Best way to implement an array of 64+ IR detectors?

outofstep · 2009-03-02 20:40

I've got·a pro dev board and am using the 40 pin version of the stamp.

If it was a simple push button type input matrix, I could simply use the voltage divider setup of the POT command.

By default the IR detectors (link)·Vout +5 when not detecting, and 0 when they do detect; I am unsure of how to set up the circuit. Also, because the IR detectors are diodes, the POT setup shouldnt work (if running the IR detectors in parallel,·to make up for the constant on state)·because the voltage read stage should always read back 0.

The only way I've come up with is to use 8 8x3 encoders. However that will eat up 24 I/O lines. Is there a more elegant/efficient way to set this up?

I'm not sure if I explained everything correctly or not, just let me know if you need me to clarify something.

Post Edited (outofstep) : 3/2/2009 8:51:48 PM GMT

Mike Green · 2009-03-02 21:09

The 74HC165 is a parallel input, serial output shift register that can be easily used with the Stamp's SHIFTIN statement and they can be cascaded. There are articles in the series of Nuts and Volts Columns downloadable from Parallax that discuss I/O expansion. They're also described (with sample code) in the StampWorks manual (Experiments #23 and #24), also downloadable from Parallax (www.parallax.com/Store/Microcontrollers/BASICStampModules/tabid/134/txtSearch/stampworks/List/1/ProductID/144/Default.aspx?SortField=ProductName%2cProductName).

You can cascade 8 x 74HC165 to provide 64 inputs using only 3 I/O pins (clock, data, parallel load). The 74HC595 provides similar output expansion.

Post Edited (Mike Green) : 3/2/2009 9:14:58 PM GMT

outofstep · 2009-03-06 16:52

Thank you very much for your reply and the included information. Very helpful.

outofstep · 2009-03-06 17:38

Okay, reading up on the SHIFTIN command it looks like I'll have to break it up into 4 parts. With this code snippet, can you tell me if I'm on the right track for shifting in 64 inputs?

code said...

tophi VAR Word
toplo VAR Word
bottomhi VAR Word
bottomlo Var Word

SHIFTIN 0, 1, MSBPRE, [noparse][[/noparse]tophi\16, toplo\16, bottomhi\16, bottomlo\16]

-edit-

It looks like I can break up the data coming in however I want. For an 8x8 matrix of sensors, I could assign an entire collumn or row a variable of its own.

SHIFTIN 0,1, mode, [noparse][[/noparse]row1\8, row2\8, row3\8, row4\8, row5\8, row6\8, row7\8, row8\8]

Post Edited (outofstep) : 3/6/2009 6:15:08 PM GMT

kleekru · 2010-08-30 00:04

OK, I'm cascading 74hc165's... the first IC is working fine... and I've used the schematic and code listed in the StampWorks manual to wire and code the second 74hc165... I've even used the suggestion above to create different variables for each chip and shift them in 8 at a time... or at least I think that is what I'm doing. Unfortunately, while all 8 bits are working fine for the first chip, when I try a switch on the 9th bit, I get nothing.

Mike, you've been very helpful in some of the other posts. How do I change my code here to reference a switch hanging off of the 2nd (and eventually 3rd, 4th, 5th...) chip?

Here is what I have:
'

[ I/O Definitions ]

Clock PIN 0 ' shift clock (74HC165.2)
SerData PIN 1 ' serial data (74HC165.7)
Load PIN 2 ' output latch (74HC165.1)

'

[ Constants ]

DelayTime CON 100

'

[ Variables ]

pedals VAR Word
xInputs VAR Word ' switch data

'

[ Initialization ]

Reset:
LOW Clock
HIGH Load ' make output and high
DEBUG CLS,
"Switches FEDCBA9876543210", CR,
"

", CR,
"Status"

'

[ Program Code ]

Main:
DO
GOSUB Get_165X2 ' get switch inputs
DEBUG CRSRXY, 10, 2, BIN8 pedals, BIN8 xInputs ' display current status
PAUSE delaytime
IF pedals.BIT0 = 1 THEN
HIGH 13
LOW 14
ELSEIF pedals.BIT1 = 1 THEN
HIGH 14
LOW 13
ELSEIF pedals.BIT8 = 1 THEN 'THIS PART IS NOT WORKING for BIT8-15!!!
HIGH 14
HIGH 13
ELSE
LOW 14
LOW 13
ENDIF
LOOP

'

[ Subroutines ]

Get_165X2:
PULSOUT Load, 5 ' load switch inputs
SHIFTIN SerData, Clock, MSBPRE, [pedals\8, xInputs\8] ' shift them in (IS THIS RIGHT?)
RETURN

