Optical Resistor Transistor Logic

[CreatorLes]

I love blinky LED's, don't you? We add them to our circuits in abundance just for the fun of seeing them light and dim according to the signals in our music circuits. Now, given that fact, wouldn't it be really cool if circuits could be made of LEDs? Well they can!

For the longest time, years and years it's been, I have wanted to make logic circuits for Lunettas and other designs using LEDS for outputs and CdS cells for inputs. This would be optical logic so it would matter how you oriented the I/O connections and watch out for light crosstalk (maybe an enhancement too). I'm already losing you so let me show you a picture worth a thousand words, attached.

If we take those inputs, hook them up to the supply voltage and replace them with CdS cells, then the incident light would turn the transistors on and off. Adding an LED in series with the output resistor should give our logic gate optical output which can be physically directed at the input of another optical logic gate.

That's the plan at least. I'm going to do a little testing now and hopefully I can get something working, then I will return with a report on the results.

Les
Les make it happen!

---

rtl-nor4t.gif
Description:	traditional RTL
Filesize:	1.79 KB
Viewed:	11464 Time(s)

---

[CreatorLes]

OK, I did a simple test for starters. I wanted to find out if the component values I chose (10k Ohm CdS cell and white LED) would light up properly in room lighting.

I found on the web that the recommended voltage drop for a white LED is 3.3V, any less than that and some of the white LEDs may not light. Well I happen to have four rechargeable NiCd batteries so I figured three or four of them connected in series would suffice as a power supply.

So I taped them up with blue painter's tape that I have from 3D printing and took one CdS cell and one LED, connected them in series at the positive terminal of the LED (CdS cells are non-polarized), and taped them to the battery ends.

Due to the awkwardness of the connection technique, I had intermittent contact but sure enough the LED lights bright and strong! Then I covered up the CdS cell and the LED glowed dimmer than with it fully lit by room lighting and also from some light from the LED itself. So yeah I know that four NiCd cells act as a good power supply for this project, plus we also know that the LED lights brightly under these conditions and can be made to dim.

That's all fine and dandy, but have I created a propagating circuit? No! For example it would be great to make a logic gate from this arrangement. What would happen if we added another CdS cell in series or in parallel with the first one?

This would create a multiple input circuit with logic-ish configuration but would it operate like a logic gate? The answer is no and I'll tell you why - it's because it attenuates with each stage in propagation. To get a strong signal we need to use a transistor.

Once we add an NPN or PNP transistor, there are a few configurations that will work as a propagating, inverting logic gate. So now I'm off to figure out how to set up such a thing. I'll be back to report soon.

Les
Les make it happen!

---

RTL CdS biasing LED - 1.jpg
Description:	Simple Test of Bias Values
Filesize:	1012.9 KB
Viewed:	489 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[CreatorLes]

I have made a functional inverter or at least what I believe to be a working inverter and I'll share a schematic with component values soon. In the mean time I would like to share the following RTL guide written by the famous Don Lancaster:

http://www.tinaja.com/ebooks/rtlcb.pdf

I have only skimmed the beginning of this doc and already I feel that I have learned a few things.

Les

[PHOBoS]

Assuming you are going to use it for musical purposes (lunettas) you might run into the problem that CdS cells have a rather
slow response time, so maybe photo transistors would be a better option. And since your goal seems to be discrete logic using
transistors, than this would also 'cut out the middleman'.

[CreatorLes]

[CreatorLes]

Ironically, I was just looking into that when you posted! I had been too hasty to check the datasheet, and now that I have I see the typical Rise time is 55 ms and the typical fall time is 20 ms. This means that the response to a square wave would be more like a ramp at 1/70ms = 13 Hz! Way too slow for music but OK for control functions.

Also as PHOBoS points out, a phototransistor is in fact a transistor so the circuit becomes even more simplified, nice!

Les

[CreatorLes]

So thanks to my stupidity and the suggestion of PHOBoS (thank you PHOBoS!), I will switch to PhotoTransistors, however I currently have none and will have to order them.

I order my parts by sending a list of links to my sister and she does procrastinate, plus shipping is a week. So probably it will be two weeks or so before I will have some PHOBoStransistors in hand. See what I did there? lol

OK, Les out for a while

Les

[blue hell]

not any olde metal-can trannies laying around? you can file of the top off of those to make them light sensitive.

[CreatorLes]

I do have a Joe Knows transistor kit, it probably has some metal can transistors. I do have some 3D printing work queued up for now though, I'll order about 20 of the real deal and go with that.

Thanks for the idea Jan!

Les

[CreatorLes]

Ah, you know, I need to rethink this. I have white LEDs and CdS cells. I do not have phototransistors for a while. So why don't I think creatively about the CdS cell alternative?

For example, the rise + fall time of 75ms is obviously for a full transition from low to high of an unspecified resistance change. Then what of small signals? Generally you can run electronics with slow speeds at large signal and they will be faster at small signals.

So like for example, if only 1/10 the resistance change is needed, will that occur in 10% of the time? Or is the opposite true, that the resistance change is delayed for a time and then it all sort of suddenly transitions? Typically we have the latter case.

Well it just so happens that I have wired up an Arduino board (Metro Mini from Adafruit) with an LED on pin 2 and a CdS cell on pin A0. If I apply a tone and experiment with biasing the optical output, will the CdS cell respond more quickly? An improvement of 20x would be sufficient for Bass signals, I think, and I like bass signals! Because it's all about the bass, right? OK well I'll try that tonight.

Les

[CreatorLes]

i dunno, maybe this is all wrong. Maybe I should focus on e-NABLE, that is charging up with other people joining the project...

Les