z axis feed back for under a dollar? - variable capactive-> freq-> mic

A while ago I posted a thread about a drip controlling valve being a very valuable first hack.
I did that because we are finding drips to be finicky to work with altho there still really great at being very finite.
That post was an expensive way to get drips to behave... this is a less expensive, and possibly better one.

The more I think about it this idea the better it gets, without getting more expensive!
Im alwase on the look out for an idea that is practically free, easy to modify, and takes up only a few grams when packaged in the peachy kit.

This one seems to be one of those so with that Ill get started.

A typical tone generating circuit charges and discharges a capacitor, it can be built just 2 transistors and a few other components at very low cost.
The smaller the capacitor the faster it charges and the higher the pitch is coming from the tone generator.

A capacitor is basically composed of 2 conductors with a very thin insulative layer in between them ... for example you can make one by sandwiching a piece of wax paper in between 2 pieces of aluminum foil. Slowly sliding one piece of foil off the wax paper would cause the capacitor to slowly drop in its capacitance.
If that varying capacitor was the one being charged and discharged in the tone detector, as you slid one of the foils sheets off of the wax paper the tone would very predictably rise in frequency.

now what if we did this....

Laminate one of the foil conductors so it has a varnish like coating.
Place that laminated foil strip along the side wall of your peachy printers top resivior, so that only the laminated portion is submerged in salt water.
Again you have 2 conductors both in very close proximity but electrically insulated from one another.
One conductor is the foil, the other is the salt water and these to conductors are insulated from each other via the very very thin layer of lamination on the foil.

As the salt water rises more and more of it covers the foil, increasing the capacitance very predictably.

variable capacitor--> tone generator = variable freq--> microphone input --> frequency mapped to height code = AWESOME PRINTS

I like this because its:
1 very hackable ( want a bigger printer, just get a bigger piece of tinfoil and coat it in varnish. )
2 its very low cost
3 it could be used in combination with the drip feed to help calibrate the size of a drip.
4 it still utilizes the mic input
5 it works despite the shape of your container. ( measures height not volume so tapered containers and fancy flower vases would just work )
6 it packs supper small and light.
7. it seems like the feed back would be far more explicit, liner, repeatable and predictable than drips, ( atho perhaps not a finite which is why its great that we can do both!)
8. You can try it now! and be a big help in the development of the peachy printer BEFORE you get your printer!

These are all great points but the first and last ones are my favourite.
I really mean it! It would be a big help if some of you amazing people in this forum would grab a simple transistor or 555 tone gen circuit, varnished or laminated tin foil and a container of salt water then Hack Post and chat about this idea until we know if its as good as it sounds!

a huge thank-you in advance to anyone that actually try's this and posts lots of pics and videos, for all of us to learn from.

PS here are some ideas that could make this idea better.

1 The foil could be rolled into a spiral shaped tube to get more surface aria in the water.

2 multiple layers of foil could be placed on the back and side walls of the printer to increase surface aria

3 insulated wire with just the tips striped could be placed at 4 or more heights in the upper resivior, each time one connects to the salt water it could make the tone jump up an octave (via conecting differnt resistances that charge the capacitor. each time a new resistor connects the tone would jump up in pitch. This could give us the explicite height data, making it posible to calibrate the printer automatically every print!

4 adding a walled off part of the bottom reservoir ( a wall with a small hole in the bottom )
you could make a resin free aria and add the variable capacitor to the bottom print aria as well, comparing the capacitance's might leave you with self relative instead of environment relative feed back, making it more reliable feed back.

5 placing a long wire 4mm away from the surface of the varnished aluminum would help reduce the amount of salt water electrons have to traverse to get close to the other electrode.
Doing this could reduce the effects of varying salinity and make for a better capacitor.


PPS

Any one want to take a crack at writing a python script that listens to the mic/ line in, returns a frequency and maps that frequency to height?
At the end of the day we need to know how well a computer can tell us the height of the water by listening to a real time microphone audio jack input.


Thanks again everyone, your ideas, hacking, and community nature are the best things that ever happen to peachy printer :)

valuable info from another thread:

Quote:

---

Originally Posted by User_Defined

......As for sensing the Z height, I think that the suggestion of using a strip of foil on the side of the top reservoir is absolutely brilliant. I would likely use copper foil as it is easy to solder wires to, and comes in standardized thin tape rolls.

This would only work with relatively thin dielectric walled containers, and would likely need a calibration step. You would also need to have another wire inside the reservoir to complete the circuit. Once calibrated, the capacitance should be linear with salt water height.

I don't know what the Peachy uses for electronics, but the way I would implement this is to use a 555 timer as a capacitance meter, and then send the square wave signal to the computer. Audio works best since you can just pulse modulate your drip received signal so every droplet interrupts the current state of the Z height. Imagine two people, one constantly expressing their detailed undying love towards the current height of the reservoir, and another impulsive and insensitive colleague occasionally interrupting them to announce when a droplet has been received.

Schematic: http://www.eleccircuit.com/simple-ca...-using-ic-555/
(One of many out there)

I would go for the newer 2MHz capable 555's since the capacitance may be in the pF range.
........

