Raspberry Pi - How to control motors...

[Duane Degn]

Ah I'm sorry, I hadn't realised you mailed him first. I should read more carefully

Well I assume he's getting advertising revenue for the video, and changing it might break existing links...

Not a problem. It gave me an excuse to rant some more.

I just don't understand why this hasn't been fixed. Do YouTube videos really make enough money to justify destroying Raspberry Pi boards?

[Duane Degn]

UGH!

Apparently this isn't the only thread where this problem has been brought up.
Augustus fried his Pi used to control a RC car because he followed the directions in the video.

hades1989 fried his Pi after following the video to connect the L298N to his obstacle avoiding car.

How many more are there?

Please, please, please, stop the carnage and fix the video!

[Duane Degn]

I'm starting to feel sick.

Mach_5 wonders why 5V is coming out of one of the enable pins (which of course isn't an enable pin). Too bad he didn't measure it before killing his Raspberry Pi.

hpack is scared too use his L298N h-bridge after it burned out the SoC.

Please Matthew, don't make me continue. This is just too sad.

It will only take a few seconds to make the video private. Do it now. Please.

[joan]

I suggest you complain to the Raspberry Pi Foundation. They link to the video. I do not understand why the Foundation has not acted.

[Duane Degn]

I suggest you complain to the Raspberry Pi Foundation. They link to the video. I do not understand why the Foundation has not acted.

I was was trying to figure out which email and how to word the email but fortunately the video has been made private.

I'll be able to sleep much better tonight knowing more boards are not being burnt out.

I really don't like being so negative.

BTW, I posted a "How To Use a L298N" video myself. I was trying to help someone with their L298N board and they told me about this (TRPG) video. After seeing the error, I decided to make my own video.

I used a QuickStart board with a Parallax Propeller chip (also 3.3V) but the principles are the same.

[raspberrypiguy1]

Hello gentlemen,

I have now made the offending video private after looking at various circuit diagrams and talking to the chinese manufacturers behind the L298N...

It would appear that my board is different to some of the others with the way the EN pins have been set up and this has been a most perplexing problem for me to deal with... I thank Jim Darby for his help with the diagnosis.

As to why I didn't pull it down straight away? I've been in a tent for the past week doing my Duke of Edinburgh award and whilst I could reply to email I was not able to access YouTube...

I hope that you'll be pleased to know that another video on how to control motors is upcoming - I'll be using the Ryanteck Motor Controller Board as this is a far easier option.

Thank you for highlighting your concerns but please do not hurl abuse and try to make trouble here and there - I have dealt with the issue and please remember that I am an unpaid volunteer of the community who makes tutorials for the benefits of others - in no way was I trying to be malicious and destroy Pis. As I previously mentioned my L298N is different to some of the other models and so in my own home testing (of which there was plenty) I did not come across any problems. I had always thought that those wires were unnecessary but not lethal.

As The Raspberry Pi Guy I would like to apologise for any stress that has been caused however I think that at 13 years old at the time of creation I did not know any better - I was simply showing the world what worked for me.

Yours in trust,

Matt
The Raspberry Pi Guy

[joan]

If you had heeded the warnings given on day 1 you wouldn't have had any abuse.

[raspberrypiguy1]

Hello Joan,

I was not aware of the gravity of the issue on day 1 and the reason why I left it up there when people were complaining that it had fried Pis was because of my own home testing and discussions with my chinese manufacturers who insisted that my video was OK! They had made apparent changes to the Enable pins and told me that I had nothing to worry about in my video.

Again I apologise for aggravation I have caused,

Matt

[Duane Degn]

Back in May Joan suggested you measure the voltages on the pins.

You want us to believe your L298N board which looks just the same as the ones with 5V next to the enable pins has different traces than the rest of ours?

If you connect 5V to the 5V screw terminal and measure the voltage on the pins behind the enable pins you don't read 5V on your volt meter?

Also in May, BSM Doug told you exactly what the problem with your video was and finished his post with :

Please, Please, remove the current video and replace it with a better one

Oh the abuse you should get which you haven't for negligently destroying so many boards.

I am convinced you've done much more harm than good with your "unpaid volunteer" videos which have ads and are also sponsored.

