Preliminary layout and schem for proximity control... check?

[NaBo]

I had a few questions, which can be found in this thread... but I was too antsy to wait for answers (read: i'll do anything to avoid studying) so I went ahead and drew a schem and layout in ExpressPCB.  It's this application note from maxim, without the "on/off trigger stage" at the end.  I'm hoping like this it'll give a smooth control over the LED brightness in an optoisolator.

Can someone check over what I have already for me?  Is the extensive ground planing a good idea since it's a single-sided board and they suggest using a board with an embedded ground layer.  Am I right in connecting all unused pins directly to ground (except for the specifically declared "NC" pin in the MXO45 oscillator)?

I'm kind of proud of this pcb... it's my second and it's starting to show some type of logical organization.  I'll be MUCH prouder if I end up putting together a decent proximity control unit out of it, too.  Suggestions and comments appreciated and needed.

Hope this is everything ya'll need...  I'd really appreciate any input on this before I start tinkering next week

1.The modified schematic, (basically just chopped off after TP2... made this to link the PCB to in express)

2. PCB and Parts Layout

3. Relevant Datasheets:
-MXO45 Oscillator, (layout accomodates the half-sized one... same price and smaller)
-MAX407, dual op-amp, (only half used in this application, but same price as the single ap-amp and in and outs on same side)
-MAX912, dual comparator
-74HC86N, quad XOR gate

[Transmogrifox]

The schematic looks solid to me.  I didn't have time to browse the PCB layout.

It looks like a capacitive stud finder to me!!!  You're a stud NaBo.

Thanks--I actually benefitted the most by being introduced to that comparator.  I have been looking for something with better than 20 ns response time in a DIP package for exploring PWM variable resistors, and for a PWM amplifier.

[The Tone God]

Before you jump into this and etch boards and the like try bread boarding it. You don't want to put alot of time and effort going into something that doesn't work in the end. Someone even in that linked thread said they couldn't get it to work. Maybe you should find out why.

My reaction when I look at that circuit is alot of parts (some specialized) to do little. There is a 5v-1Mhz signal which I don't think is going to offer much range. The circuit itself is similar, if not overly complex, to what some of the simple theremin circuits do which is have a oscillator with an antenna attached feeding a gate which has a second oscillator to serve as a nulling signal. Nothing special. You should notice that the app note's goal is not about making a proximity detector per se but an example for bragging on how the opamp is tempature compensated.

All of my circuits that I have designed use significantly less parts. Some of them use as little as three parts. I will recommend keeping away from circuits that use coils.

My original advice still holds. Do some research and design a solution for the problem instead of grabing circuits not knowing what they do and hoping that they work. This will be a good learning experience.

Andrew

[NaBo]

Mmhmm, don't worry... this will be an ongoing learning experience for me.  Most of my stuff is in storage or packed right now though, so I can't really do much experimenting at the moment... but I definitely (and rather obviously) can't wait to get started!  So i figured just in case i get this thing up and running on breadboard, i'll have a decent layout I can modify a bit if necessary, and toss into an enclosure and start working on the opto's!  (little overeager, hunh?  I know  :wink:)

It's more than likely going to be a long and protracted, sometimes frustrating process, but if the end result is a useful control unit... extra components or not, it's definitely time to play some serious guitar for a couple months.    I sometimes forget my primary goal is making music... all this stuff isn't JUST a stepping stone to get there, but thats definitely a part of what drives me

I registered for my distance ed electronics technician course last week and im gonna go pick up the materials on friday.  I understand it comes with one of those fancy-software-circuit-emulators too!  So in addition to the wealth of knowledge i hope to amass over the summer while studying something im actually really interested in, that should help me out in my tinkering.  We'll see how the story unfolds!

