my first soldering project is a simple (midi-)controller with a realtively large number of potentiometers (~50) of which the values are read by an arduino to send to the music software. I use 16-channel multiplexers to "bundle" the signals into the existing analog inputs.
The problem: I think that I have short-circuits (bad soldering) and cannot even trouble-shoot the connections properly because the amateurish design. I.e. the ardiono turns off as soon as I attach the power and ground pins.
Question 1)
Do you have any advice on how to improve the design to make it more cleaned-up?
Question 2)
My thought was, that instead of powering the individual potentiometers in parallel (current approach, see attached image) they could be powered in series to avoid cable clutter. Does that work at all? Can the arduino power that many potentiometers in series?
Would I be out of order if I said that's a fookin mess?
OK, you're new to this so you are forgiven.
2 things.
I suspect that what you mean by series and parallel are not what I think of as series and parallel. It would help if you could provide a circuit diagram. Hand drawn and photographed is fine.
You need to organise the wires to run in rows and columns then tie them together with cable ties or similar. You need to plan a route to each pot for the wires. Typically you would come off the pot, along a row then down a column keeping the wires tucked in between the pots. Each pot on the same row or column should have wires taking the same route. make sure that the wires are just the right length to go where they need to go with only a small bit of spare, and all the same.
I see you did stick the wires in the holes of the pot connectors.
I hope you didn't solder them there yet, otherwise you have to throw away all the pots.
Soldering them in the holes will get flux between the metal connector and the carbon track, making the pot very unreliable.
Next time get pots with solder lugs instead of pins. These ones with pins are made to mount on a circuit board.
If you have used 10kB pots, then each pot draws 5/10k = 0.5mA, so 50 of them draw 25mA.
Not a problem.
And no, you can't connect them in series.
Tip: use a long wire, and carefully make a cut through the insulation every few centimeters (pot distances).
Then pull the insulation a few millimeters apart, and wrap that bare wire once or twice around the pot pin.
Solder, and move to the next pot.
That way 5volt/and ground to the pots are neat, and only two wires are used.
The wiper can be done with individual wires, but mount the 74HC4067 board near the pots.
Leo..
You probably need a separate power supply. No, you can't wire them in series.
If you don't have a multimeter you should get one.
Do you know [u]Ohm's Law[/u]? (Current (Amps) = Voltage/Resistance)
With 10K pots, each pot will "consume" 1/2 milliamp. So, 50 of them is only 25mA and that should be OK powered from the Arduino. 1K pots might be pushing it.
If you disconnect power and check the resistance, 50 pots should measure 1/50th the resistance of one pot.
If the pots are wired correctly, the current shouldn't change when you rotate the pot.
OK, assuming this is a short you are dealing with, I'm going to give you some bad advice. Here are two ways to possibly ferret out a short:
Using a bench supply that can pump out some serious amperage, but also can operate in a current limit mode, set the current limit to some low value, like 100mA or so, then remove the Arduino, and apply 5 Volts, from the Bench Supply, across the 5V and Gnd points on that array of pots. Notice what voltage the Power Supply [PS] drops to, and if it's reading 0V or nearly zero volts, crank up the current limiting until you get a voltage indication on the PS. If you get up to about an Amp and still 0V on the PS [or a voltage too low to be useful], then go to suggestion #2 [below]. Otherwise, using a voltmeter, set to the lowest range that it has [most Voltmeters can go as low as 200mVFS or lower], connect the Negative probe to the PS Negative terminal, and probe around and follow the voltage to where it is at the halfway point.. The short will be near that point IF the short is a higher resistance than the wires. If the voltages you are reading don't change enough to reveal the short, try cranking up the current limit, but, again, if you get to around an amp, go to #2. You can also !CAREFULLY! feel for warmth in the wires and pots. If any of the wires/pots are getting warm, !CAREFULLY! follow the warmth, hopefully, to the point of the short. If the short has a similar resistance to the wires, then this probably won't work.
As you watch for smoke, keep cranking the amperage [on the Current Limit] up. If the PS maxes out, either get a beefier PS, or start cranking up the voltage until you see smoke, something explodes, our you loose your nerve. If the current reading on the PS suddenly drops, then either you blew out the short, or burned something else open. it also might not be a bad idea to do a little math, and determine the max current and voltage the pots can take -- so you don't blow them out -- otherwise, if you see smoke rising from numerous pots, you've probably run out of head-room and should abandon this method. Also, that would be a good indication that something else is wrong.
Again REALLY BAD IDEA!!!!!!!!!!! -- like, maybe, a last resort measure!
--But, I've actually used this technique to find/clear shorts on a PCB -- so it can work -- but it's kinda dangerous -- especially if you don't know enough about electronics to take the proper precautions, or have an intuitive gut to guide you.
NOTE: It may not be a bad idea to rig up a way to quickly kill the power -- like, plug the PS into a power strip that has it's own switch. That way, if something starts to smoke, or goes up in flames, you can kill the power right away. Might also not be a bad idea to have a Fire Extinguisher in easy reach, and do this in a well ventilated area [like in a garage, with the garage door open].
PerryBebbington:
Would I be out of order if I said that's a fookin mess?
[...]
I suspect that what you mean by series and parallel are not what I think of as series and parallel. It would help if you could provide a circuit diagram. Hand drawn and photographed is fine.
A fookin mess it is, most definitely. Now it's about making it better!
Concerning the "powering in series" (obviously I used the terms inadequately) I made a quick photoshop sketch (without proper knowledge of all the proper signs that are usuall used in circuit diagrams) to show what I mean. From my unexperienced point of view it should work like that, no? What do you think? As far as I can tell this would also translate to what @Wawa suggested in his answer:
Wawa:
And no, you can't connect them in series.