[raspberrypiguy1]

Back in May Joan suggested you measure the voltages on the pins.

You want us to believe your L298N board which looks just the same as the ones with 5V next to the enable pins has different traces than the rest of ours?

If you connect 5V to the 5V screw terminal and measure the voltage on the pins behind the enable pins you don't read 5V on your volt meter?

Also in May, BSM Doug told you exactly what the problem with your video was and finished his post with :

Please, Please, remove the current video and replace it with a better one

Oh the abuse you should get which you haven't for negligently destroying so many boards.

I am convinced you've done much more harm than good with your "unpaid volunteer" videos which have ads and are also sponsored.

I have removed the offending video. I have apologised. I have spent a great deal of time researching and investigating why this has happened. I have contacted lots of the people involved and rectified their problems. I hope that we can please draw a line under this.

On the other hand I have had 1000s of emails from people thanking me for my original motors video and professing how it has helped with their projects... So many have been empowered by my work. It is not as if the video was a deliberate assault on people's Pis! I am deeply sorry for the trouble it may have caused but please just listen to some of these extracts... Each line is a different quote.

This is the most informative and pleasant, tutorial I've seen in a while. Thank you for being awesome.

Thanks!

Outstanding tutorial. Well documented, well spoken, and professionally presented. Worked perfectly

Very good job mate. God bless u Pradeep Live u (pvt) ltd. Sri Lanka , uk (Bromley )

Excellent tutorial. Thank you.

I have done as much as I can to fix my mistake and have apologised profusely... This issue will not happen in my upcoming video.

I'm going to go and get on with the mountain of school coursework that I have currently,

Matt

[uberscubajim]

I'll post this in the other threads as well...

I've taken a long look at the (now private) video and there are a couple of errors in it. The first one is where The Raspberry Pi Guy states that the maximum output of a GPIO is 50mA when in fact it's 16mA with a total of them all of 60mA (if I recall correctly).

However, the key issue is with connecting the 5V as well as the real enable to the GPIO. As has been pointed out this is a mistake and a fairly serious one.

However, if you have a board with no jumpers on it then you could well assume that you need to to set both the pins labelled enable high to enable the output. After all: two output pins per side so two enables sort of makes sense doesn't it? It's a fairly easy mistake to make.

OK, we can talk a lot about testing and getting others to check but for a 14 year old it's an understandable mistake. And TRPG did indeed test it and it all worked for him on several different set ups. It's a much more logical thought that a few people had wired it up wrong.

Yet, we've all wired Pi stuff up wrong and been amazed it all worked afterwards. I've run workshops where I've been amazed that none of the Pies were fried.

Which got me thinking. Sure, putting 5V into a GPIO is wrong, it's bad and may kill your Pi. But it often doesn't. Maybe something else was happening.

I have an idea about that. I sat down and looked at photos of the various versions of the board and then read the datasheet for the chip and a horrible idea came to mind.

The chip itself is rated to have its logic side powered at up to 7V. In fact reading the data sheet the "5V" input can be anything from 4.5V to 7V so the 5V in on the board probably ought to be marked "4.5V to 7V". Similarly, the 12V input can be anything up to 50V! Provided you don't exceed 2A. So far, so good.

However, often people might want to power their motors of 4AA batteries: exactly like in the video. This is a reasonable choice. It would appear that some of the boards have a jumper that allows you to power the chip from other power input (the nominal 12V one).

In fact, if you're powering 6V motors from a Pi this makes a lot of sense. No need to connect the 5V and the inputs just work.

BUT, and this is the nightmare, if that jumper comes preinstalled then never mind feeding 5V into a GPIO input, you're feeding 6V into the Pi's 5V line! Someone even mentioned using a 9V battery! If you put 9V onto a Pi's 5V line you are going to kill it!

This, I strongly suspect, is why a few people had such disastrous results.

Thoughts anyone?

Jim.

[joan]

My understanding is that the on-board 5V regulator will take up to 30V from the motor supply and safely convert that to 5V for the logic supply (if that jumper is fitted).

[uberscubajim]

Well spotted!

