Forums > Technical Discussion > Luxeon Poi...

Posted: 17th Jul 2005
im sorry but i don't get it. i mean, you set the bits to 1 with the 'wait1second' thing, right?



but why does RB04- RB06 are 'turned to 1'? why not another one?



in a normal code there's something like

"MOVLW b'01000000'

MOVWF PORTB"



i guess im totally missing the point



*edit*

i would really appreciate if someone explained the code step-by-step... i think that'll take me much further...
EDITED_BY: anonomatos (1121644030)

"Don’t know how long, this one’s gonna take;
I could fail, but I’d rather be a fuckup, than a fake"

Posted: 17th Jul 2005
Just got back from Glade, and I'm about to do some work on the ASM tutorials. I probably won't be able to do much, as I also have to rebuild my prototypes before Falmouth, and transfer the whole LEDPoiwiki onto a faster server.



The 4bit.asm is just a binary counter, the original code did this to the port:



| RB7 | RB6 | RB5 | RB4 | RB3 | RB2 | RB1 | RB0 |



00000000

00000001

00000010

00000011

00000100

00000101

00000110

00000111

00001000

00001001

00001010

00001011

00001100

00001101

00001110

00001111



The changes to the LEDs are made by these lines:



movfw Count ;Move Count into W

movwf PORTB ;Write W to PORTB



The value of Count is written to the port, where it is expressed as a binary number, even though it's working in decimal. Numbers written to ports always come out as binary. The two swap lines just change which group of 4 bits the changes happen on. Without them the bits would change as above, and you wouldn't see anything happening, as you said you'd hooked up LEDs to ports 4:6. With them the first 4 and last 4 are swapped around, like from this:

00001011



to this:

10110000



so that you can see the changes.



It's a good idea to become very familiar with all the instructions, and exactly what they do to registers and bits. You'll end up using nearly all of them.





There will be 1 second period that the LEDs are all off, as it always starts from 0 (00000000). You could change it so that the Reset subroutine contains:

movlw b'00000001' ;put the value '1' in the W register

movwf Count ;copy the W register to the Count register



instead of:

clrf Count



then it will always start from 1.
EDITED_BY: polarity (1121646792)

You aren't thinking or really existing unless you're willing to risk even your own sanity in the judgment of your existence.

Green peppers, lime pickle and whole-grain mustard = best sandwich filling.

Posted: 17th Jul 2005
okay, so it count 16 times... starting with 0.
but still don't get it.
it counts 16 times, for each port 1 time, right? (i figured, 16 ports for I/O, and 2 ports for VDD and VSS will make 18 pins... like my pic has).

so it goes from 1-16, and 3 of the 16 times a Led is turned on (and sometimes two or 3 at the same time)...

still don't get it... oh well. ill wait a bit and hope your explaination helps a bit

"Don’t know how long, this one’s gonna take;
I could fail, but I’d rather be a fuckup, than a fake"

Posted: 17th Jul 2005
Thats 16 pins. Two ports that each have 8 pins, as they are 8 bit ports.

As it's counting in binary 16 possible combinations of ones and zeros can be expressed using four pins of a single port, which is why only four of the eight bits change between 1 and 0.

One LED has a value of 1, another 2, and the third one has value of 4. You add the values together to get what the Count value is.

Because you haven't got a fourth LED hooked up on RB7, you can't see it's condition. This LED would have a value of eight, so the highest you could count is 8+4+2+1, which is 15. Include 0 and that's 16 different values (computers always count 0 as one of the numbers). At the moment you aren't seeing that extra eight, so it will just count from 0 to 7 then reset. Binary counting makes a simple pattern, where the bit with value 1 cycles off once, then on once, the bit for value 2 cycles off twice, then on twice, the bit for value 4 cycles off four times then on four times, and so on for all the bits.

The LEDs don't actually count to 16, The value of Count is tested before any changes are made to the LEDs. If it's at 16, then the value is reset to 0 before the change is made to PORTB.