-UD

---

Unfortunately I dont have time to wire it all up and try it out, but it is a simple circuit!

- Some dielectric constants:
http://hyperphysics.phy-astr.gsu.edu...bles/diel.html

- parallel plate capacitor calculator:
http://www.daycounter.com/Calculator...lculator.phtml

Quick numbers: 2mm thick glass sheet 1"x4" total capacitance 57pF

Dont use varnish, I know a lot of good painters, none can make an even coating :p For thinner layers use thin shrink tubing maybe, shrunk around a metal tube?

57pf is not a lot to work with, BUT Theremins work with less! (So I know it is possible).

if you want a usable tone from 500Hz to 12kHz you need to have a capacitance in parallel with your sensor to give 12000Hz as a baseline, then from 12000 down to 500 will be the strip.

I used this 555 for a client:
http://www.intersil.com/content/dam/...icm7555-56.pdf
Page 5 Figure 2A to get 50% duty cycle always.
Page 6, equation: f = 1/(1.4*R*C)

Calculation: solve for R with C = x + 57pF producing 500Hz (or some lower value if you want)
Same R with C = x producing 12000Hz.

This is a pair of linear equations and so I wont go over how to solve it because HighSchool:
R = 22Mohm, x (the starting capacitance in parallel with strip) = 2.5pF, and the strip is 57pF when fully covered.

I don't think the resolution on this will be enough as a standalone system, but if you have the drip plus this you can have it constantly re-tuning your drip rate vs. time and get a very accurate readout.

ok enough slacking off for the day!


Someone check my work and then try this out!




-UD





EDIT:

Ok one problem I am seeing is stray capacitance changing the readout since 2.5pF is very small.

Maybe sandwich some foil inside some tape to get a higher capacitance, just don't paint it!


-UD

OK, don't paint the strip with varnish, then might I suggest a strip of aluminium tape covered with a layer of packing tape but use varnish around the edge of the packing tape to seal it from the water, if laid out carefully it should be very accurate.

do this in both containers for a real time comparison

Quote:

---

Originally Posted by Aztecphoenix

OK, don't paint the strip with varnish, then might I suggest a strip of aluminium tape covered with a layer of packing tape but use varnish around the edge of the packing tape to seal it from the water, if laid out carefully it should be very accurate.

do this in both containers for a real time comparison

---

Just tape some foil to the container wall... or laminate some foil or foil tape such that the edges of the metal are covered by the tape. You cant guarantee that the edges will be varnished, any bit of salt water touching that metal will cause a short, and varnishing thin flexible things is just asking for cracks and leaks.

Thin packing tape might work well, no bumps though! Personally I would go with something thicker and more consistent like kapton tape. It is made to be precise and is pretty rugged. Not to mention the polymer material might have a super high dielectric constant.

Oh I forgot to mention you should probably have the wire going inside the saltwater as earth/chasis ground. Also, get that wire to be as close as possible in the salt water to the strip. Aaaaaaand try to have the foil connection be as close as possible to the 555 chip to minimize stray capacitance to other objects.





-UD

If there are other capacitive loads on the circuit, then they would get accounted for in calibration. I imagine they wouldn't be significant in comparison to the varying capacitence with verying salinity of the water. Each time you change the water, you'd need to calibrate the system anyway to account for any change in salinity.

Quote:

---

Originally Posted by Feign

If there are other capacitive loads on the circuit, then they would get accounted for in calibration. I imagine they wouldn't be significant in comparison to the varying capacitence with verying salinity of the water. Each time you change the water, you'd need to calibrate the system anyway to account for any change in salinity.

---

Changes in salinity would not affect the capacitive nature of the system, only resistive. Since the charge discharge resistor is in the Mohm range, this would be a negligible change on the system. You would likely only need to calibrate the system once.

What would affect the system is say, having the foil in the container and then running a wire too far outside where people and things can impact the stray capacitance and change the baseline freq while it is working. If you can get it from 2.5pF baseline to 25pF, it should be much more stable and this is totally possible with tape and not 2mm of glass.

I drove to a local electronics shop this morning

I found some aluminum foil tape that was "50 mm"

but they only had kapton tape up to "50 mm" ...

Searching on ebay I was finding 60 mm kapton tape that would cover the aluminum tape perfectly, but that would be a couple weeks away.

Trying to decide If I can get the 60 mm tape in time ... or if overlapping kapton tape would be okay for this?
(would the overlap have to be consistent the whole length?)

the 555 timer they had was a regular one not the 2Mhz 555 timer that was suggested it might be needed for low capacitance's involved in this

I guess googling tone generator circuit might help me on that side

That specific chip I linked is actually 1MHz, I dont think a 2MHz chip exists.

The thing which worried me about the standard 555 is that I am not sure if the circuit config in the datasheet to produce a fixed 50% on/off will work. My client tried using a standard 555 and claimed it didn't work, but since it wasn't me doing it I have doubts about his claim.

