Hackaday alum [Caleb Kraft] tipped us about a nice hack he got to see at the Open Hardware Summit this year. It is a flexible haptic strip made from a LED strip.
Cheap flexible printed circuit boards aren’t easy to find, so [Jacob] basically switched all the RGB LEDs of his strip with shaftless vibratory motors. The LEDs were addressed using WS2801 LED drivers so the hack also consisted in shorting the current feedback resistors. As a result, the motors will use as much current as the driver can give and [Jacob] can individually drive each motor. Luckily for him there already was an Arduino library called fastSPI to drive the strip, so he managed to make a nice haptic device in no time. In case you were wondering, the maximum number of motors you could drive is 32.
Our own [Eric Evenchick] also saw a lot of great project demos during his time at the OHS.
[via EE Times]
This seems like it would be a really good way of easily creating one of those ‘North is this way’ belts.
Or “midi”fy playing glass water tunes.
Good grief! What a terrible article title. You could have used ‘swapping’ instead of ‘switching’, which wouldn’t bring up the idea of, you know, a switch, controlled by a motor. And it’s not even about swapping out the LEDs. It’s a haptic interface.
Perhaps something like “Haptic feedback with an LED strip”. “Haptic feedback” tells us what it is, “with an LED strip” piques our interest to find out how it’s done. Yes, some people will get annoyed when it turns out that the LED strip was heavily modified.
For goodness sake HaD, hire some decent editors!
Hey Andrew,
I’m very sorry to disappoint you, I’m not a native english speaker as you may be.
Maybe you could contribute to HaD by applying to the last offer we published?
It has nothing to do with being a native speaker. The text in the body of the blog post was completely fine. However, the title of the post is hugely important and you chose to talk about the act of changing the LEDs to motors which is irrelevant. In my opinion you should talk about what the project does (it’s a haptic interface).
Thank you for changing “switching” to “swapping”.
You sir, are a douche.
Thank you for your constructive and helpful comment. If you were an engineer you would realise the importance of carefully choosing words and using them correctly to make your ideas clear to another person. Much like you would carefully choose components if you expected your circuit to work.
Good day to you, sir.
Doesn’t have anything to do with an engineer degree.
You honestly don’t speak a second language. and i agree with Not_A_Grammer_NAZI. If you were a real engineer you would focuses less on how something is worded and more on what it does and don’t tell me about choice of words as i find the choice of your words lacking
Seriously? Why do you have to be such a pain in the ass? What on earth will changing the title change? Either way, the use of both words depends on the dialect where the speaker is from. It’s not his fault that you’re from some southern american redneck republican state in the bible belt.
And it’s “an LED strip”, not “a LED strip”.
Bah!
Simmer down now. I promise it will be OK…the mean man won’t bother you again. No more tears, only dreams now.
Our Jimmies are eternal
I’ve got jimmies of steel
That would depend on how you look at it, ‘LED’ can be seen in three ways.
LED has become such a commonly used abbreviation that it’s approaching the status of laser, which is commonly considered a word. In this case writing ‘a LED’ would be correct, however you probably would want to start using lower case letters, so it should really be ‘a led’.
You could also see LED for what abbreviates, light emitting diode, in this case ‘a led’ would also be correct.
If you however see LED as the letters L.E.D. it would be correct to write ‘an LED’.
No. ‘a LED’ is incorrect, for the reasons I give below. ‘a light-emitting diode’ is correct, because we are using the full words. ‘a led’ is also incorrect, since nobody actually does that, and if they did, they’d be wrong.
So if people did it it would be ok, except it wouldn’t because they would be wrong? Catch 22?
Why ‘a’ is not right? http://www.englishclub.com/pronunciation/a-an.htm
Good question. Thanks for the link. If you look at the last example “FBI agent” you can see why ‘a’ is not used. It’s because of the sound of the letters (or words). “FBI” is pronounced “eff-bee-eye”, so the initial sound is a vowel, so we use ‘an’: “an FBI agent”.
Similarly, “LED” is pronounced “ell-ee-dee”, again with a vowel sound at the beginning, so we use ‘an’: “an LED”.
Some people say “LED” should be pronounced “led” (as in past tense of ‘to lead’) but it’s an initialism, and most people say “ell-ee-dee”.
http://en.wiktionary.org/wiki/LED
Yes! Just like an SCUBA diver!
No. SCUBA is an acronym. LED is an initialism. We pronounce “SCUBA” as a word. We pronounce “LED” with the sounds of the individual letters.
