Temp RH Discrepancy

You're asking us to troubleshoot an unknown system from a remote location....but you did say the temp and humidity are independently controlled....

Someone opened the chamber door, got a text message, had to reply immediately and left the door open all the while.

That would be perfectly in sync with today's world. The all-important cell phone takes pre-eminence over everything else.

With that big of temperature change and no measurable humidity change, I would suspect measurement error. Maybe the temperature logger had a glitch. With temperature changing and a constant amount of moisture in the air the humidity would certainly have changed. Or maybe the temperature changed faster than the humidity sensor could respond.

that would be my guess, as the humidity should attempt to go with everything else constant (semi-closed system), and temperature falling.

How quickly did the temperature drop?

Is this rate of change possible without opening the door?

Try replicating the situation by cutting power to the heaters, and, monitoring temperature and humidity.

If, as you said, "They r controlled independently" why would you expect the RH to change?

They [a] r [e] controlled independently.

Why the temperature changed is your problem to solve.

What is the ambient temp and humidity just outside you climatic chamber?

Knowing this could give you strong indication of instrument error or open door causes.

There is a pretty narrow range of RH and temp combinatioms external to your climatic chamber for which an open door could result in your data without instrument error.

As an example, if the ambient relative humidity external to your climatic chamber was >=80% and the temp was 28 C, it would be very difficult to get the readings you describe without a measurement error.

.

The previous assumes RH and temp are measured at close to the same location. With distance between the measurements the readings could occur from a small openning in the chamber near the temperature measurement but remote from the RH measurement without measurement error.

.

Can you explain how you are controlling RH in more detail?

You are asking the same questions as me.

some pointers though

1) no one opened the chamber

2) The sensor was working fine when checked last, its a normal PT-100 and has been calibrated.

3) there is a very remote chance that a water droplet/condensation may have fallen on the sensor.

Question:

Is point three a valid argument?

I am attaching the zoomed pic. Please note that the complete protocol takes 10 days and this was the only glitch on 5th day, before things became smooth again.

Yes, as the chance is not identically equal to zero.

If the water shorted out the leads on the sensor either at one end of the other of cabling, then yes, it would lead to large negative systematic error, problem solved.

When a temp sensor is wet, it is a 'wet bulb', not a dry bulb; half of a wet bulb/dry bulb RH sensor.

A dry bulb that becomes wetted continues to report temperature, but it is wet bulb temperature, not dry bulb temperature.

Good answer.

The humidity control does seem problematic.

The chamber is probably poorly insulated, at least in a location such that subcooled condensate could build up then drop or pool onto the sensor. Given the step change in temperature followed by the gradual recovery, if the chamber was indeed not opened, movement of a mass of subcooled condensate onto the sensor, as an explanation is at least consistent with the changes.

Sensor, or sensor connections? IF connections, then he has the identical problem observed, most likely.

I tend to agree with Jpfalt on the connections. Seems improbably that a connection fault would gradually recover.

I am open to knowledge upgrades. Have you experienced connection faults that have gradually recovered?

Not so much, but I seem to recall at least once of having a Pt100 drop out due to wet connection, enough the indicated temperature dropped all the way to default value of 25 °C (more like sensor disconnect than a short).

I would expect sensor input shorted would show something way out of sync with reality, much lower than ambient.

Thanks for the reply. I was considering the drop as wet bulb temperature and not dew point. The RH sensor is rotronic capacitive type. RH is generated only by steam and a PLC controlled solenoid valve.

My question is, shouldn't the RH have shown a sharp drop once the temp started increasing?

I checked my PT-100 after the test, it was perfectly fine and giving correct resistance values.

The abnormality has created a lot of trouble in subsequent activities. Hence the persistence.

The system RH response should be thus: decrease--> time constant for valve to open --> transfer function of steam into chamber --> response time of capacitive RH sensor instrument --> time constant for valve to close. If that cycle time is faster than the system changes in RH, then all is good, and you have a "tight" control loop. If not, then RH can "be all over the place".

Temperature sensor drop out with RH steady probably means that your system did not change all that much, but that the indicated values were misleading.