Most of the boards seem to have that reg on them, but not all. I've also spotted one with the reg but no jumpers! I've no idea how that's supposed to work, maybe it's a 5V out?

We've seen 5V in fry a GPIO input but never an entire Pi. I'm just trying to think if there's something we've missed.

Take a peek at http://www.ebay.co.uk/itm/SainSmart-L29 ... 1158904462

It clearly has a 5V reg and the enable are labelled far better. But what controls where the regulator's output goes? If anything it looks like the 5V is an output.

However, I also just found http://www.ebay.co.uk/itm/Arduino-Motor ... 1255436110 Look at the VD=VS jumped that's installed by default!

[joan]

Well spotted!
...

Actually grahamed pointed that out to me http://www.raspberrypi.org/forums/viewt ... 76#p543576

There are a lot of different models of this design of motor driver board. Some come with jumpers, some without.

[uberscubajim]

Actually grahamed pointed that out to me http://www.raspberrypi.org/forums/viewt ... 76#p543576

There are a lot of different models of this design of motor driver board. Some come with jumpers, some without.

I think this is the key issue here is that The Raspberry Pi Guy's honest mistake was horribly magnified by a tiny minority of boards out there. That's the problem here: there are a load of very similar boards of highly varying quality, configuration and default jumpering.

And don't get me started on build quality... I got a stepper motor controller that had worse soldering than that performed by a friend's seven year old daughter!

Jim.

[Duane Degn]

I think this is the key issue here is that The Raspberry Pi Guy's honest mistake was horribly magnified by a tiny minority of boards out there. That's the problem here: there are a load of very similar boards of highly varying quality, configuration and default jumpering.

There are a lot of models of L298N board but the little red ones like the one in TRPG video and mine all have the 5V next to the enable pins. This is to let the L298N boards easily be used with stepper motors.

So some boards don't have jumpers but I bet if you check the continuity between the 5V screw terminals and the two pins behind the enable pins you'll find them connected on all boards which look like these.

The boards need 5V (or close to it) to power the logic portion of the L298N chip. Many boards have onboard 5V regulators so you just need to provide Vin and not a 5V source.

If one doesn't use the onboard regulator then you need to connect an outside 5V source to either the screw terminal position or the one of the two 5V pins behind the enable pins. In all cases there will be 5V on the two pins behind the enable pins.

It's not too hard to figure out how the boards are laid out if one probes around with a multimeter.

I'm sure the initial mistake was honest though IMO, incredibly ignorant. Understanding what the jumpers on a PCB should be a prerequisite before using the board let alone pretending to instruct others on its use.

While the initial mistake was honest, the refusal to listen to the many warnings and objections brought up by others is incredibly negligent. IMO, Matt was enjoying being The Raspberry Pi Guy too much and let his hubris blind him to his short comings.

The many sad tales this video has caused can be read by searching for the video URL in this forum.

[uberscubajim]

After all this talk of keeping 5V (or, indeed, more) away from the GPIO pins I've just found out that I've been feeding 5V into the serial port on on of my Pies. Oops!

Ironically this was from an ARM based board where I've been experimenting with coding for a ARM Cortex-M3 series processor. Both run at 3V3 but I'd left the buffer chips on the M3 set to 5V output.

No magic smoke though....

[joan]

After all this talk of keeping 5V (or, indeed, more) away from the GPIO pins I've just found out that I've been feeding 5V into the serial port on on of my Pies. Oops!

Ironically this was from an ARM based board where I've been experimenting with coding for a ARM Cortex-M3 series processor. Both run at 3V3 but I'd left the buffer chips on the M3 set to 5V output.

No magic smoke though....

You can feed any voltage you want into your Pi. We don't particularly care - unless you advise others to do the same.

[uberscubajim]

I most certainly wouldn't advise others to do the same. I was just a little embarrassed that as I'm usually very careful I managed to make such a mistake. As far as I know this is the first time I've done so.

[Duane Degn]

Matt,

I'm very glad you took down the video but there is still more you could do to minimize the damage.

I really think you should make a correction video immediately. The video's title should be something like "Urgent! Important Correction to Earlier Video: Raspberry Pi - How to control motors...". I think this video should be on your title page for few weeks but left on your channel indefinitely. I think it's very important to get the message to these hundreds of thousands of viewers about the danger of connecting 5V to a GPIO.

