glassonion
Electrical
- May 7, 2008
- 3
Has anyone had issues with Auma valve actuator position feedback with a MicroLogix 1200? My issue is quite odd:
The signal is 4-20mA into a Micrologix analog input module. As usual, my code is a SCP block alone in a rung. Parameters in the SCP block were determined by running the valve all the way closed, which showed a raw data of 8268. All the way open was, as expected, 32767. This was scaled 0 - 1000 (multiplied by 0.1 at the HMI).
The results were only correct at the valve limits. In between, the scaled result was not linear. For instance, when the valve was ran to 25% open, the position read 38%. At 50%, the scaled value was over 65%. But as the valve approached 100% the error reduced to no error at 100%.
Then I experimented with the raw input values. I left the 100% value at 32767, but started changing the 0% value. What I came up with was in order to get a linear response, the closed raw value had to be 11200. Since the valve is sending 8268 at 0%, The scaled result was a negative, so I had to clamp it at 0%. Just off 0%, however, the scaled result was very close to linear.
Any ideas on that one?
The signal is 4-20mA into a Micrologix analog input module. As usual, my code is a SCP block alone in a rung. Parameters in the SCP block were determined by running the valve all the way closed, which showed a raw data of 8268. All the way open was, as expected, 32767. This was scaled 0 - 1000 (multiplied by 0.1 at the HMI).
The results were only correct at the valve limits. In between, the scaled result was not linear. For instance, when the valve was ran to 25% open, the position read 38%. At 50%, the scaled value was over 65%. But as the valve approached 100% the error reduced to no error at 100%.
Then I experimented with the raw input values. I left the 100% value at 32767, but started changing the 0% value. What I came up with was in order to get a linear response, the closed raw value had to be 11200. Since the valve is sending 8268 at 0%, The scaled result was a negative, so I had to clamp it at 0%. Just off 0%, however, the scaled result was very close to linear.
Any ideas on that one?