How can you keep Chest freezer on or off longer?

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

tylo_k2008

Well-Known Member
Joined
May 30, 2012
Messages
82
Reaction score
1
Location
Phoenix
Hey. So I have a chest freezer. It is the magic chef 5.5cf. And I have a love thermostat, which is hooked up to an outlet (Keezer plugged into outlet).
At first my chest reezer was on for about 5 minutes and off for about 14. Then I insulated it and taped it all off now it stays on for about 7 minutes and off for about 24.

What I am wondering is- what can I do to keep the chest freezer on longer, or off longer? I heard with the chest freezer keep turning on and off that it can drastically short the life of the freezer. If there was a way where I could just keep the freezer on all the time at 37 degrees that would be great! But unfortunately I don't think thats an option. Any tips or suggestions I can try to keep my chest freezer lasting as long as possible. I would definately hate to have to start all over with a new freezer, rebuilding the collar and everything. Thanks!!! :rockin:
 
If using a PID, you can set the on/off temps further apart an allow the fridge to cyle less. If using an STC-1000 or similar heater/cooler from Ebay and others, give us a model number to help decipher your issue. Hope this helps you get set-up and brewing.
Wheelchair Bob
 
But he has a LOVE controller. How do you get that to have a larger differential in temperature than the standard +/- 2*?
 
There is no possible way your beer temp changes 4 degrees in 7 minutes. You need to get that probe measuring the beer temp not the air temp. I tape mine to the fermenter and then cover with several layers of bubble wrap.
 
I have the love ts2-010 love thermostat. My set point is set to 37 degrees and the thermostat ranges anywhere from 35-38 degrees. Also I do not have a fermenter. I currently have 1 - 1/6th keg and 2 cornie kegs. The thermostat probe hangs in the middle of the chest freezer, roughly 8 inches or so from the bottom. Hope this helps you help me. :)
 
"The thermostat probe hangs in the middle of the chest freezer, roughly 8 inches or so from the bottom."

You are just measuring air temp and that is why it is cycling so fast. Tape the probe to one of the kegs and the tape a covering over it. I like bubble wrap, but any insulator will do.
 
If using a PID, you can set the on/off temps further apart an allow the fridge to cyle less. If using an STC-1000 or similar heater/cooler from Ebay and others, give us a model number to help decipher your issue. Hope this helps you get set-up and brewing.
Wheelchair Bob

If you are using a PID, you should be able to adjust the Interval values (the I in PID).

If you are using a cheap chinese digital controller - which is a PI controller (Program and Interval, no Derivative) - you can still set the Interval values.

If you have, for example, the extremely cheap Willhi WH-7016c, look at the Hysteresis and Delay functions in the manual:

http://www.willhi.com/Documents/WH7016Cdatasheet.pdf

These are standard functions in digital thermal controllers. I haven't looked at the Love manual, but if it lacks this sort of function, it's worthless.
 
Ok. Thanks Billl for the suggestion. I bubble wrapped it and taped it to the keg. I am not sure how thick the bubble wrap should be around the probe. Anyways it went from being on for 7 minutes and off for 24 minutes to now being on for 11 minutes and off for 26 1/2 minutes. So the cycle moved a little bit. Ive heard people that have their cycles up to 45 minutes. I am not sure if thats true of not - But I know I want to get the longest cycle possible. And for those who keep question my Love Thermostat it is the Dwyer Ts2-010 Love Thermostat which I bought for $50 right off the Dwyer website. Any other suggestions or ways I can get the cycle to be a little longer?
 
Put your temperature probe in a glass / bottle of water at the middle of your keezer. I have a little "bump" in mine where the compressor motor is. I put a pint glass of water there and the probe in the water. This will give you a close representation of beer temperature and not just the air temperature. If you don't have a fan to circulate the air from bottom to top you may want to get one setup as well. Cold air will sink to the bottom so your kegs may be colder at the bottom than the top.
 
Put your temperature probe in a glass / bottle of water at the middle of your keezer. I have a little "bump" in mine where the compressor motor is. I put a pint glass of water there and the probe in the water. This will give you a close representation of beer temperature and not just the air temperature. If you don't have a fan to circulate the air from bottom to top you may want to get one setup as well. Cold air will sink to the bottom so your kegs may be colder at the bottom than the top.

I use a pill bottle that I drilled a hole through, and after I stuck the probe in I sealed it on both sides of the lid with silicone caulk. It too has water in it, but it's sealed to keep it from spilling, and I can also move it around if I want to check the temp in different areas of the keezer.

Here's a pic of it-

526ca5ac.jpg


It's sitting on the bucket in the back. And it too normally hangs about halfway down.

You can also see I have some L brackets mounting a small computer fan in there as well.

