Multiplexing project

[alec_t]

@alec: couldnt I just have 16? because I would only be applying power to the pos and neg of the grid?

I don't know how you intend to apply power just to the pos and neg of the grid; or did you mean to rows and columns?

 

[mrn]

@alec: Yeah, the rows and columns

 

[mrn]

I was also wondering on the transistors, for the drivers: when a transistor gets a signal from the 595, it switches allowing the external power to go threw and drive the coil, however how do I switch the transistor back? or does it only hold the state as long as it has power? I wasnt sure.

 

[mrn]

@Chris: Why couldnt I use 8 transistors for the negatives and 8 transistors for the positives? Why would I need shifters?

 

[ChrisP58]

To switch the positive voltage to the rows you need PNP transistors. PNP transistors are switched by pulling their base pin negative with respect to the emitter. If the emitter is at 12V from ground, then any base voltage less than 11.4 from ground will turn them on. Since the outputs of the 595s go between 0 and +5 volts, the PNP transistors will always be on.

The level shifter will be an NPN transistor with a grounded emitter. The collector of that npn will drive the base of the pnp through a resistor.

 

[BeerBelly]

Why couldn't I use 8 transistors for the negatives and 8 transistors for the positives? Why would I need shifters?

We were talking about a 64 x 64 matrix. You could use 64 digital NMOSFETS for the negative lines(rows?) and 64 digital PMOSFETS for the (columns?).
These transistors are a reasonable price and may be the way to go for a small prototype.

 

[mrn]

@Chris: So basically there is no way to turn off the PNP so you need a NPN to basically invert the signal of the PNP? the PNP wont switch back to the first state after the signal form the 595 is gone? Basically the signals from the 595's wont be constant they will have breaks in between.

@BeerBelly: Ya they are cheap, around a $1.50 on mouser, however 128 of these (64 and 64) would get pretty pricy. Is there any other simpler/cheaper solutions

Unless of course I can see if I can get the coils to work with 5 volts from the 595's, I do more winds on the coil, currently I only have 8 winds, however this would require more wire, obviously.

 

[BeerBelly]

@BeerBelly: Ya they are cheap, around a $1.50 on mouser, however 128 of these (64 and 64) would get pretty pricy. Is there any other simpler/cheaper solutions

Unless of course I can see if I can get the coils to work with 5 volts from the 595's, .

These digital NMOSFETS are $5.10 for 100 of them
https://il.mouser.com/ProductDetail...GAEpiMZZMvplms98TlKY6v%2b9m7cya7PfMZkJSjtm40=

These digital PMOSFETS are $7.60 for 100 of them
https://il.mouser.com/ProductDetail/Fairchild-Semiconductor/FDV302P/?qs=sGAEpiMZZMvplms98TlKYyI%252bsdYAyN1NEIV6nMu%252bqrM%3d

These MOSFETS may not do the job, you need to measure and experiment some.
Other more powerful MOSFETS are available at just a little higher price.
You are going to need a spike suppression diode across each MOSFET also.
$3 for 100 https://il.mouser.com/ProductDetail/Rectron/1N4007-T/?qs=sGAEpiMZZMvplms98TlKY27629dpP7N7zrIE40%2focdo%3d

The 595s will not drive a coil by themselves period.

 

[ChrisP58]

It's going to be difficult to choose the right power transistors until you know how much current is really needed. Guessing will either give you a part that is too small but cheap, or overkill and expensive.

In the meantime, let me ask this. How big is this thing going to be? And how is it to be built? Will the whole thing be on a single PCB, including the actuators, or will you have a PCB for the electronics and a bundle of wires going up to the actuator array? Will the circuitry be handbuilt on perfboard, or will it be an etched printed circuit board(s)?

 

[mrn]

@BeerBelly: Wow okay, I was looking in the wrong spot then. I thought the 595's can output 5 volts though? Or is that pushing it?


@Chris: Right, I am working on that now. Well the grid is 8 inches by 8 inches there will be 8 little coils/valves per inch. The entire thing will probably be at least 8 inches tall.
For now I will just have the electronics probably bread boarded, and just sitting next to the machine.

However in the future if this first build goes well, I would like to create a single board which houses all the electronics, maybe slim it down as well and replace the 595's with a 1 or 2 chip solution to save on space.

the coils will be tiny valves almost microfluidic like, however not as small. the coils will be separated from the other electronics, I will just have a bundled cable from the columns/rows, to the electronics(for now anyways)


I will be in town today, I will see what they have at radio shack, as far as electronics. Probably allot more expensive there though.

 

[mrn]

I had another quick question, I just bought some .50mm gauge wire, and forgot to check if it was insulated.

Someone told me I could use like craft paint or nail police, and coat the wire with this to form a good insulated. Would this work?

Also I know you need magnet/insulated wire for a electromagnet, but I've read somewhere that you don't absolutely need it, so basically the wire would be come one chunk of metal and draw a ton of current, considering these will be tiny electromagnets, could I just use this un-insulated wire?

 

[ChrisP58]

To check if the wire is insulated, just touch your ohm meter probes to it a few inches apart. There should be no connection. The wire DOES need to be insulated, or the winding will be just a big conductive tube. You need to separate each turn from it's neighbor to ensure the current goes around and around. Each turn builds a stronger magnetic field.