Tip: use a long wire, and carefully make a cut through the insulation every few centimeters (pot distances).
Then pull the insulation a few millimeters apart, and wrap that bare wire once or twice around the pot pin.
Solder, and move to the next pot.
That way 5volt/and ground to the pots are neat, and only two wires are used.
The wiper can be done with individual wires, but mount the 74HC4067 board near the pots.
Leo..
@PerryBebbington, @Paul__B and everyone else: Cheers for the great replies! That's really assuring and keeps up my motivation for this project.
@ReverseEMF that's it. now it makes sense too so parallel was what I was doing from the start and my optimization was just "parallel with a LOT less cables". thanks for the clarification.
now I modified the wiring of the first row of pots. however, I get no meaningful reading on my arduino from the modified pots. I think I destroyed them by un- and resoldering. but it's a bit suspicious:
the voltmeter applied like shown in the pic below gives me a reading of 5 volts at the last pot in the row. so I assume current is flowing in appropriately. none of them produce a meaningful signal reading even when I apply power directly with separate wires (I checked with a fresh pot that it's not the Arduino signal input that's broken, it works).
I also unsoldered the second row of pots but didn't resolder them yet. All of them work fine when I apply power individually with debug cables. so those seemed to have survived the initial soldering + the unsoldering just fine. also all the pots I checked from the lower rows work.
Q1) Can it really be that most (or all) pots survive 2 soldering sessions but all 6 that underwent a third soldering session get destroyed? Just seems so unlikely. I didn't change the way I applied the soldering iron or anything else.
Q2) I think that is really the question: When the voltmeter reads 5 volts but the pots don't produce a meaningful signal, then is it for SURE that the pots are broken? Or could it still be faulty wiring?
oystersauce:
Q1) Can it really be that most (or all) pots survive 2 soldering sessions but all 6 that underwent a third soldering session get destroyed? Just seems so unlikely. I didn't change the way I applied the soldering iron or anything else.
Was answered in post#2.
Solder/flux should not come near the pressed contact between the solder lug and the carbon track of the pot.
Leo..
Disconnect the power. Put your meter on ohms and measure the resistance between one end of each pot and its wiper (centre contact). The resistance should be equal to or less than the rated resistance of the pot. Turn the knob, the resistance should change in a smooth manner. If either or both these tests fail then its knackered.
Solder/flux should not come near the pressed contact between the solder lug and the carbon track of the pot.
Leo..
right, you already mentioned it. but it's exactly what I did for all the rest of the pots and they still work. anyways, this is not an attempt to disagree with your answer. I need to get new ones anyways. do you have a recommendation for a specific product? I chose the ones that I used until now because of their ridiculously low price and hope to find something for less than 1€/$ per piece - as I'm always on tight budget and my controller to me is more a proof of concept than a quality instrument.
PerryBebbington:
Disconnect the power. Put your meter on ohms and measure the resistance between one end of each pot and its wiper (centre contact). The resistance should be equal to or less than the rated resistance of the pot. Turn the knob, the resistance should change in a smooth manner. If either or both these tests fail then its knackered.
Niiiiiiice, thank you. I never thought I'd get so excited about electronics!
The big takeaway lesson here is never do so much wiring without testing. Just do one row and then test.
What you have in the two ways you wired up the power and ground to the pots is known as star wiring where each pot goes back to a common point. This has its place to isolate heavy ground currents but not here. What you have now done is known as chain wiring and is the correct thing to do under the circumstances.
But remember hardware or software, do a little and then test. That is what all the pros do.
Grumpy_Mike:
The big takeaway lesson here is never do so much wiring without testing. Just do one row and then test.
What you have in the two ways you wired up the power and ground to the pots is known as star wiring where each pot goes back to a common point. This has its place to isolate heavy ground currents but not here. What you have now done is known as chain wiring and is the correct thing to do under the circumstances.
But remember hardware or software, do a little and then test. That is what all the pros do.
Lesson learned. For my second attempt I rewired only one row and do testing now.
ReverseEMF:
A couple of questions:
Are you connecting one pot to one Arduino input [i.e. one wiper to one input?
What do you mean when you say "Meaningful Signal"? How are you arriving at this conclusion?
For testing I connect one pot directly into one Arduino analog input, yes. Wiper is the middle terminal which "sends out" the signal I'm assuming - I'm using that. With untouched pots that works fine and gives expected readings from 0 - 1023 when turning the knob.
with "meaningful signal" I meant the signal readings that the middle terminal (wiper?) produces from a single pot directly connected to the arduino input (which I output directly to the serial monitor in intervals of ~150ms). The apparently broken pots do not respond to turning the knob. I observed weird behaviour where the read value is 0 for some seconds, then ramps up to 1023 within something like 2 seconds, after some second ramps down to 0, continuing like this until I stop the assessment - without me touching the know at all! As far as I remember now, I also observed apparently broken pots that just output steady signal somewhere between 0 - 1023, say ~470, that doesn't change when turning the knob. Does this answer your question?
Thank you everyone for your highly helpful answers and suggestions, without which I would stand pretty much in the dark still. I will get a set of new pots which is going to take some time and then apply what was said in here.
Wawa:
[...]
Next time get pots with solder lugs instead of pins. These ones with pins are made to mount on a circuit board.
[...]
These have the kind of terminal shape (i.e. solder lugs) that are appropriate for this task, right? This time I want to be on the save side 8)
Those are the right ones for connecting wires to them, yes. Solder the wires to the tags.
EDIT
Do you know if you need linear or logarithmic? If it's been discussed here I missed it (I've assumed all the way through linear, but that was just my assumption).
Now it's about making it better!
