Bouncing Temperatures

[rcrabb22]

The following data log graph shows my temperature bouncing up and down like a heartbeat between 37F and 41F 10 times within 4 minutes in my kegerator. The stock BCS-460 thermowell is mounted in the lid of a 1 qt mason jar full of water. I also tried a temperature probe from Derrin at Brewers Hardware with the same results, although his probe using his coefficients measured a uniform 3 F colder, but the same temperature swing activity. The probe is ~30ft from the BCS unit and I used 24G shielded audio cable to augment the probe cable length.

Is this normal? Can the temperature of the water fluctuate that much? I don't think this type of operation is right. What am I doing wrong?
How does one create a state machine that won't actuate the SSR on/off repeatedly? I do not want to implement a delay state hack to compensate for this behavior.

A little help please!

More data (different data than graph) at 5 sec intervals
Temp Resistance
39.3 24483.3
40.3 23874.1
39.3 24457
39.1 24602.5
38.5 24937.9
36.9 25955.6
39 24629.1
39.2 24509.8
40.4 23785.2
39.4 24417.5
39.2 24523
39 24642.4
37.8 25390.7
36.7 26085.2
39 24615.8
39 24615.8
39.4 24365
40.4 23797.9
41.4 23187.9
39.2 24536.2
39.1 24562.7

[ECC]

No, I don't think thats real. I don't believe the temp could fluctuate that much in a 5min time frame. It would be useful to see the output plotted as well, since the interference is most likely caused by the compressor. If it is interference from the compressor, I would expect to see the temp reading quiet when the compressor is off, and jump around when its on. Do you see the temp settle down when the compressor is OFF? On thing to try would be to ground one end of the shielding, this will make your shielded wire more effective at blocking out electrical interferenece.

What type of control are you using for the output? I think Differential Control would be the best for a kegerator. A Differential Swing of at least 2-3deg.

To 'slow it down', adjust the "PID Sample Period" on the System settings page. This will change how often the controller reads the temperature, and how often it makes a decision to switch the output. For example, if you are controlling an output Differentially cool with a swing of 4, and a PID sample period of 20sec, the BCS will read the temperature once every 20 seconds. If the temp is greater than the temperature setpoint plus 2deg, it will turn the output ON. After another 20sec, it will read the temp again, and if the temp is less than the setpoint minus 2deg, it will turn the output OFF.. If not, it will leave it as it was (ON), and check back in other 20sec.

Hint: If you set the upper bound on your chart to 50deg you will get nice round numbers on the left hand side. There are 20 divisions divided equally between the upper and lower bounds. The datalog only stores rounded integers, so decimals aren't really needed.

[rcrabb22]

Thanks for the help!!

I relocated the BCS unit closer to the kegerator and am able to connect the probe directly to the BCS unit and temperature measurements have become more stable. I am plotting a new graph and will post when I have more data.

I am not using the BCS to control the kegerator. I was logging it's current performance as a baseline before switching to the BCS. I will try grounding the drain wire and reattach the extension cable after I log some data.

I calibrated my Brewer's Hardware temperature sensors based on real measurements as the reported temps from them did not reconcile with my Thermapen. I am relying on the Thermapen as the baseline. Once I did that the temps reported by the BH temperature sensors were more in line with the Johnson A419 currently controlling the kegerator. Before calibration the BH sensors were measuring around 2 F below the Thermapen readings at 104 F and 158F but 2 F above the Thermapen at 32.1F.

More to come.....

[rcrabb22]

Here are new graphs and thermo test chart after connecting the temp probe directly to the BCS unit (removed the shield extension cable) and adjusted the coefficients after actual temperature measurements using a Themapen as the reference temperature measurement. I will add the SSR to the mix tonight or tomorrow and use the BCS to control the kegerator. more to come.....

39.1 24589.2
39.1 24602.5
39.1 24602.5
39.7 24247.4
39.5 24338.8
39.5 24378.1
39.3 24443.8
39.2 24509.8
39.2 24523
39.2 24536.2
39.2 24536.2
39.2 24549.4
39.2 24536.2
39.2 24536.2
39.2 24536.2
39.2 24523
39.2 24509.8
39.3 24457
39.2 24523
39.2 24509.8
39.3 24457

[rcrabb22]

More data just for fun. The graph shows both the BCS thermowell in a 1 pint mason jar of water and the Brewer's Hardware temperature sensor in a white labs vial packed with gel from a freezer pack. They are located right next to each other and they track pretty close once I adjusted the coefficients of the BH sensor based on real measurements with a Thermapen digital thermometer. My plan to cut over to the BCS for kegeratore control was delayed. I am trying to find a decent project box to mount the SSR. Once I find one I'll cut over and show new data. More to come.....

[ECC]

That looks much better.