Thanks for the input you two- both of you guyse's'ese'sses work has taught me a lot, and will continue to as I return to it from time to time with a better understanding of whats really going on.  I remember the first time I saw your schem for the proxyossi... I was all " uh... :?  ....  :?:"...  now I'm starting to say "ohhh, so THAT's what's going on here".  The Rock 'n' Control article is gold,  it's really helped me a lot with implementations of LDR controls and opto's in the past few months, and I suspect that trend will continue.  So yeah... thanks guys!

[Transmogrifox]

I think you should give it a try.  It will work fine as long as it's stable, and I think with a good crystal oscillator, metal film resistors and temperature stable (NP0) capacitors, you should be able to use it fine.  You may need to add extra gain on the output, but I don't think that's a real big deal if the input is pretty dependable.

[The Tone God]

Thanks NaBo from a fellow Torontoian.

Here is a fun little circuit I came up with that you can experiment with. Take a sensative comparitor (or opamp) tie the + input to power. Connect a resistor (say 100ohms) from power to the - input. Connect a small cap (say 10pF) from the - input to ground. Connect a metal plate to the - input. Watch the output when you touch the plate. You can throw an LED with an appropriate current limiting resistor off the output to see whats happening if needed.

This won't work beyond bare skin (maybe a thin glove) or give you any serious range but its simple and useful to understanding the situation.

Now ask yourself why does it work ?

Andrew

[Paul Perry (Frostwave)]

Anyone who builds this will certainly learn SOMETHING!! :wink:
And, at 1MHz, the stray capacitances on a prototyping plastic board might be a serious problem.

[troubledtom]

it's cool that you have the guts and time to try to pull this off.
        bestwishes
               - tom :wink:

[Zero the hero]

Nice to see that this application note is usefull!
Nice job with the layout!  

Here my impressions on the Maxim appl. note...
I've breadboarded it one month ago. I used a TTL square wave generator instead of a single chip solution. The oscillator runned at 1 MHz, everything seemed to work fine and with an oscilloscope I noticed that moving my and across the plate changed the duty-cycle of the oscillator, so the waveform scoped at TP1 was "moving".
Unfortunately, the variation on the duty cycle was very very small, and it achieved few millivolts of difference at TP2, not enough to control a VCA or a J-FET.
That's why I said that I had not much fortune with this circuit. Maybe I just need to remove it from the breadboard and draw a layout for it.
I believe that carefull layouting is needed for 1MHz signal routing...

I noticed that using shielded cable influences a lot the behavior of the circuit. Even the lenght of the cable plays a role. And the plate dimension is very important: bigger surface causes a bigger variation no TP2.

I have another schem floating around, Im going to build this one...
http://www.physics.gla.ac.uk/~kskeldon/PubSci/exhibits/E9/cir2.gif
for our purpose, we just need the first part of this schem: follow the signal and stop before it enters tot the VCO (pin 8 of ICL 8038)

NaBo, wanna write an article and post it on my site? I'll build this controller and see what happens.

[NaBo]

Thanks Zero!  I've already noticed a couple things i missed with the layout (like a couple pins supposed to be N.C.).  But hopefully it will lighten the work load and shorten the building time once I figure out something functional.

Good to find out what exactly wasn't working with yours.  Like Paul said, it's possible breadboards might hurt us more than help us   ... I HOPE i can get away without having to make a PCB and solder and desolder to experiment... but I'm willing to do that if necessary!

I'm definitely going to take the wire's capacitance into consideration...  6" is a ton of wire in stompbox terms, so i'll try reducing the length and increasing the 33pf cap to maintain the delay capacitance.  I'll have to check the capacitance of the shielded cable i have... rg174 and some rg59 i think.  I'll also experiment with different antenna dimensions... it's nice cuz you can still etch and drill the one's that don't work well!

It'll be a while before I can start testing this out, a longer while til I get the really precise resistor values from mouser, and probably an even longer while before the thing works half-way decent ... I'll definitely keep a log of everything though, and anyone who wants to post that anywhere can do so when it's done.