I've got to say, mine's staying pretty steady at 4 degrees c, and it only seems to be going on once an hour...I'm trying to time it, but I cant seem to catch when it's on....
 
Revvy,
Did you post your building process? Or get that plan from a DIY site? (-or buy it as is?)
Sweet looking setup.
 
you will also find that the fuller the kegerator the less it will cycle. All the cold kegs have considerable mass and helps to keep the temps from changing rapidly.
 
Put your temperature probe in a glass / bottle of water at the middle of your keezer. I have a little "bump" in mine where the compressor motor is. I put a pint glass of water there and the probe in the water. This will give you a close representation of beer temperature and not just the air temperature. If you don't have a fan to circulate the air from bottom to top you may want to get one setup as well. Cold air will sink to the bottom so your kegs may be colder at the bottom than the top.

I don't mean to be a jerk - i might sound like a jerk lately because something has got me in a bad mood so if i come off as rude, I'm sorry.

But I'm dumbfounded that people will invest in high-tech industrial process controllers and then use stone age methods to manipulate them.

If you want a Love TS2 series controller to stay off longer, increase the value of the c0 parameter (minumum time for compressor to be OFF).

If you want it to run for a longer length of time when it switches on - another aspect of managing the duty cycle of a compressor - increase the value of the c1 parameter (continuous cycle time).

This is rocket science, and we can do it. We live in a post-lunar-landing era, and we don't need a jar of water to make our thermostat behave in a more sluggish way.
 
Revvy,
Did you post your building process? Or get that plan from a DIY site? (-or buy it as is?)
Sweet looking setup.

Nah, I haven't posted my process, it was really no different than any other one on here. I just took the best ideas from all the 9 million other threads on here and went with the ideas that felt right.

13cf40df.jpg


The guts, tap shanks, secondary regulator, and electrical- the temp controller, the temp probe (sitting in water inside the pill bottle on the fermenter bucket) and the power outlet in back.

526ca5ac.jpg


The rear, where the freezer and fan plug into.
34635be0.jpg


The hinges for the lid.
af37551c.jpg


The fan to circulate air.
440dbd0e.jpg
 
Thanks. I'm proud of it. I totally trashed my condo building it since I don't have a garage or work space. And I still haven't cleaned the place. But I think it's pretty cool.

:mug:
 
And i concur that revvy's build is very keen. Except for the pill bottle full of water. I know it works, I just also know that the thermal controller has the ability to obviate it.

Please excuse my pedantry, I am a quality assurance manager. I think i should go relax and have a home brew.
 
And i concur that revvy's build is very keen. Except for the pill bottle full of water. I know it works, I just also know that the thermal controller has the ability to obviate it.

Please excuse my pedantry, I am a quality assurance manager. I think i should go relax and have a home brew.

Well, countless people on here, many who probably know just as much if not more than you (and definitely a LOT more about this than me) have made a great case for submerging the temp probe, not to as you say trick it but in order to get a more accurate temp reading since, and this makes sense to me, that we are interested in controlling the temps of a FLUID and not the surrounding air....

One of the things about this hobby is that usually there are more than one way to skin that brewing cat. And just because we may not agree with why something is done, more than likely other "experts" see things differently. Like we say on here quite often, "ask 10 different brewers the same question, and you'll get 12 different answers, and they will all be correct.

I also bet that there aren't a few "quality assurance managers" around who do the same type of job as you, but completely differently. And probably if they followed you around on the job for a day, looking at what how you'd do things, they'd probably need to have a homebrew after themselves.

All I know is that My compressor went on for 10 mintues, and now 48 minutes have gone by and it STILL hasn't come on again, so perhaps maybe the little pill bottle of water must be doing something right.

Or it's the thickness of my collar, or my insulation, or my fan, or the silicone caulk I used, or lucking out on figuring out the right settings on my Stc1000, or all of the above.

So go have your homebrew, probably brewed in ways that would probably cause apoplexy in some of our members because of one thing or another that YOU did. That's just the nature of the beast. We always don't agree on how to do things...and usually the various ways ALL achieve the same end.:mug:
 
Well, countless people on here, many who probably know just as much if not more than you (and definitely a LOT more about this than me) have made a great case for submerging the temp probe, not to as you say trick it but in order to get a more accurate temp reading since, and this makes sense to me, that we are interested in controlling the temps of a FLUID and not the surrounding air....

I think it's not accuracy so much as normalization.

The data that you get from a thermal sensor in free air is as accurate as ever, but what it's telling you is not necessarily what you want to know. Unlike a liquid caught in a thin channel or a needle attached to a bimetallic spring, it reacts immediately to the changes around it.

Laboratory types call a thermometer "traceable" when, among other things, it averages the temperature it reads over a pre-determined duration of time, so that when they open the door of a refrigerator the number on the display gives them a useful impression of the temperature it has been reading instead of a reaction to the door being opened.

