Analog problems

[Austin]

I am using 3 of the analog inputs on the T1616+.  All 3 are 4-20mA from  tranmitters, and the appropriate resistors are installed. The input to ADC(1) is quite stable.
I am using ADC(3) and ADC(4) for the other two inputs.  If I connect them individually, I can get the correct readings. However, once I connect both of them, one sinks to 0, the other explodes to 4092. It is also not consistent which drops to 0, and which explodes to 4092.

Your help is appreciated.

[support]

When a 0-5V analog input is not connected to any voltage, it "floats" and tend to be "pulled along" by the voltage in an adjacent analog pin and assume a value close to the adjacent pin. This is quite normal. As long as you have connected the pin to a stable voltage it should give a stable reading.

I don't quite understand what you mean by "explode" to 0 and 4092. It appears to me that there is some wiring connection issue here. If both ADC 3 and 4 are connected to different stable voltages then ADC(3) and ADC(4) should give value equal to a fraction of 4092 in proportion to the applied voltage.

[Austin]

I think that by verifing that ADC(3) & ADC(4) are operating correctly when the adjacent input is not connected or is shorted shows that the wiring is correct.  In addition, the wiring of ADC(3) & ADC(4) is identical to that of ADC(1), with the exception of the resistor values.

The problem occurs, if for example I have a stable voltage connected to ADC(3) that gives me a reading of, say, 1000.  The input to ADC(4) is shorted, giving a value of zero.

The short is then removed from ADC(4), and a stable voltage applied to ADC(4).  The previous stable input to ADC(3), that was reading 1000, now either drops to zero, or goes up to 4092.

If ADC(4) is now removed and shorted, ADC(3) returns to it previous stable value.

The scenario is the same when ADC(4) is used, and ADC(3) is shorted.

[support]

Did you use a meter to measure the actual voltage that is applied on input 3 & input 4 when you recorded abnormal readings? It is important to verify what is the actual voltage applied to the screw terminal to be certain if the readings that the PLC reported on its analog input matches the physical voltage values.

[Austin]

Problem solved
What we were measuring was the ORP of a solution in a tank at 2 diffeent locations in the tank.
We had switched the 2 ORP signals from ADC 3 & ADC4 to ADC1 & ADC2.  If we had the probes in 2 different solutions, the readings were bang on. But as soon as there were in the same solution is when the problem arose.
We spoke to the manufacturer of the probes and they confirmed that this is a problem, since the probes will have the same reference point.
We decided to go back to our orginal plan of using a single probe.