You aren't thinking or really existing unless you're willing to risk even your own sanity in the judgment of your existence.

Green peppers, lime pickle and whole-grain mustard = best sandwich filling.

Posted: 18th Jul 2005
so that would explain the delay indeed.
so, if i got it correctly. the fourth led (which isn't there) cycles off and on for 8 times?

if i got this all right... this code is a very shitty one to start building from? if it's got a permanent 'build-in' delay it's not very handy.

"Don’t know how long, this one’s gonna take;
I could fail, but I’d rather be a fuckup, than a fake"

Posted: 18th Jul 2005
Right on all counts.

It's just a good piece of code to test that the circuit is hooked up right.

You can always disable the delay by putting a semicolon infront of the call instruction though. Then it'll run so fast that the LEDs appear to be always on.

You aren't thinking or really existing unless you're willing to risk even your own sanity in the judgment of your existence.

Green peppers, lime pickle and whole-grain mustard = best sandwich filling.

Posted: 18th Jul 2005
I already did that with changing the 250 ms delay into a 1ms delay in the test code. You get a FREAKY strobe by doing that!

Though that kind of strobe is not what I wanna be using. So... problem is... i don't have a code to start building from. But I guess it would be the easiest to start with a strobe...
If I wanna do a strobe with 3 colors (1 at a time) each with a delay in between which is not noticably when you hold the poi still... only when spinned.

Well.. since i don't have anything to start build from, i thought this would be the best start (and it would be nice if you guys correct it and make additions):

STROBE_1 MOVLW B'00010000'
MOVWF PORTB
MOVLW B'00100000'
MOVWF PORTB
MOVLW B'01000000'
MOVWF PORTB
GOTO STROBE_1

This should be the first thing. Now I need something that turn them on and off for a split second. 1ms should do it I guess...
but how do I do that?

MOVLW B'01000000'
MOVWF PORTB
CALL TIMER

Then something like:

TIMER MOVLW D'1'
MOVWF ???
...

that's basicially where i'm stranded now :P

"Don’t know how long, this one’s gonna take;
I could fail, but I’d rather be a fuckup, than a fake"

Posted: 18th Jul 2005
You've got everything you need there really. From what I can tell, you want to make your poi strobe red, green, blue? So take the original code you had - the one that worked. Now strip out everything bar the bits that do the basic setup. Now engage your brain... you want to do a red green blue strobe yes? Start with a simple loop to do that:

StrobeLoop
MOVLW B'00010000' ; Set the red bit (pin RB4) to 1
MOVWF PORTB ; Write the value to the port

MOVLW B'00100000' ; Set the green bit (pin RB5) to 1
MOVWF PORTB ; Write the value to the port

MOVLW B'01000000' ; Set the blue bit (pin RB6) to 1
MOVWF PORTB ; Write the value to the port
GOTO StrobeLoop

This assumes that the Red LED is connected to pin RB4, green to RB5, and blue to RB6.

That will run around the loop as fast as possible. If you want to slow it down, add a delay routine in:

StrobeLoop
MOVLW B'00010000' ; Set the red bit (pin RB4) to 1
MOVWF PORTB ; Write the value to the port

Call DelayRoutine ; Wait for a bit

MOVLW B'00100000' ; Set the green bit (pin RB5) to 1
MOVWF PORTB ; Write the value to the port

Call DelayRoutine ; Wait for a bit

MOVLW B'01000000' ; Set the blue bit (pin RB6) to 1
MOVWF PORTB ; Write the value to the port

Call DelayRoutine ; Wait for a bit

GOTO StrobeLoop

That will give you a delay between changing colours. The longer the delay, the slower the colours will change.

Cake or Death?

Posted: 18th Jul 2005
You can also make the delay a lot simpler if you don't need it to be accurate to a certain number of milliseconds.

Code:

---

DelayRoutine
movlw d'255'
movwf Count ;give Count the value 255
DelayLoop
decfsz Count ;decrease Count , if it's 0 skip the next line
goto DelayLoop
return

---

You can always put one loop inside another to make the delay much bigger.