Good look with your's Zero.  Looks like the 1/2 a theremin one... it's striking how similar the design is... can definitely spot the building blocks that are similar between the two.  Hopefully we both come up with winners!

[Zero the hero]

About resistors, for 499 an 4,99k you can parallel two 1k and two 10k 1% resistors.
For the 4530 ohms rez, I have to check what values I used (2 parallel resistors, if I remember well).
I used a short (15cm) shielded cable that I found in a VCR. Its capacitance was about 15pF, nice value here.

I have no time to build the theremin like controller I posted: I got all the components but I'll be busy till friday.

[The Tone God]

Unfortunately, the variation on the duty cycle was very very small, and it achieved few millivolts of difference at TP2, not enough to control a VCA or a J-FET.

Thats pretty much what I expected. I doubt playing with the layout or few part values will really add any serious enough voltage swing to drive anything.

I have another schem floating around, Im going to build this one...
http://www.physics.gla.ac.uk/~kskeldon/PubSci/exhibits/E9/cir2.gif

The circuit is quite similar in function to the app note and to other theremins in general hence why I commented on how the app note is overly complicated. I don't think that circuit will give a great range either but it depends some what on your antenna design. You will have a fun time tuning that circuit for null and it will drift sometimes require recalibration.

It should be pointed out those theremins run in the audio range so the signal will leak into the audio stream. Thats the thing with many theremins. You are manipulating the output signal directly as opposed to controling a parameter of the output.

Try out the circuit I mention above. After you guys try it I'll post another circuit for you guys to test out and learn from.

In the meantime if you guys are still serious here a few things to read. Read them to understand the problem NOT to try to duplicate.

http://www.electroline.com.au/elc/feature_article/item_042002c.asp

http://www.sensorsmag.com/articles/hot1099/main.shtml

Andrew

P.S. Are you guys beginning to understand why I made the suggestion of a light controlled "gheto theremin" for simplification reasons ?

[Zero the hero]

What about this one?

http://www.bik.com/

[NaBo]

Thats pretty much what I expected. I doubt playing with the layout or few part values will really add any serious enough voltage swing to drive anything.

Aww man, and you were gonna let me order parts and build it and everything?    I'm just kidding.  I guess I really should just bone up on all the information i can first, so I can hopefully approach this with knowledge in the brain, a blank paper, and a pen.  Besides, I have way to many other projects to finish as it is :wink:

Try out the circuit I mention above. After you guys try it I'll post another circuit for you guys to test out and learn from.

How could I refuse this offer?!?!  Sounds good to me!  I'll try it out soon and let you know what I come up with.

In the meantime if you guys are still serious here a few things to read. Read them to understand the problem NOT to try to duplicate.

Yeah... something tells me if I just built it and it worked right away I'd feel kind of cheated...  I'm sure it might have dissipated as soon as I brought it to jam and made my friends shit themselves, but I think I'd rather go about this how you suggest it.  Teach a man to fish...

P.S. Are you guys beginning to understand why I made the suggestion of a light controlled "gheto theremin" for simplification reasons ?

Yeahhh... but I already feel like I've been there, done that  When I box up my light-controlled Aww-Wah pedal, I'll have to post some pics and clips.  I think I've done a pretty good job of thinking that project out thoroughly, so hopefully it turns out how i've envisioned it  

Anyway, thanks for your help and guidance.  I'll tell you how that little circuit works out and what I think I learned from it soon

[Jaicen_solo]

Tonegod: I simulated that little circuit in MultiSim, and couldn't really get anything to work. Basically it output DC at 5v (using a 5v supply).
Do I need to do something to get it oscillating??

I've been working on designs for a proximity detector myself, using a modified theremin pitch controller. Although it uses inductors, that's not a problem since I can wind my own.
A guy I found sent me a schematic which should work, but as of yet i've not prototyped it, i'm a little intimidated by it to be honest!
What i'd like to know is why you guys seem to prefer CMOS oscillators to discrete transistors. Is it because they're more predictable??