Here is what I recommend with what you have:

Get the standard 555 circuit working with a more traditional 555 oscillator configuration.
I found this on youtube:
https://www.youtube.com/watch?v=stN-ZzHfiO4

I used a 22Meg resistor with 57pF from the 2mm glass. Let's assume you get 570pF, since the equation is linear youll want to be getting a resistor in there between 1 and 5Meg. If you have a variable 20 turn pot of 10 Meg it would be perfect as you could tune it to within audio range. Also, buy a few 200pF capacitors so that you can get it working first without fumbling with the salt water until it is ready.


Oh, and I forgot that not everyone has an oscilloscope! Maybe look for a 555 to speaker circuit like this:
http://www.eleccircuit.com/many-simp...ound-circuits/

OR buy an oscilloscope for $250 because the'yre awesome (probably go with the speaker for now).

Carefully cut your foil tape with a ruler and exacto-knife so that you get a more narrow even strip such that your kapton will overlap it.

You'll need to probably fold over one end of the foil tape because you can't solder to aluminum foil (don't try, you'll ruin the iron tip).

Then stick the foil tape to the inside wall of the top reservoir as vertical as you can. It isn't critical since you'll need to calibrate it anyways eventually and it will all still be linear. Make sure to have the folded tape part way above your water level.

You need to completely and reliably cover that foil tape with the kapton tape. No air bubbles! Line it all up and slowly press it on?

You will need to put one capacitor wire from the breadboard into the salt water (ground?). use a long wire, have it in close proximity to the foil tape but not too close (maybe tape it right beside with your kapton). The other wire of the capacitor you should clip on the foil. Try to keep the whole thing tight (make sure to protect against splashes from the water on the electronics).

Maybe before even adding the salt water, you could have another piece of clipped foil and cover the sensor to test it without the water.

Good luck!


Oh, and if you do overlap the tape, it needs to be perfectly even. I see more success in cutting the foil to be even.

Good stuff. Maybe use more aluminum. I know the salt water high voltage capacitor, aluminum foil wrapped around a plastic bottle, is 1800pF. http://www.instructables.com/id/Make...p10/Completed/

And a typical Leydon jar made of glass had about 1000pF capacitance. That is an easy value to work with. Something like 30 square inches of foil, minimum would be needed.

I like this whole idea as well. A sound card can pickup 20-20khz. A quart jar has about 1000 ml in it and 20 ml per drop, that is 20,000 drops per quart or liter. This might be accurate enough to distinguish a 1 drop height change. In real world, it would be less, but 1/10th of that is still accurate enough. For instance, 2000 steps (up to 20,000 possible) in about 4 inch (100 mm) height change in water, would be 0.05 mm Z axis accuracy. That is a 50 micron accuracy, with a possible 5 micron accuracy.

Quote:

---

Originally Posted by Spelljammer

Good stuff. Maybe use more aluminum. I know the salt water high voltage capacitor, aluminum foil wrapped around a plastic bottle, is 1800pF. http://www.instructables.com/id/Make...p10/Completed/

And a typical Leydon jar made of glass had about 1000pF capacitance. That is an easy value to work with. Something like 30 square inches of foil, minimum would be needed.

I like this whole idea as well. A sound card can pickup 20-20khz. A quart jar has about 1000 ml in it and 20 ml per drop, that is 20,000 drops per quart or liter. This might be accurate enough to distinguish a 1 drop height change. In real world, it would be less, but 1/10th of that is still accurate enough. For instance, 2000 steps (up to 20,000 possible) in about 4 inch (100 mm) height change in water, would be 0.05 mm Z axis accuracy. That is a 50 micron accuracy, with a possible 5 micron accuracy.

---

What about a laminate of several layers of aluminium and kapton, would that work?

I had been thinking about this. So tried it out with a foil wrapped container:
Attachment 1453

With a 4.7 Meg Ohm resistor I can hear the tone - and here is what I could see:
Attachment 1456Attachment 1457

add some water (next post because of image issue.)....

So added water and got:

Attachment 1458Attachment 1459

And clearly the tone changes.

Video posted here of filling the container. http://youtu.be/YpQnxEugeLQ

So, there is a lot of noise. I might be able to lower the resistance value to see if the noise reduces. For some reason, the act of filling the container seems to change the tone, regardless of level. While water is pouring in, the tone changes, as the level goes up, that changed tone drops, then when the pouring stops the noise steps again. I don't know if the stirring of the water as the pour happens has anything to do with that.

If anyone can find me any good PC sound card signal analysis (e.g. frequency sampling tools) I can plot the changes against height of water.

The issue that I see here is that you wrapped the whole container with foil, and that creates a large surface area which can store charge (like those globes you put your hand on and your hair stands up). It might not be that much compared to the inside capacitance, but it could introduce some tone jitter.

Did you make sure to pour salt water in?? (or was is simply normal tap water?)

The act of pouring a lot water in might cause weird ionic effects within the water to throw things off. Also keep in mind when water is being poured in, it makes an electrical connection between you and the system via the water stream, and that will throw things off too.

This is why the inside salt water probe needs to be earth grounded (and you as well).

You should do it in shots.: Take a shot glass and transfer water a bit at a time, slowly, to compare the tone.

Also make sure you have enough salt in the water.


Awesome work!

- UD

Ya! Awesome Work! colehard
from what I saw this is still looking really promising