I see you are using irony to make a point. Well done.
they can help:
http://www.aspergers.com/
Well technically,
If I remember my grade school grammar correctly;
an is used prior to vowels and a is used prior to consonants.
So a LCD would be correct even though the letter “L: is
pronounced with an e at the beginning such as “EL”.
So while An L sounds and feels much better, A L
is the technically correct way to write it.
Or did they change that over the past 40 years or so?
You remember perfectly, but you didn’t understand. “An” is used in front of vowels, but only vowel sounds. “LCD” is pronounced “ell-see-dee”, so we must use “an”: “an LCD”. “a LCD” is incorrect. This is explained in japala’s link above.
Here’s another example. ‘u’ is a vowel, so we say “an umbrella” but we also say “a unicorn”, because the initial sound of each word is different, even though it’s a vowel.
I don’t mind people bristling at these messages. Some people will learn from them, others will remain ignorant. It’s your choice. Educate yourselves.
“You remember perfectly, but you didn’t understand”
No actually, I don’t recall hearing or reading that.
Maybe that was taught during the several weeks
I was in the hospital being treated for one flu after
the other.
I also had a difficult time with Multiplication.
Then a couple years ago I was laying in bed
and realized that if the teachers had taught
us how to read the equation rather than
memorizing tables, I would have known that
2 X 3 meant to add 2, 3 times or add 3 2 times.
Memorization has it’s merits, but explaining things
so that kids can understand it gets you much further
ahead when teaching them. Then they don’t need to
have answer to 12 X 12 memorized.
@Andrew LOL … Cool hack though.
What about back EMF?
Adding a protection diode may be a good idea.
I agree with you, some protection is in order if he wants this to last. I wonder if there has been any tests conducted on how well one of these LED driver ICs can control a motor.
The article does mention a Schottky diode as a flyback across the motor terminals, so I’d assume they are there, even though I don’t see them in the pictures.
I’d be more worried about the motor current; especially the peak current when the motor is turned on but not yet moving, as this current would only be limited by the resistance of the motor windings, and the internal resistance or voltage drop in the LED drivers.
*an EMF ;)
Great… Now all someone has to do is weave some of these into an undergarment.
already done! http://www.youtube.com/watch?feature=player_embedded&v=T6vul95hwOY
You must be naive good sir. There are far more insidious products on the market already (for women).
Does SPI not imply that there has to be a connection from each motor/slave to the µC?
Wouldn’t that mean that there has to be 32 connections going to the µC?
I’m a noob so please correct me if I’m wrong.
FastSPI is a library to drive a bunch of different types of LED strips which use chainable controllers which may or may not be using an SPI-like protocol, which is where your confusion stems from.
thx
not if each ic has its own address.
So the ICs would forward (like port forwarding) the instructions to the IC with the correct address?
Another question. In the picture there are only 4 connections to the µC Board.
So there is only MOSI and SCK (MISO is not needed cause the ICs don’t send things to the µC) right?
Each LED has a 24-bit shift register (8 bits each for RGB). The bits are clocked in at 100 nSec. Now, there are two basic ways you can make a shift register work: clock the bits in, then latch them to the outputs, or enable the latch so as the bits shift in, they “ripple” down the outputs.
From the behavior of the LED strips I have been using, I suspect there is a counter for the number of bits shifted in, and when it hits 24, the bits are latched to the output.
There also *appears* to be a timeout on the bits changing, so when you have two or more LEDS in a row, so you are shifting sets of 24-bit words, the logic recognizes the “end” of the bit shift because nothing was shifted for a period of time, which is ANDed wit the counter to latch the outputs.
I have not looked at the data lines vs latch with a scope. I am just looking at gross behavior. But this is how each LED is “addressed” – the first 24-bits clocked in end up at the Nth LED when N sets of bits are shifted in.
Cool hack with the motors…
Looks like with a few Ultrasonic sensors and this belt you could build a haptic 360° “vision” for blind people
Noe this IS a hack, and if you are in need of something like this a very handy one.
Thats pretty clever, and not something I’d have thought of.
I have a few of these pancake motors floating around here taken from old phones, also these are sometimes used in PDAs etc
Yeah, haptic vision would be feasible as would a gaming force feedback “helmet”.
I have one a LED strip with WS2801 ICs myself and figured out from the datasheet and reverse engineering that the WS2801 works in constant voltage mode and that a series resistors limits the current through the LED to 10mA. The WS2801 can supply in constant voltage mode to 100mA max. When I connected a motor it runs weakly. When I shunted the series resistor the motor run powerful for a split second and then turned off. May it be that the driver is overloaded?
I figured out that when I supply the led strip with 4V the motors run more happy than from 5V