius

Oh wow okay, I didn't know that the 3V3 drew power from the FTDI chip.

Is there any way which I can avoid this problem? I mean plugging the nunchuck into the 5V would work if your conclusions are correct I suppose?... although I do not want to damage the nunchuck as I've read that too much voltage can shorten the life/ruin it. Any other ways?

Thanks again.

EDIT: A member from another forum has recommeded that plugging in a 9V battery to the barrel jack may make a difference?

...Also, just tried using the 5v to power up the Nunchuck instead of the 3V3 with everything still connected to the Mega, still disconnects itself from the computer however the FTDI chip does not get as extremely hot as before when it was connected to the 3V3. HMM.

I would recommend a regulator to power the nunchuck seperately.

EDIT: A member from another forum has recommeded that plugging in a 9V battery to the barrel jack may make a difference?

Shouldn't do.

Mowcius

Add in a 3V3 regulator? I have one however I have no idea how to do it. Any suggestions for good tutorials? I'll also have a google around.

The best solution would be to create a voltage divider using 2 resistors from the 5V line on your Mega.

Depending on how much current the nunchuck requires will affect the best resistor values to use, but the ratio will need to be pretty much 2:1.

Just ensure you get them the right way round - a multimeter is your friend but if you get it the wrong way round, I suppose you'll just end up under-powering the nunchuck with about 1.6V~ :)

Hm okay, so basically, quoted from PaulS - "With a voltage divider: +5v ---/\/\/\/--- + ---/\/\/\/--- Gnd * R1 | R2* * |* * +3.3V* The voltage drops to Vin * (R1/(R1 + R2)). Use any values of R1 and R2 that add up to about 10K. "

What values would I be looking at do you think? Sorry by the way, I'm not the best at remembering electronics!

The best solution would be to create a voltage divider using 2 resistors from the 5V line on your Mega.

No you should never never NEVER power stuff using a potential divider. >:(

The best way is to use a three terminal regulator from the 5V supply.

Add in a 3V3 regulator? I have one however I have no idea how to do it.

Connect the Vin of the regulator to 5V, connect the Vout of the regulator has the 3V3 voltage on it. Connect the ground of the regulator to the ground of the arduino. Then you need a 0.1uF capacitor between Vin and ground and another between Vout and ground.

Thanks a lot, Mike.

Will try hooking this up tomorrow (& post results) as it's now 11:40pm and I have a 6am start tomorrow, eek.

Thanks all for your help, I would give each of you a 5 minute hug had you live in NZ.

Laura.

Mike,

In regards to

No you should never never NEVER power stuff using a potential divider.

Can you explain why rather than just outright dismissing it please?

Richard,

The changing load changes the voltage output. It is simply ill suited for circuits with any current going through them. (Unless you have some new insight nobody else has discovered in the last 100 years.)

Not at all - I've used potential dividers to power low voltage sensors but admittedly I have never had anything like the load that Lawrah is talking about, and never really used anything in that configuration for any period of time.

The reason I asked Mike to expand was purely to learn for myself - he said don't do it and suggested an alternative, but didn't say why it was a bad idea.

Now I will take that onboard and learn from it - something I couldn't do without it being expanded upon.

As for the sarcasm, that was just unnecessary. People come here to learn, and elitist comments like that just turns people away.

Morrolan

Can you explain why rather than just outright dismissing it please?

Ok here is the reason. A potential divider only works because the two resistors are just that a constant resistance. Any load you put on the bottom resistor is in parallel with it and changes it's value. Therefore the voltage of the load varies with the current the load takes. This is the sort of supply noise that makes things not work.

http://www.thebox.myzen.co.uk/Tutorial/De-coupling.html

Those with long memories might remember TV sets that were powered by large potential dividers that got very very hot. This is because in order to stop the effect described above you have to have a very large current going down the divider chain compared to the current you want to take out. This is in the order of 10 to 100 times. So for example suppose your 3v3 load takes 50mA you would have to have at least 10 * 50 = 0.5A going down your resistor chain. This means the resistors have to be very low and you generate a lot of heat and use a lot of power.

So while a potential divider can work for signals it should never be used for anything drawing power. It is an idea that should be dismissed out of hand.

Not at all - I've used potential dividers to power low voltage sensors but admittedly...

