Digital Potentiometer not fading LED

Question

So I am trying to test what will eventually be a Arduino controlling 3 110V motors that are currently controlled by an analog pot. But more on that later.

I am trying to test out using a digital pot (never have before) and am running into some issues with tutorials I have found online. For now, I want to control a LED with the digital pot (Spec Sheet) so that the LED fades in and fades out in a continuos loop.

Here is the tutorial I followed. It uses a different chip model but it seems from the pot I am using that the pins are the same. (I could be wrong?)

I double and triple checked the wiring and it matches below.

Wiring Diagram

Here is the code I am running (slightly modified from the tutorial, also tried the tutorial code)

#include <SPI.h>
byte address = 0x11;
int CS= 10;
int i=0;
void setup()
{
 pinMode (CS, OUTPUT);
 SPI.begin();
}
void loop()
{
 for (i = 0; i <= 128; i++)
 {
 digitalPotWrite(i);
 delay(10);
 }
 delay(500);
 for (i = 128; i >= 0; i--)
 {
 digitalPotWrite(i);
 delay(10);
 }
}
int digitalPotWrite(int value)
{
 digitalWrite(CS, LOW);
 SPI.transfer(address);
 SPI.transfer(value);
 digitalWrite(CS, HIGH);
}

As a result of the two above, I have an LED that stays lit (at about 50% it seems) and does not fade in/out at all.

Being relatively new to wiring and EE in general, is there a better way to replicate an analog pot without using a digital pot? Can I just use PWM to control it? Why would the LED be just staying lit?

Thanks!

• jfpoilpret: The spec sheet you provide is very general and covers very different models of digital pots. What is the exact reference of yours?
• cgaubuchon: Sorry @jfpoilpret, I am using the MCP41X1 reference as the chip is a MCP4131 via Amazon
• David Cary: Yes, you can just use Arduino PWM to control an LED. See arduino.cc/en/Tutorial/Fade .

Answer 1

After studying the datasheet differences between MCP41100 (digital pot used in the tutorial) and MCP4131 (the one you get), I found two major differences between both:

1. Difference in resolution:

   • MCP41100 has 257 steps resolution (8 bits + 1): you can see that at the beginning of its datasheet
   • MCP4131 has only 129 steps resolution (7 bits + 1): you can get this information on page 2 of its datasheet

2. Difference in commands formats sent through SPI; both use one "command byte" followed by a "data byte", but have a diffrent format for their command byte:

   • MCP41100 command byte is X X C1 C0 X X P1 P0 where X is unused, C1 C0 is one of 4 possible commands, only 2 being valid commands actually: 01 (Write data), 10 (Shutdown), finally P1 P0 should be always 01 for an MCP41100
   • MCP4131 command byte is AD3 AD2 AD1 AD0 C1 C0 D9 D8, where AD3-AD0 is the address of the "register" on which to perfom a command. For the MCP4131, the most important address is 00 which means "volatile wiper 0" (this is what allows you to change the resistance value of the pot), C1 C0 is one of 4 possible commands (read data, write data, increment, decrement), in your example the command should be 00 (write data); finally D9 D8 are the high order bits of the value to write to the register, but D9 is unused and should always be 0. For 257-steps pots, D8is used to have the full resistance scale, otherwise it should be 0

Now that we know aht your MCP4131 expects as commands passed through SPI, we can take a look at what the tutorial program passes:

00010001 value

The first byte can be broken down as such:

• 0001 address of the pot wiper register 1 (not 0): this wiper does not exist on MCP4131 hecne writing to it will have no impact...
• 00 command bits meaning "Write to register specified in previous address"
• 01 high order bits (D9-D8) of the value to write to the wiper register.

Although I don't have an MCP4131 at hand, I think you should adapt the program as follows:

byte address = 0x00;

And that should be fine! Indeed, now your program passes:

00000000 value

Which means "write value (between 0 and 128 which matches exactly the 129 steps resolution) to the wiper register 0 of the port".

Note that the MCP4131 has more features than the MCP41100 in terms of commands, for instance you could use the special commands "increment" and "decrement", with the advantage they need only one byte (the command byte), e.g.

void loop()
{
 for (i = 0; i <= 128; i++)
 {
 digitalPotIncrement();
 delay(10);
 }
 delay(500);
 for (i = 128; i >= 0; i--)
 {
 digitalPotDecrement();
 delay(10);
 }
}
void digitalPotIncrement()
{
 digitalWrite(CS, LOW);
 SPI.transfer(0x04);
 digitalWrite(CS, HIGH);
}
void digitalPotDecrement()
{
 digitalWrite(CS, LOW);
 SPI.transfer(0x08);
 digitalWrite(CS, HIGH);
}