Mike Green · 2010-08-30 06:09

You're shifting 8 bits into each of 2 variables ("pedals\8","xInputs\8"). That's .BIT0 through .BIT7 of each. If you want to shift the whole 16 bits into one variable, leave off the "\8" and PBasic will assume you mean "\16" because the variables are WORDs. If you use "pedals,xInputs", that's 32 bits total and would assume you have 4 74HC165s.

kleekru · 2010-08-30 07:46

Thanks Mike. I also tried using a single variable (pedals\16) but then they dont seem to respond when I have a switch against one of the pins (I'm using pedals.BIT8 for example). I'm expecting this to be something simple but I've triple checked and completely rewired and I get the same lack of response. I've tried another chip, same thing. I'm using the same wiring and code as in the Stampworks manual but can't reference the 2nd IC. Grrr.

kleekru · 2010-08-30 10:09

Here's the code with the updates (still not working)

'

[ Variables ]

pedals VAR Word
xInputs VAR Word ' switch data

'

[ Initialization ]

Reset:
LOW Clock
HIGH Load ' make output and high
DEBUG CLS,
"Switches FEDCBA9876543210", CR,
"

", CR,
"Status"

'

[ Program Code ]

Main:
DO
GOSUB Get_165X2 ' get switch inputs
DEBUG CRSRXY, 10, 2, BIN16 pedals ' display current status
PAUSE delaytime
IF pedals.BIT0 = 1 THEN
HIGH 13
LOW 14
ELSEIF pedals.BIT1 = 1 THEN
HIGH 14
LOW 13
ELSEIF pedals.BIT7 = 1 THEN
HIGH 14
HIGH 13
ELSE
LOW 14
LOW 13
ENDIF
LOOP

'LOW 12
'HIGH 14
'HIGH 13
'PAUSE 50
'LOW 14
'LOW 13
'ELSEIF switches.BIT0=1 AND switches.BIT1=1 AND switches.BIT5=1 THEN
'SEROUT 15,6,["#0 P500 #1 P1400 #2 P1600 #3 P1400",13]
'HIGH 13
'HIGH 14
'HIGH 12
'ELSE
'SEROUT 15,6,["#0 P750 #1 P1500 #2 P1500 #3 P1500",13]
'LOW 13
'LOW 14
'LOW 12
'ENDIF

'PAUSE DelayTime ' pad the loop a bit
'LOOP

'

[ Subroutines ]

Get_165X2:
PULSOUT Load, 5 ' load switch inputs
SHIFTIN SerData, Clock, MSBPRE, [pedals] ' shift them in
RETURN

Switches FEDCBA9876543210

Status 0000000001110000

The switches on the left 8 bits are not working. I have a switch on the pin that lines up with what I thought would be #8 (1st input from the second chip).

Any and all suggestions are greatly appreciated.

Mike Green · 2010-08-30 11:06

Almost all the time when something like this happens, it's due to a wiring error. Someone is referencing a diagram in a tutorial or manual and the diagram is correct, but they've copied it wrong or wired it wrong and they can't understand why it doesn't work. If you can, have someone else check your work or painstakingly go over your wiring, make a list of what's connected where, then compare that against the diagram in the text marking each wire as you match them up.

kleekru · 2010-09-21 23:36

Whew, I got it working with 2 74HC165's by completely re-wiring. Wasn't fun but it is nice to get some motivation again. On to the next step... adding another shift register. I suspect cascading a 3rd one is just like cascading the 2nd so I've wired it the same connecting Pin 9 of the 3rd to Pin 10 of the 2nd, 1 of the 3rd to 1 of the 2nd and 2 of the 3rd to 2 of the second....

However, now I'm baffled by the code. Above outofstep suggests adding separate variables for each 74HC165:
"pedals1 VAR Word
pedals2 VAR Word

SHIFTIN 0, 1, MSBPRE, [pedals1\16, pedals2\16]
"

If I use this approach, how should I write the following:
1) The code to view all 24 (or more) inputs:
DEBUG CRSRXY, 10, 2, BIN16 pedals1 ' ??how do I reference the pins in pedals 2??

2) How do I reference each of the input pins on the 3rd register... I'm assuming I can use pedals1.BIT0 - pedals1.BIT15 ...or... pedals2.BIT0 - pedals2.BIT15

Am I thinking about this correctly?

Mike Green · 2010-09-22 07:11

1) DEBUG ..., BIN16 pedals1, BIN16 pedals2

2) Correct. The Manual also discusses bit arrays. You can refer to consecutive bits using a subscript and this can span adjacent variables like "pedals1.bit0(i)".

Each WORD variable has 16 bits.

Best way to implement an array of 64+ IR detectors?

Comments