Code:

---

DelayRoutine
movlw d'255'
movwf Count1 ;give Count1 the value 255
DelayLoop1
movlw d'255'
movwf Count2 ;give Count2 the value 255
DelayLoop2
decfsz Count2 ;decrease Count2, if it's 0 skip the next line
goto DelayLoop2
decfsz Count1 ;decrease Count1, if it's 0 skip the next line
goto DelayLoop1
return

---

You aren't thinking or really existing unless you're willing to risk even your own sanity in the judgment of your existence.

Green peppers, lime pickle and whole-grain mustard = best sandwich filling.

Posted: 18th Jul 2005
If I would be able to vote for the most helpful HoP-members you two will tie for the first place

im gonna try some of that code tonight

"Don’t know how long, this one’s gonna take;
I could fail, but I’d rather be a fuckup, than a fake"

Posted: 18th Jul 2005
Just a quick note on protecting your electronics

I made test pieces for electronic led staff/poi/light bars a couple years ago now and covered them in hot melt glue from hot melt glue gun(!) this stuff goes on easy and can be smoothed down VERY CAREFULLY WHEN NOT AT FULL TEMP (wet your finger/palet knife first too) and remains rubbery and allows the light out. The texture of it means it acts very well as an impact absorber without losing much of the light output.
this could save having to replace them when they collide or get knocked.

I'd rather go Poi with myself!

Posted: 18th Jul 2005
good point casting it in resin is a bit too ambitious... glue gun might work

"Don’t know how long, this one’s gonna take;
I could fail, but I’d rather be a fuckup, than a fake"

Posted: 18th Jul 2005
This should be about the code... but it sticks with the first LED... not going to the second or third one... what did I do wrong?


LIST P=16F627A, R=DEC

__CONFIG _WDT_OFF & _PWRTE_OFF & _BODEN_OFF & _MCLRE_OFF & _CP_OFF & _INTOSC_OSC_NOCLKOUT & _LVP_OFF

include "P16F627A.inc"

;--------------------------------------------------------------------------
; Subroutines
;--------------------------------------------------------------------------

Ram EQU h'20'
Count1 EQU Ram+0
Count2 EQU Ram+1

Init clrf PORTA ;all of porta low
clrf PORTB ;all of portb low

bsf STATUS,RP0 ;change to bank1

bsf PCON, 3 ;set INTOSC to 4Mhz

clrf TRISA ;all of porta outputs

clrf TRISB ;all of portb outputs

bcf STATUS,RP0 ;back to bank0

DelayRoutine movlw d'255'
movwf Count1 ;give Count1 the value 255

DelayLoop1 movlw d'255'
movwf Count2 ;give Count2 the value 255

DelayLoop2 decfsz Count2 ;decrease Count2, if it's 0 skip the next line
goto DelayLoop2
decfsz Count1 ;decrease Count1, if it's 0 skip the next line
goto DelayLoop1

StrobeLoop MOVLW B'00010000' ; Set the red bit (pin RB4) to 1
MOVWF PORTB ; Write the value to the port
Call DelayRoutine ; Wait for a bit
MOVLW B'00100000' ; Set the green bit (pin RB5) to 1
MOVWF PORTB ; Write the value to the port
Call DelayRoutine ; Wait for a bit
MOVLW B'01000000' ; Set the blue bit (pin RB6) to 1
MOVWF PORTB ; Write the value to the port
Call DelayRoutine ; Wait for a bit
GOTO StrobeLoop

END

"Don’t know how long, this one’s gonna take;
I could fail, but I’d rather be a fuckup, than a fake"

Posted: 18th Jul 2005
You need a return instruction at the end of the delay routine. At the moment it finishes the delay and then goes on to the line starting 'StrobeLoop MOVLW', so it's just looping back to the first LED. The return makes it go to the line after the call instruction, so it moves on to the next LED.

You aren't thinking or really existing unless you're willing to risk even your own sanity in the judgment of your existence.

