Automation - Round 2

FNQTrade

I run a solenoid automation system and have had amazing success in dialing in my PID's and keeping my temps within 1 degree of the setpoint. Unfortunately the clicking of the solenoids drives me nuts so I set about down the variable controlled path around a year ago. Despite numerous attempts, I really never achieved the tight temps that I have with the solenoid valve so I reverted to using my existing setup.

Once again the solenoids are driving me crazy so I want to take another crack at getting the whole thing to work. Here is my setup,

1. Municipal water set 5L per minute.
2. REX c700 PID
3. Johnson Controls VG1245AE+906GGA 1/2" two way ball valve with actuator
4. Auberins 3wire RTD probe

My thoughts on Potential reasons for problem

1. PID is rubbish (Poor algorithm, poor accuracy etc)
2. Incorrectly sized valve.
3. Poor PID settings ( I've tried a lot but cant rule it out, autotune is useless, have tried Cothermans recommendations)
4. RTD is in 5" RC 1/3 from top. Perhaps the turbulence in the RC is making the temps fluctuate. Would moving this to exit help?
5. RTD is noisy.

A few questions for those of you running automation.

1. What hardware do you run?
2. What position is your temp probe in?
3. Are your temps stable, how many degrees will it go from your setpoint?

I apologise for all of the points, just wanted to make everything as clear as possible so I can start tackling this issue. Temp does seem to jump a little up and down which I feel could be the rubbish PID or the noisy RTD. I appreciate any thoughts you guys might have.

crozdog

sounds like they are on on type units (sorry I haven't looked the specs up). have you thought about going proportional and using say 4-20mA for the control signal?

FNQTrade

Hi @crozdog, its a proportional valve setup and uses the 4-20mA control signal. I was wondering if there was an issue with my components/expectations/PID settings. I was hoping to compare them to others who have a similar system operating.

grim

@FNQTrade said: 1. PID is rubbish (Poor algorithm, poor accuracy etc)

Highly, highly, unlikely.

What do you mean by the RTD is noisy? If you shake your cables, sensors, etc - do you see the temperature readings change? If so, this is going to give you lots of headaches as it will confuse the PID. Is this the same setup you used with the solenoids (just different PID configuration)?

The thing with solenoids is they are way more forgiving. BUT, the 1245AE should be just fine in terms of sizing for municipal water on a smaller still. In normal operation, what's the valve position?

Also, are you seeing oscillation?

FNQTrade

I did a quick shake test and the RTD temp is not affected by movement. Temp does seem to go and down within a 1-2 degrees regularly, which I thought perhaps there might be some electrical interference in my control box that would create this. Could this possibly be convection currents/turbulence? Not understanding a great deal about electronics I am not sure how liable to interference these 3 wire RTD's are.

Its the same brand of PID controller (c700) but different sub model(for the prop control). I start the valve in full open position and work from that point.

grim

Nah, simplest answer is more likely, PID settings and valve sizing.

You haven't said the actual behavior though - Are you overshooting? Slow to respond? Oscillating? Random waver/occasionally upsets? Never hits the set point? Is "Jump up and down a little" mean oscillation around the set point?

Assuming that at 5 liters per minute, your still will operate in 100% reflux? Have you tried dialing it back any?

There is one smaller valve, 1245AD, if during normal operation you aren't really seeing much valve stroke.

Unsensibel

From my experiments, I think it's valve sizing. I replaced to a valve with the smallest disc (1/2") and ran in manual control. After stabilizing temps - I think it was around 60C - even the smallest change in output was causing a craving of the temp.
My take, either smaller valve 1/4", pressure reducer in the supply or pumping from a reservoir (which is the same as pressure reducing the supply).

CothermanDistilling

are you using city water(60-70psi) or recirculating(0-15psi)????

Unsensibel

At the moment, all my experiments were using city water. I have a bucket rigged with a pond pump and a float valve to maintain constant water level.

Time permitting (and my honey-do list is getting long), there's three more experiments that I see:

1. Use pond head pressure to feed into valve to control water supply to RC. Since I'm measuring in the RC I've also seen temperature variations during the run. I'll see if they'll effect tune-ability.
2. Use pond pump to recirculate between bucket and RC and allow valve to bleed hot water from system. The float would then re-fill the bucket/reservoir with fresh/cold water lowering the overall temp.
3. Use pond pump to recirculate betwee bucket and RC and allow valve to inject fresh/cold water into the loop. Bucket/reservoir would then overflow.

razerhawg

Do you have your valve on the inlet or outlet of your condenser? Probably need it on the outlet.

Mickiboi

@razerhawg said: Do you have your valve on the inlet or outlet of your condenser? Probably need it on the outlet.

+1

Unsensibel

I can plumb it in either end