All I can say is... you have been extremely lucky.

There is a good reason that there is a real 5V regulator on the Arduino board and not just a couple of resistors.

I only ever consider using a voltage divider in a voltage reference circuit, for an op amp or multi-meter measurement or as an input to an Analog to Digital Converter . If used with an A/D circuit, I still insert a Zener as a safeguard.

The 5v source I've used has always been a regulated supply, whether Arduino or other MCU's - I have used an MCU digital output through a potential divider to power a 3.3v laser on/off with no ill-effects for example.

Therefore it IS regulated before reaching the potential divider - I'm not so stupid that I would consider a potential divider adequate to power an MCU or a full circuit.

For more semi-permanent logic level conversion I have several of these: http://www.sparkfun.com/commerce/product_info.php?products_id=8745

Mike - many thanks for your expanded explanation, it all makes perfect sense now. I do admit that I was wrong advising Lawrah to use a potential divider in this instance.

However one thing that is confusing me slightly regarding the Wii Nunchuck.

On the Wii Nunchuck code from todbot http://todbot.com/blog/2008/02/18/wiichuck-wii-nunchuck-adapter-available/, he is powering the Nunchuck directly from what I believe is an Analog pin?

Many Thanks, Morrolan

he is powering the Nunchuck directly from what I believe is an Analog pin?

Those pins can also be driven in the digital mode. He is defining them as digital outputs and writing high to one and low to the other. If the load is below 35mA you can do this although the voltage won't be too stable and you will both source and sink that current through the chip.

Connect the Vin of the regulator to 5V, connect the Vout of the regulator has the 3V3 voltage on it. Connect the ground of the regulator to the ground of the arduino. Then you need a 0.1uF capacitor between Vin and ground and another between Vout and ground.

For more semi-permanent logic level conversion I have several of these: http://www.sparkfun.com/commerce/product_info.php?products_id=8745

Would I need one of these to do what Mike is suggesting? Also just a question, do the 3.3V regulator pins go "gnd, vout, vin" or "vin, gnd, vout"? EDIT: Nvm, should be vin, gnd, vout.

Also I was thinking, I did try plug in the nunchuck in the 5V pin on the Mega and it still disconnected itself like before, how would adding another regulator (3.3 I'll add in) give me a different outcome? I would think that the same thing would happen?...

Thanks a lot, Laura.

Would I need one of these to do what Mike is suggesting?

You only need one if the logic signals are different. If you are powering the mega off 3v3 then no.

do the 3.3V regulator pins go "gnd, vout, vin" or "vin, gnd, vout"?

Yes. They can go either way, see the data sheet for the exact one you have.

I did try plug in the nunchuck in the 5V pin on the Mega and it still disconnected itself like before

First time you mentioned this. This thread has got a bit tangled. The FTDI chip is overheating. It is because too much power is being taken from it, on the 3v3 line. Therefore supply the power from a regulator. If you power a device on 3v3 and the Mega off 5V you will need a level translator.

A nunchuck should work off 5V as far as I know. Can you be more explicit about how it disconnected itself?

Okay so plugging the 3V3 regulator into the 5V pin from the Mega as you stated before should be fine?

That was not the first time I mentioned it, I mentioned it in my second post in this thread:

...Also, just tried using the 5v to power up the Nunchuck instead of the 3V3 with everything still connected to the Mega, still disconnects itself from the computer however the FTDI chip does not get as extremely hot as before when it was connected to the 3V3. HMM.

Well it sounds as though I'd need a level translator then... I'll have to Google that. :S (Oh - same as logic level converter?)

Well it does exactly the same thing as when it was connected to the 3V3 (disconnects itself from the computer - you hear the 'beep' sound that Windows makes when disconnecting an external device) but the FTDI chip does not get as heated.

disconnects itself from the computer

Right, it sounds like you have a short circuit on the nunchuck and this is either pulling too much current through the FTDI when you connect it to 3v3 and the chip gets hot. Or when you connect it to 5V you are pulling too much power out of the USB port and it shuts down and so disconnects itself. I would say either your wiring is wrong or you have shafted your nunchuck.

Alrighty well I'll quadruple check my wiring as soon as I get home in an hour, I hope it is just a short circuit... I really do.

I really appreciate your time and responses.

Laura.