Midi Controller - 32 mixed type/value pots - Strange readings

I am making a midi controller with 32 turn knobs/potentiometers. In the final circuit, I will read the 32 pots using 2x 74HC4067 boards (analog multiplexers with each 16 analog in/outs) connected to an Arduino Nano.

In the current circuit I have connected all 'ground' pins of the 32 pots to a common ground wire, and all the '5v' pins to a common 5V wire. The circuit with only 3 of the 32 potentiometers is:

When I read the various wiper pins (not connected to 5V or GND) I get strange results. For the first 75% of the knob turn, the resistance at each wiper pin increased, and turning further it decreases.

I have mixed potentiometers of different types (linear and logarithmic) and of different values (50k, 100k and 250k). Does this cause this strange behaviour?

Also I don't want to draw too much current for the 74HC4067, if I read specs correctly this should be max 50mA. Are pots > 5k safe? Is 250k too much?

5kΩ is fine, 10kΩ is ideal, 50kΩ is a bit much, but should work fine in most cases, everything above that is too high, and you'll get crosstalk between channels, because it takes too long to charge the sample-and-hold capacitor of the ADC.

Are you 100% sure that you connected them correctly? Did you try it with just a single potentiometer, without the mux, directly connected to the analog pin of the Nano?

There should be no DC current flowing through the 74HC4067, the current flows from 5V to ground through the potentiometers.

It's not entirely clear to me what you are showing on the plot. Did you measure the resistance, or did you measure the voltage on the analog inputs? Do you have a multimeter to check the resistance of the pots? My guess is that the pots are fine, but that there's a wiring problem somewhere.

To test the pots and the mux, you could use this 74HC4067 AnalogReadSerial sketch: https://tttapa.github.io/Control-Surface/Doc/Doxygen/dc/d1b/1_8AnalogReadSerial_8ino-example.html You can then look at it in the Serial Plotter (Ctrl+Shift+L).

Pieter

I assume you do not realize you cannot put pots with such low values across the 5V and GND ?

Have you heard of Ohm's Law ?

Matlab

Rp=1/((1/50)+(1/100)+(1/250))

I have mixed potentiometers of different types (linear and logarithmic) and of different values (50k, 100k and 250k). Does this cause this strange behaviour?

I see 50 ohms, 100 ohms, and 250 ohms. Are these the real values of the pots ?

Rp =

29.4118

5/Rp

ans =

0.1700

Rp = parallel resistance of the 3 pots you included in your post. 5/Rp = 5V/Rp ans = the current in A drawn by the three pots from the 5V regulator.

If these are only 3 of 32 pots, and ALL of these values are too low, WHAT are the values of the OTHER pots ?

Did you remove the pots in question and measure them ? Did you measure the voltage on the wipers with a DMM (I'm sorry , but I have no confidence in your circuit or your code yet. I do trust a multimeter.)

I think (hope) that he actually means 50*k*Ω:

NardJ: I have mixed potentiometers of different types (linear and logarithmic) and of different values (50k, 100k and 250k).

thx for your replies! I did forget to put the 'k' before the 'Ω'. The pots range from 10kΩ to 250kΩ.

The circuit is not connected to 5V/GND and also not to an Arduino or the HC chip. (It is an isolated circuit consisting only of potentiometers and wires.) I am measuring resistance with a multimeter between wiper/middle pin and one of the outer pins of the potentiometer.

I now reduced the circuit to 4x B10K resistors (and some wires) |500x317

When measuring one potentiometer in this circuit between one of the wiper pins and an outer pin, the resistance slowly rises from 0k to 3.2k at 80% turn and then falls to 2.4k when fully turned. (Turning other pots does not affect the reading).

Other multimeter gives same readings. When measuring a single pot not connected to other pots, it reads from 0 at 0% to 10k at 100%. I also double checked the connections to make sure they indeed are as in de circuit diagram above.

Think this rules out wrong connections, low quality potentiometers and bad multimeter.

Any ideas?

Do all the pots behave the same or just that one you tested ?

All pots behave correctly if not connected to the other pots. If I build the circuit in previous reply, all 4 have a max resistance of 3.2KOhm midway and above expected 2.5KOhm.

Matlab

Rp=1/(1/10000+1/10000+1/10000+1/10000)

Rp =

2500

What this says is that four 10k pots is a total parallel resistance of 2.5k ohms. This should be obvious since if you parallel TWO 10k resistors you get 5k, so if you then parallel TWO 5k resistors , would you not expect to get 2.5k ?

As far as I can tell Ohm's Law still applies here. The value of the pot is only correct if you have not put it in parallel with another pot because the reference for the wiper is the total resistance across the pot which changes every time you put add another pot in parallel.

"When you have eliminated all possibilities but, whatever remain, however illogical , must therefore be the truth." Sherlock Holmes

The ohmmeter is a current source that measures the voltage across the test leads. Look at how the current divides and flows through a given pot (upwards and downwards) and then through all the other pots.

"Back to the drawing board..."

Pots connected in parallel like that do well as voltage sources. The confusion happens when one does not understand the resistance readings.

Pots connected in parallel like that do well as voltage sources. The confusion happens when one does not understand the resistance readings.

When I read the various wiper pins (not connected to 5V or GND) I get strange results. For the first 75% of the knob turn, the resistance at each wiper pin increased, and turning further it decreases.

How (EXACTLY) did you obtain the plot for this ?

If you write out Ohm’s law in the circuit, you get something along the lines of


Where x is the potentiometer position, R is the resistance, and n is the number of pots.

If you plot this for linear and logarithmic pots, you get something like this


log-potentiometer-parallel.png

lin-potentiometer-parallel.png