over voltage on 10 volt analog input

[russ]

I was setting up an input to an IO-A14-A02 expansion module with a V290 and the input was over voltaged. The actual amount of voltage is not known but may have been as much as 20 VDC. The system ran for awhile in this configuration and then began giving problems. In particular the power up values within the linearization that this input was being fed to would take on random negative numbers. The voltage input has been corrected to be within the 10 VDC level but the problem with the linearization continues. Has this problem been seen before due to an over voltage? Is something was damaged as a result would it just be the expansion module?

Thanks,

Russ

[Joe Tauser]

Russ,

Yes, I have blown up analog inputs by directly applying 24V, but this is usually in the 4-20 mA mode and it fries the little resistor inside. The voltage input impedance is 1 Meg and the maximum rating is 20V.

You have to start at the beginning by looking at the raw values returned by the module. If you apply 0-10V to the module, you should get 0-4095 in the mapped MI register. Disconnect all the inputs and check them one at a time to make sure you're not getting a ground loop anywhere.

Joe T.

[russ]

Thanks Joe,

I will take a look at the input values. When I was watching them they looked good and stable and below the 4095 max, however I didn't note if they were following the 0-4095 with 0-10V. Would this cause the other MI's in the linearization to randomly go negative.

Russ

[Walkerok]

If we go away from an asumption that the input chanel has failed or been dmaged. It is possible for the linearization to create negative numbers that would show up on screen periodically. If you are doing the calculations using only MI locations to store the information on the linearization. If I remember correctly the prebuilt linearization function has values that it will not show correctly becuase of overflow when the calculations are made exceeding the length of a standard 16 bit MI. If you have not done so all ready, try making your own linearization calculations in the ladder logic. If you have a number that exceeeds 16,535 durring the calculation process it will run over and may or may not show a negative value (depending on the exact size of the number). If any number is too large to fit in an MI location use the appropriate math blocks and store and manipulate at least a part of the linearization calculation in a 32 bit structure using ML locations to prevent the overflow that could be happening in a 16 bit MI location set of calculations.

Of course if the analog input chanel is damaged then this won't help but if, like Joe said, you get 0-4095 raw values when inputing a 0-10VDC signal then the linearization calculation is more likely cause for your periodic negative values you are seeing on screen.

Keith

[Joe Tauser]

Keith-

You point is valid on overflowing the MI. It is with great humility that I must point out that the maximum value for a 16 bit register is 2^15 - 1, or 32,767. If the linear block spits out value of 32,768 or more the sign bit kicks in and all kinds of goofy negative numbers show up.

Joe T.

[Walkerok]

MIssed it by that much :-) Thanks Joe, I apparently failed my 2 bit math class. Pun Intended

Keith

[russ]

Kieth and Joe,

Thanks for all the input. I've been busy running in 5 directions at once but will check into all this soon. I'll let you know what I find.

Russ