If uninsulated yes, you can coat it with something, but you need to make sure that you give it a good, uniform coat without any pin holes. But doing that without making the insulation too thick might be a pain. The best solution is just to buy "magnet wire." While you may not find it at your local hardware or craft store, it should be readily available at any electronics store or online supplier. Also motor repair shops will use it, but may not have the small gauge sizes.

It is manufactured specifically for winding magnetic coils, including solenoids, motors, transformers and inductors. While regular hookup wire generally comes only in even gauge sizes, magnet wire comes in the odd numbered sizes, and some suppliers even make 1/2 gauge sizes of magnet wire.

There is also a variety of insulation types that are used. Enamel has been the standard for a long time, but I prefer polyurethane-nylon as it is heat stripable with a hot soldering iron. Enamel needs to be scraped.

The NEMA code for what I use most often is MW-80-C

* National Electrical Manufacturers Association.

 

[mrn]

@Chris: Aw didn't think of that.



I was thinking of have a small tub or container full of paint, and just run the wire threw it rather then taking a brush and painting it, and to make sure I get every last bit of it covered. The only problem I see with this is drips and dried access paint, I will have to try it with a small length of wire and see how it works.

 

[mrn]

So it turns out the paint isn't working, so i'm just using black electrical tape for now. However right now I am having a hard time trying to find a big enough power source to test the coil with, a 9 volt battery doesn't do much, a 5v 800ma charger does worse and I tried a old PC power supply but when ever I connect the coil, the power supply seems to shut off or something, witch is odd I don't know why because I just have a paper clip with 24 gauge wire rapped around it about 6-7 times. the wire is rolled-up inside electrical tape. I'm guess the psu shuts off because it is drawing to many amps, however I don't know why this tiny coil would draw so much.


However if you know the old trick of jumping a old pc psu, if I jigle the jumper wire on the main 20 pin connector, the psu will kick on for a second then shut off, and one time during this test, I got a small piece of metal to jump to the coil witch is exactly what I need.

I looked at the specs on the psu and it says 12v 13 amps and 5v 25 amps

on all the 4 pin connectors of the psu I found that the one black and yellow is 12v and one red and black wire is 5v if that makes sense

so basically then I will need 5v 25 amps to shoot threw each coil. This is of course with very few winding's, however if my math is correct, a winding of 4 inches long around the diameter of a paper clip, will allow me to do 300 coils with a 100 ft length of 24 gauge wire.

I bought the wire for $9.00 at a small hardware store, it was a ripoff the more I think about it, I found 312 feet of magnet wire on radioshack for the same price afterwords.

 

[ChrisP58]

The current draw of a DC coil is all about the DC resistance. 24 gauge copper wire has a dc resistance of ~25.6 ohms per thousand feet, so your 4 inch length is about 8 milliOhms.

The Ohms law equation for this is Volts/Resistance = Amps.

Even if you run it on a 1.5Volt flashlight battery, it theoretically would draw 180 amps. But even a D-cell can't deliver that much current, so the terminal voltage will sag considerably. Probably down to near zero.

I'm afraid that, to make a practical coil, you will need to wind it with many turns of much smaller wire.

You can get resistances for other sizes **broken link removed**:

 

[mrn]

@Chris: First off, a very good resource I will keep this handy

You would think though that the PSU would just limit the current and not shut off, the little power adapter 5v 1 amp still works it just limits the current i'm assuming?

I actually got a six inch piece of the 24 gauge wire, with electrical tape around it, and wrapped it around a paper clip, I used that small 5v 1 amp power supply/adapter

I can pick up a piece of metal about the size I need from about 1/10 of a mm away, this is not that far however, in this case, Its close enough, i'm thinking the electrical tape is pretty thick, so if I can figure out some other type of insulation, then I might get even more strength out of the coil, that or just get some magnet wire which I eventually will do.
I could also find some sort of iron core, instead of a paper clip, that would help with the strength.

Again I know 1/10 of a mm sounds crazy but its just about right.
I now just have cut the weight of the tiny valve I have in half. Time to get out the magnifying glass

 

[BeerBelly]

Iron(a nail) for the core of the magnet would probably work better than a paper clip. Use a fine piece of magnet wire more like 10 or 20 feet long instead of 6 inches.

Think salvage.
Fine magnet wire can be salvaged from an old wall wart power pack or finer wire still from a torn speaker. An old stereo or tv will keep you from placing 100 orders when experimenting.

 

[ChrisP58]

There are power supplies that are current limiting. Most bench top instrumentation type supplies are like that. But the behavior of the current limit function is that they reduce the voltage to whatever level keeps the current at the limit. Most circuits don't like to be run at a voltage lower than spec, and PCs are no exception. So, most purpose made supplies protect against overload conditions by shutting down, rather than reducing their output voltage.

With your 6 inch piece of #24,(at 0.013 Ohms) a 10 amp current limited supply will push 10 amps through the coil, but the voltage across the coil at that point will only be 0.13 volts.

 

[mrn]

@Chris: so basically using 24 gauge wire is only dampering the power of the coil? I guess I will switch to magnet wire then, the only problem though is I will need alot of wire times 6 thousand coils.

I just tested it with the multimeter, im getting 0.01 amps and 0.02 volts I don't understand though because based on stuff that I have read online, people say 24 gauge wire will work.

@BeerBelly: That's actually what I did first with a old power adapter, however I switched to the bigger wire I didn't think it would make that much of a difference, but I guess so.

 

[BeerBelly]

The point of using the finer wire is to get more windings on the iron core.
Also there is a lot longer wire in a pound of #40 wire than a pound of #20 wire.