Part operation not matching spec?

[Topher925]

I've run into a dilly of a pickle using the MAX5494 Digital Potentiometer from Maxim. Its a simple two pot per chip IC which has a rated full resistance value of 10k with 1024 steps, or so the datasheet says. The relationship between wiper to ground resistance (Rwl) is suppose to follow this formula,

Rwl = D/1023*Rtot + Rz

where D is the digital value being sent, Rtot is the total resistance (10k), and Rz is an error offset. So I have this thing integrated into a project for work and much to my surprise it doesn't seem to work! The pot only seems to be linear in the first 1k to 4k region and after that its resistance goes all over the place. For example, sending it a value of 512 should give a resistance of around 5.5k but its actually about 6.5k. A value of 750 gives a resistance of 16k. A value of 900 gives 11k (way over 10k!), but a value of 1023 gives 7.5k.

I thought it might be code being screwy but I've scoped the SPI bus and double counted the bits and my code is solid. I just can't figure out what the problem is and I'm suppose to be testing this thing tomorrow. Any idea what it could be? Do you think the manufacturer just may have screwed up a batch of chips (I've got several, they all do it but with random values).

 

[jim hardy]

5494? Is that the programmable voltage divider IC?

You said nothing about how you use it in your curcuit.

Did you check to see that you don't run afoul of this precaution in the datasheet ?

Note that the programmable
voltage-divider is not intended to be used
as a variable resistor. Wiper current creates a nonlinear
voltage drop in series with the wiper. To ensure temperature
drift remains within specifications, do not pull current
through the voltage-divider wiper. Connect the wiper to a
high-impedance node.

datasheet page 12

The selection guide you linked distinguishes between voltage divider and variable resistor IC's

The MAX5494/MAX5495 are dual, 3-terminal, programmable voltage-dividers;
the MAX5496/MAX5497 are dual, 2-terminal variable resistors;
and the MAX5498/MAX5499 include one 2-terminal variable resistor and one 3-terminal programmable voltage-divider.

From your description it sounds like you're using the 5494 voltage divider IC as if it were a 5496 variable resistor IC.

just a guess.

old jim

 

[Topher925]

5494? Is that the programmable voltage divider IC?

You said nothing about how you use it in your curcuit.

Did you check to see that you don't run afoul of this precaution in the datasheet ?

datasheet page 12

The selection guide you linked distinguishes between voltage divider and variable resistor IC's
From your description it sounds like you're using the 5494 voltage divider IC as if it were a 5496 variable resistor IC.

just a guess.

old jim

Jim, thanks for the post. Yeah, I realized that just about 2 minutes after making the thread and yes that is the problem (facepalm!). I was in a rush when designing this particular board and just skimmed over the first few pages of the datasheet and thought it was what I wanted. Guess I should read the whole datasheet next time.

Fortunately, I can still make it work, its just going to take a hit in resolution.

 

[jim hardy]

Glad you got it !

Facepalm - what a great phrase.

I say "These lessons put knots on my head, and i have plenty of really big ones."

 

[alphysicist]

Based on the information provided, it seems like there may be an issue with the digital potentiometer itself. It is possible that there was a manufacturing error or a batch of faulty chips. Another possibility is that the potentiometer is not compatible with the specific project or setup it is being used in. I would recommend reaching out to the manufacturer for assistance and possibly testing the potentiometer in a different project or setup to see if the issue persists. It is also important to double check the connections and ensure that the potentiometer is being used correctly according to the datasheet.