Green peppers, lime pickle and whole-grain mustard = best sandwich filling.

Posted: 18th Jul 2005
assembler is not complaining and compiles it... but it's not working with 'RETURN' after 'goto delayloop1'

though it gives 24 warning messages (though no errors)... maybe this'll could help you guys help me?



strobe1.asm



Warning[216] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 1 : Radix superceded by command line.

Warning[205] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 3 : Found directive in column 1. (__CONFIG)

Warning[205] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 5 : Found directive in column 1. (include)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 16 : Found opcode in column 1. (clrf)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 18 : Found opcode in column 1. (bsf)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 20 : Found opcode in column 1. (bsf)

Message[302] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 20 : Register in operand not in bank 0. Ensure that bank bits are correct.

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 22 : Found opcode in column 1. (clrf)

Message[302] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 22 : Register in operand not in bank 0. Ensure that bank bits are correct.

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 24 : Found opcode in column 1. (clrf)

Message[302] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 24 : Register in operand not in bank 0. Ensure that bank bits are correct.

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 26 : Found opcode in column 1. (bcf)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 29 : Found opcode in column 1. (movwf)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 31 : Found opcode in column 1. (movwf)

Message[305] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 32 : Using default destination of 1 (file).

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 33 : Found opcode in column 1. (goto)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 34 : Found opcode in column 1. (decfsz)

Message[305] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 34 : Using default destination of 1 (file).

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 35 : Found opcode in column 1. (goto)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 36 : Found opcode in column 1. (RETURN)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 39 : Found opcode in column 1. (MOVWF)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 40 : Found opcode in column 1. (Call)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 41 : Found opcode in column 1. (MOVLW)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 42 : Found opcode in column 1. (MOVWF)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 43 : Found opcode in column 1. (Call)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 44 : Found opcode in column 1. (MOVLW)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 45 : Found opcode in column 1. (MOVWF)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 46 : Found opcode in column 1. (Call)

Warning[203] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 47 : Found opcode in column 1. (GOTO)

Warning[205] C:\PROGRAM FILES\VELLEMAN\EXAMPLES\NIEUWE MAP\STROBE1.ASM 49 : Found directive in column 1. (END)

"Don’t know how long, this one’s gonna take;
I could fail, but I’d rather be a fuckup, than a fake"

Posted: 18th Jul 2005
You need to make sure instructions are indented (not in column 1), by putting spaces in front of them (you can use tabs, but it gets messy when you cut and paste).

Labels are OK to be in the first column though.

Check here for an example.

You aren't thinking or really existing unless you're willing to risk even your own sanity in the judgment of your existence.

Green peppers, lime pickle and whole-grain mustard = best sandwich filling.

Posted: 18th Jul 2005
the code seems to be OK now... on my test-board/pic-programmer the code is working.
but in the circuit of my poi not. and the strangest thing is... a few times it did work. I don't get it. When I write some other code on it, everything is fine. but this one doesn't work all the time.

ARGH!

"Don’t know how long, this one’s gonna take;
I could fail, but I’d rather be a fuckup, than a fake"

Posted: 20th Jul 2005
I've had some SERIOUS headbreaking over my code. Luckily someone told me I forgot the " ORG 0 ;reset vector "

So... now it's working. Just a little bit more frustration and I would have given up...



luckily, the slow-strobe thing is working now

"Don’t know how long, this one’s gonna take;
I could fail, but I’d rather be a fuckup, than a fake"

Posted: 20th Jul 2005
I've reached that point soooo many times whilst building my hyperlights!

Cake or Death?

Posted: 21st Jul 2005
Now I'm going to try some PWM'ing, but not without your help!

