Pot interaction

W

waltsar

New Member
I have two 10 K pots wired in parallel, CW to CW and CCW to CCW across a 4.5VDC P/S. The wipers go to b.4 and b.5 (configured as ADC input pins) on a Picaxe 14M2 controller chip. I find that the pots are extremely interactive. When I change either pot up or down, it causes the function controlled by the other pot to change as well as its own, and vice versa. I have used this configuration in the past and never noticed this interaction - that doesn't mean that it wasn't happening, I just have never seen this.Picaxe Pots - 3.jpg
 
Hi,

No, the circuit you've shown shouldn't be interacting like that. The cause is very probably that there is a low impedance (e.g. a short-circuit) from one of the wipers (or the connected PICaxe pin) to ground or to another component. Probably a hardware issue such as a solder bridge between pins at the PICaxe or elsewhere; It could be a software issue, but you haven't given us anything to check.

The best way to determine if it's hardware or software is to try a simple test program and rotate each pot in turn (which should swing uniformly between values from 0 to 255), for example:
Code: 
#picaxe 14m2
#no_data
#terminal 4800
do
   readadc c.4 , b1
   readadc c.5 , b2
   sertxd (cr,lf,"c.4= ",#b1,"  c.5= ",#b2)
   pause 1000
loop

Note that the supply rail is also used as the ADC reference voltage, so if the supply rail is varying (for example due to loading when a pot is rotated), then the voltages won't change in the "expected" way.

Cheers, Alan.
 
Wildly interacting would suggest a software problem, mildly interacting suggests power loading.

As Alan says, try his simple code to check the hardware, and post your code for review.
 
Thank you, gentlemen. Do you know what the input resistance/impedance is of an ADC input? I would expect a DC resistance of at least 1 meg, which would make this circuit virtually interaction-free.
 
Hi,

The ADC input is primarily a capacitive "sample and hold" with the maximum recommended source impedance to be 10 kohms, e.g. a 40k pot at mid-track (i.e two 20k in parallel = 10k). Higher resistances are also acceptable if a capacitor is connected across the ADC input to ground, so indeed we wouldn't expect the input impedance to be the cause of the issue. But if the software happens to (temporarily) configure the ADC input as a Digital Output, then the impedance could be only tens of ohms to ground.

Cheers, Alan.
 
I should have mentioned that the 14M2 chip is mounted on an AXE117 project board. On the board, each input pin has a 10K resistor to common, so the effective input DC resistance is 20K across the 2 pot source - would that cause the large interaction? Also, I noted a low resistance to ground on the wipers of the pots. I completely isolated the board from all other circuitry, including the power supply, and found 106 ohms from the 4.5V pin to 0V. I checked a known good board, and the same pins have a 40K resistance, closer to what I would expect. Can I assume that the board is compromised enough that I should get a new chip? And, by the way, where is a good USA source for Picaxe hardware?

edit: Looking at the wiring diagram of the AXE117 board, pins b.4 and b.5 are hardwired to the input of a ULN 2003 darlington buffer chip. The input to each channel has a 13K to ground DC path. Again, will this low impedance cause the interaction of the two 10K pots feeding pins b.4 and 5?
 
Last edited:
As others have stated it is most likely a soldering issue, take a picture with your phone of the soldering and zoom in (or attach here).

It's an unusual pin choice to select B.4 and B.5 on this board because of the darlington driver - the C.x pins would be more logical. We assume you are using the row of holes between the two chips on the AXE117 for the B.4/B.5 connections (not the buffered outputs on the right side). What happens if you remove the ULN2003 from the socket? Also the AXE117 PCB has the option for an optional 10k/22k for a 3V safe output on output 5 (B.5) - are these extra resistors fitted?
 
Thanks for all the input guys, it's greatly appreciated! I'm going to take a couple of days off from this project out of sheer frustration. I worked for many years as a systems engineer, but never as a circuit designer and/or programmer - it's a lot easier cook-booking a system together using beta-tested components and sub-systems.
 
Hi,

Resistors between the wipers and ground may upset the linearity of the potentiometers' output, but will not introduce crosstalk. Most integrated circuits have "(electrostatic) protection" diodes connected internally between nearly all I/O pins and the supply and ground, so multimeter resistance measurements can be very misleading, with or without a power supply connected. The only reliable (external component) test is to remove ALL the ICs, and obtain consistent measurement values with both connection polarities of the probes, and preferably for more than one Ohms (scale) range. Failure of the PICaxe chip itself is very rare.

But one of the strengths of the PICaxe is its integrated debugging environment, so you can download and run a test program that uses only the programming interface, in literally a minute or two. Then post the program and its SERTXD output to the forum, if you can't interpret the results (obviously with a comment on how the pot(s) were set and/or changed, for the test). Just be familiar with the "Hard Reset" procedure, if the PICaxe has been previously programmed, particularly with a (now) "unknown" program.

For the USA, I believe Peter Anderson (RIP) was a well-regarded supplier, but his son may have now ceased trading. For most parts of the World, the (UK) PICaxe Store can be a reliable fall-back, but high shipping fees and unpredictable Tariffs may have become an issue for the USA. :(

Cheers, Alan.
 
The easiest and most accurate;
With a DMM, monitor one of the pot’s wiper voltage while you adjust the other one.
Swap the pot being monitored and adjust the other.
Report your findings.

As it has been mentioned previously, they should not interact.
 
Alan; it is correct. Professor Anderson passed away a while back. His son continued with the Picaxe store for a while, but has since liquidated it.

My last Picaxe order from an US-based retailer was from RobotShop, has the most usual Picaxe devices.
 
Are those pots really connected as in the picture? If not they might just devide the 4.5 Volt between them, which would explain the interaction. Just a thought...
 
waltsar,

If you post a picture of both the front and back of the board it will enable forum members to do a visual inspection.
 
You must log in or register to reply here.
Back
Top