0

I am trying to repurpose a replica GPO 746 using a Raspberry Pi, however the rotary mechanism is rather different to that in an original 746. The circuit has 12 LDRs arranged in an array (3 columns, 4 rows), a shutter attached to the dial, and an LED above the shutter. The shutter is arranged so that when the dial is rotated, a single LDR is illuminated by the LED. The finger stop incorporates a microswitch which powers the LED. I've included a rough circuit diagram below.

The row and column wires were previously connected to an HM9102D, however I cannot reliably replicate the setup using a Raspberry Pi. Reading the GPIO pins via multiplexing seems to have inputs triggered at random, even when the LED is off.

How can I reliably read the inputs from the dial?

schematic

simulate this circuit – Schematic created using CircuitLab

Foxocube
  • 101
  • 2
  • Can you define Req of LDR with Light ON/OFF? and thus make window comparators reliable for logic level detection? Can you ensure wiring does not pickup noise using twisted pair AWG 30 Magnet wire? Can you detect source/destination of noise current loop? Do you have noise filters for FL light flicker or DC ripple?? – Tony Stewart EE75 Apr 16 '17 at 10:46
  • The Charlieplexing style requires unidirectional elements of the matrix. LDR is bidirectional. – AltAir Apr 16 '17 at 11:28
  • @TonyStewart.EEsince'75 Pretty much no to all of that... – Foxocube Apr 16 '17 at 11:35
  • Then you know what you need to learn Generally Vishay/Sharp/Infineon PD's are less sensitive (hi Z) than LDRs, but far more accurate. So start with IR or Red LED and black or clear PD (respectively) and measure I=V/R with light on /off with close distance using narrowest angle LED. – Tony Stewart EE75 Apr 16 '17 at 11:37
  • @utu2012 Sorry, got charlieplexing and multiplexing mixed up. I've corrected the post now. – Foxocube Apr 16 '17 at 11:39
  • Learn optical design , then use of PD's and LED's with gap Loss (inverse squared) then mW(optical)/mA(electrical) . PD is ~0.4 A/W – Tony Stewart EE75 Apr 16 '17 at 11:44
  • @CyberJacob, the multiplexing requires unidirectional elements of the matrix too. – AltAir Apr 16 '17 at 11:59
  • It's not a restriction, it could be a diode switch, PD, LDR, Cap, R anything in a matrix Row/column scanner as long as V and SNR is in range. – Tony Stewart EE75 Apr 16 '17 at 12:03
  • @CyberJacob, One solution is to put a Schottky diode in series with every photoresistor [link](https://electronics.stackexchange.com/questions/29076/unstable-readings-in-photoresistor-matrix-with-arduino-ok-with-ohmmeter-what) – AltAir Apr 16 '17 at 12:05
  • @utu2012 How was the HM9102D reading it? – Foxocube Apr 16 '17 at 14:01
  • @CyberJacob, Something like this [page 2-34,Рin description tab,Description](http://pdf1.alldatasheet.com/datasheet-pdf/view/103711/UMC/UM91260.html) – AltAir Apr 16 '17 at 14:27
  • @utu2012 not sure if that link is going to the right place? – Foxocube Apr 16 '17 at 15:34
  • @CyberJacob, [direct link to pdf](http://pdf.dzsc.com/UM9/UM91260.pdf) – AltAir Apr 16 '17 at 15:40
  • @utu2012 sorry, I'm not seeing how that would help. Could you explain in a little more detail? Thanks – Foxocube Apr 16 '17 at 18:42
  • Let us [continue this discussion in chat](http://chat.stackexchange.com/rooms/57206/discussion-between-utu2012-and-cyberjacob). – AltAir Apr 16 '17 at 18:44

0 Answers0