I ripped the 'Simple PWM code' from Polarity's site... and im trying to figure it out.



rgbfade4.asm



as far as i can tell it's a counter that count to 16. each step it takes it increases the time of '0' between '1' and another '1'.. When it reaches 16 it resets.

but as far as i can make out of the text on the site of polarity there is no timing in the code? but it just increases the counter, he doesn't need timing for that? :S and i can't find the place where it writes the result to a certain port (like it should be RB04-06 in my case)



help is MUCH appreciated



*edit*

in a day im on vacation for a week. guess you guys are at falmouth at the moment (so not able to reply )... so ill be back 30th of july... hope to see some feedback by then!!


EDITED_BY: anonomatos (1121986358)

"Don’t know how long, this one’s gonna take;
I could fail, but I’d rather be a fuckup, than a fake"

Posted: 1st Aug 2005
I'm back again from vacation.

At the moment i'm starting to get aware of the PWM possibilities. I've seen several examples, but they are all very complex.
How about this one:

PWMloop Count = 128
X above Count?
No, turn light on
Yes, turn light off
X= X+1
Goto loop
(this should dim the light constantly to 50% brightness)

And if I want the light to dim and come up again i should be adding a DECFSZ or something similar that counts every time 256 cycles are passed. If a 'round' of cycles is passed it should increase or decrease 'Count'.

Any tips? I'm trying to figure out the appropriate codes for the instructions, though there are quite hard to find.

"Don’t know how long, this one’s gonna take;
I could fail, but I’d rather be a fuckup, than a fake"

Posted: 26th Feb 2006
Hello... ran across this whilst googling for something else, and I thought I might as well link to my luxeon poi... Having spent however many hours banging my head on circuits & software to get mine going it seems fair to pass on whatever I've figured out. There's some code at the bottom for PWMing and so forth.

I have an upgrade in the works which lets me program patterns into the PIC via the serial port from a little desktop application which you draw them in, but it's too flakey for general release at the moment.

Anyway back to lurking.

Posted: 26th Feb 2006
Nice, another one 'going Luxeon'.

If you have any questions just PM me or one of the other techies (actually, they know far more than me, but still )



I've gotten MUCH further since my last post!



https://projectarclight.blogspot.com
for my thingies (havent been updated for ages properly btw)

"Don’t know how long, this one’s gonna take;
I could fail, but I’d rather be a fuckup, than a fake"

Posted: 26th Feb 2006
That polycarb tube link on your site is really helpful, I've been looking for a supplier who don't need me to buy 100 metre segments or whatever.

If your design has a bar with several LEDs along (as in the pics) it have you considered programming it to display text? I built a project like that once to put text on my bike wheels, although that had about 20 LEDs in a line to get a decent resolution.

Posted: 26th Feb 2006
Yes, I did some text experiments. But to be honoust, that project failed. It was far beyond my league and got messed up with the design. I'm now working on 2 staffs: red/yellow and green/blue with each 32 leds... update coming soon... this time it WILL work

"Don’t know how long, this one’s gonna take;
I could fail, but I’d rather be a fuckup, than a fake"

Posted: 28th Feb 2006
Blimey, not seen this thread for a while!! I'm looking forward to seeing those staffs of yours in action anonononomatos

Cake or Death?

Posted: 28th Feb 2006
hyperlight, you start sellin your poi yet? how far away are we?

Posted: 1st Mar 2006
Not yet... I'm nearing the end of the design phase for the final version of the electronics though (I hope). It's been slow going because I have a 9-5 job on top of trying to develop these toys. They're definitely progressing though, and if the first boards don't throw up any serious issues, we might be seeing full production HyperLights in the not-too-distant future!

Cake or Death?

Posted: 20th Jun 2006
They really do look fantastic, I hope to see them at Play.

LED's they get hot enough to need a heatsink though? yikes, whats the battery life expected to be?

Blinded by Hyperlights, please donate generously

Posted: 20th Jun 2006
The photo's and videos really don't do them justice - just you wait and see



They get hot because the LEDs in them consume about 3W of power! As for battery life, have a look in [Old link]:

 Written by: *HyperLight

They run off 2xAAA's... the time they run for depends on which patterns you're running. I've not tested all that thoroughly but I would guess upwards of an hour from 900mAH batteries.

Cake or Death?