There's also the list of emails you received about the motor control video. You could reply to each email telling them about the 5V error.

You probably also still have access to the many comments on the video. This would be another way of contacting past viewers.

Every single board connected the way shown in the video either has already failed or will likely fail in the future if the connections haven't been changed. This is true for all the little red L298N boards like the one shown in your video. It doesn't matter how the jumpers are set, either the 5V will be provided by the regulator, or the 5V will be provided from an external source.

I'm baffled how you can say the problems your video caused to the people involved have been "rectified". I seriously doubt you're in the financial position to rectify the damage this video has caused, but there are the steps listed above which could be taken to reduce the continuing damage.

[Ravenous]

I'm very glad you took down the video but there is still more you could do to minimize the damage.

Come on, I think he's done enough. Let's not waste forum space forcing people into making ever more grovelling apologies. I doubt anyone's keeping offline copies of the offending video.

[BMS Doug]

Hello gentlemen,

I have now made the offending video private after looking at various circuit diagrams and talking to the chinese manufacturers behind the L298N...

It would appear that my board is different to some of the others with the way the EN pins have been set up and this has been a most perplexing problem for me to deal with... I thank Jim Darby for his help with the diagnosis.

As to why I didn't pull it down straight away? I've been in a tent for the past week doing my Duke of Edinburgh award and whilst I could reply to email I was not able to access YouTube...

I hope that you'll be pleased to know that another video on how to control motors is upcoming - I'll be using the Ryanteck Motor Controller Board as this is a far easier option.

Thank you for highlighting your concerns but please do not hurl abuse and try to make trouble here and there - I have dealt with the issue and please remember that I am an unpaid volunteer of the community who makes tutorials for the benefits of others - in no way was I trying to be malicious and destroy Pis. As I previously mentioned my L298N is different to some of the other models and so in my own home testing (of which there was plenty) I did not come across any problems. I had always thought that those wires were unnecessary but not lethal.

As The Raspberry Pi Guy I would like to apologise for any stress that has been caused however I think that at 13 years old at the time of creation I did not know any better - I was simply showing the world what worked for me.

Yours in trust,

Matt
The Raspberry Pi Guy

Hi Matt,

Please stop accusing people of hurling abuse or making trouble, that has never been anyone's intention.

you are indeed an unpaid volunteer of the community, a trait that you share with all of the rest of us, you try to make tutorials for the benefit of others without being willing to listen to criticism of those tutorials when it is evident to others that your tutorial is flawed.

As has been mentioned you have been told many times in the past that there were issues with your video. You may recall that I sent you an e-mail detailing the errors, to which you replied that you knew more about the subject than me and were going to ignore my advice.

The Hubris involved in naming yourself an expert in the subject "Raspberry Pi Guy" can lead to beginners trusting you perhaps more than it is wise to trust a 13 year old. (as you were when you made the offending video).

My advice to you is to learn from this, check your facts and offer your ideas for peer review before general release.

you will find that this forum is full of people who will be happy to help out, review your video and offer constructive criticism to help you completely understand the subject and release a perfect video.

[Duane Degn]

I'm very glad you took down the video but there is still more you could do to minimize the damage.

Come on, I think he's done enough. Let's not waste forum space forcing people into making ever more grovelling apologies. I doubt anyone's keeping offline copies of the offending video.

Did I call for more apologies?

According to Matt there are thousands of Raspberry Pi boards already hooked up as instructed by the video.

All of those boards connected to the h-bridge as instructed in the video are in extreme danger of being burnt out (if they are still functioning).

Based on Matthew's estimates of the number of viewers who had already followed his instructions, his video could end up destroying thousands of boards.

The connection shown by the video isn't just risky, it almost certainly disastrous (to the board) if left as is.

I think a correction video to warn those who have already made this dangerous connection a good idea. Apparently you and Matt disagree. I disagree with you about such a suggestion being a waste of forum space.

[jamesh]

OK, this is going to turn in to a flame war.

I believe its all fixed now, so locking. Anyone have anything to add? Please PM me.