Has anybody actually got a theremin style plate to work??

*b

[puretube]

yes: Z. Vex !  

[Jaicen_solo]

LOL! Behave yourself.

I meant a DIY solution. I plan to use mine to control a number of pedals using a 'distributor' box.
Actually, I'm surprised Zach isn't marketing the 'Probe' as a standalone product.

[puretube]

? too little stompboxes with "standardized" CV-input jack available on the market?

(I`d gladly help out...)

[Jaicen_solo]

Yeah, i've thought that but the design i'm thinking of uses variable resistance, which can be tweaked to give a range of control. I don't understand why guitarists don't want realtime control over EVERYTHING! Am I the only one??  
The synth community seems to be more demanding of things like that.
Maybe it's the sound engineer in me, but I like to automate my effects, I like to be more expressive (though that could be to detract attention from my poor skills!).
Any help you give would definitely be appreciated!

*b

[The Tone God]

How could I refuse this offer?!?!  Sounds good to me!  I'll try it out soon and let you know what I come up with.

Cool. If your part shopping get a 1M pot, some diodes (1N4148 or smiliar), and a MOSFET (our common BS170 or 2N7000 should be fine).

Yeah... something tells me if I just built it and it worked right away I'd feel kind of cheated...  I'm sure it might have dissipated as soon as I brought it to jam and made my friends shit themselves, but I think I'd rather go about this how you suggest it.  Teach a man to fish...

Build a QT/CT sensor that work right away ? That would impress me greatly. The reason I suggest not trying to copy it is that it is very complicated and to be honest beyond what we would need/want. In those articles you should read the stuff about what the sensor senses in terms of the human touch interface. Skip the charge transfer,  burst,  nulling, etc. stuff since that applies to the sensor they are trying to sell. (Remember alot of datasheets/app notes are technical flyers to promote the company's product.)

Yeahhh... but I already feel like I've been there, done that  When I box up my light-controlled Aww-Wah pedal, I'll have to post some pics and clips.  I think I've done a pretty good job of thinking that project out thoroughly, so hopefully it turns out how i've envisioned it  

I didn't know someone had already tried it. I had some ideas for an article about it but I won't be able to get to it anytime soon. Glad to see someone gave it a go.

Tonegod: I simulated that little circuit in MultiSim, and couldn't really get anything to work. Basically it output DC at 5v (using a 5v supply).
Do I need to do something to get it oscillating??

Nope. It doesn't oscillate nor does it need one. The comparitor/opamp is running in the standard comparitor mode so it should stay high until your finger touches the plate. Probably because your not simulating the touch over time is why your are not getting a result. Note: This is one of those times where building the circuit takes less time then trying to simulate it.

I've been working on designs for a proximity detector myself, using a modified theremin pitch controller. Although it uses inductors, that's not a problem since I can wind my own.

I have this habit of making sure I can reproduce things in a reliable and consistant manner. The can be especially true when dealing with DIY stuff that I publish. Inductors are tough because of that but it not need for this case if you smart about the design. It still fun though ain't it ?

A guy I found sent me a schematic which should work, but as of yet i've not prototyped it, i'm a little intimidated by it to be honest!

I haven't look at the schematic but from the parts list and a few of the docs it does seem alittle overkill. 8038s, HV, etc. scare me.

What i'd like to know is why you guys seem to prefer CMOS oscillators to discrete transistors. Is it because they're more predictable??

Partly. They take few parts (less parts means more stabilty most of the time), usually there are gates lying around the workshop that can be used, and they usually cost the same or sometimes less the descreets. Also typically we use gates in other parts of the circuit so its convient to use a package full of them. If all I needed was one little oscillator I wouldn't pull out a big DIP package set of gates just for one gate.

Has anybody actually got a theremin style plate to work??

Yep.

I will state it now just in case anybody missed it in the last thread. I will not hand over any of my completed designs. I don't mind giving hints or helping people out with what I know when I can.

Andrew