By adjusting the software parameters of a PI (Program and Interval) controller - in this case the c0 and c1 parameters to do what the OP asked, and r0 to make it less interested in minor deviations in temperature - we can make it behave in much the same way as it behaves when you swamp the sensor with a big thermal buffer. Bolting it to a large piece of metal would have a similar effect, fwiw, if you had good enough thermal contact.

But it won't make the display on the willhi controller change it's digits slower.

I referenced being a QA manager not because it means that i am smart - i am clearly a ******* for getting talked into this management gig - but because it means that i have been conditioned to complain about things that don't seem right to me.
 
PI normally stands for Proportional and Integral, not Program and Interval.

I have a homemade temp controller and I have the 'dead band' set so that the freezer is off for 30-45 minutes before kicking on for about 10 minutes. This results in about 12 degrees F of temp swing, measured by air temp. If the OP can adjust his temp controller's dead band, I would do that just because it's easier, if not, the jar of water trick or strapping it to a keg works as well.
 
OK Tylo I don't know if this will help, or if the way to set things on a love are the same as it is for my Ebay STC-1000 but I'm at 1 hour and 22 minutes since the last time my compressor kicked off (it ran for 10 minutes) and it JUST kicked on, and the temp has only risen .5 degrees C (that's what 3 degrees f?) in that time, but the settings on my controller are;

Temp Set Value, 4.6 degrees C.
Difference Set Value, 0.5 degrees.
Compressor Delay Time 10 minutes (maximum for the stc-1000 iirc)
Temp Calibration Value 0.0
 
Ya. Thanks thats what I couldn't figure out. I had another thread asking how to set the perimeters on it, but lacked the response. The only thing I really figured out is how to set the set point to and like I said it has been staying between 35 and 38 degrees, which is what I wanted. The thing I hadn't figured out was how to make the cycles longer. I live in an apartment, so building this thing has taken a little longer then it normally would. The probe wrapped in bubblewrap hasn't helped as much as I needed it to. I think I am gonna go ahead and try it submersed in water. I think the pill bottle idea is pretty awesome! Also I just have the collar sitting on top of the chest freezer. I insulated the inside and taped it off, but fear of my chest freezer not working out, I have been trying not to screw, seal, or bolt anything into it. I believe I am already so deep in this project that I might as well start. I should probably seal it off a little better assuming that the probe being submersed still doesn't give me the results I am looking for.
 
PI normally stands for Proportional and Integral, not Program and Interval.

I have a homemade temp controller and I have the 'dead band' set so that the freezer is off for 30-45 minutes before kicking on for about 10 minutes. This results in about 12 degrees F of temp swing, measured by air temp. If the OP can adjust his temp controller's dead band, I would do that just because it's easier, if not, the jar of water trick or strapping it to a keg works as well.

indeed, I remembered that wrong.

It's just that i start to twitch when i read some of the bizarre lore about temperature control here and some other places. Immersing the thermocouple isn't a terrible thing, but, why not at least try to use the technology you already bought?

What really gets me is when people decide that regular PID doesn't work and instead hook up the alarm outputs to relays, so that they can accomplish with hardware what they could accomplish with the built-in software if they read the manual. and had someone around to explain the theory. I admit to having a good friend with a background in factory automation.

Decades ago, whole teams of people worked out how to do this, back when cpu cycles were really expensive. They even came up with ways to do it without software. More recently, some coder put in hours implementing the equations someone else derived. Hopefully, someone like me put in hours making sure his implementation works the way it ought. Seems a shame to ignore all that just because telling the controller to ignore temperature fluctuation until you are +3 over set point results in a less appealing user experience than mechanically buffering the input.
 
Ya. Thanks thats what I couldn't figure out. I had another thread asking how to set the perimeters on it, but lacked the response. The only thing I really figured out is how to set the set point to and like I said it has been staying between 35 and 38 degrees, which is what I wanted. The thing I hadn't figured out was how to make the cycles longer. I live in an apartment, so building this thing has taken a little longer then it normally would. The probe wrapped in bubblewrap hasn't helped as much as I needed it to. I think I am gonna go ahead and try it submersed in water. I think the pill bottle idea is pretty awesome! Also I just have the collar sitting on top of the chest freezer. I insulated the inside and taped it off, but fear of my chest freezer not working out, I have been trying not to screw, seal, or bolt anything into it. I believe I am already so deep in this project that I might as well start. I should probably seal it off a little better assuming that the probe being submersed still doesn't give me the results I am looking for.


I can relate to building it in a small apartment.

The biggest thing to help is to make as tight a seal between collar and freezer as possible. The thing about 100% silicone caulk that many of us use is that you can peel off the collar and caulk if you need to. Some folks use liquid nails but silicone is not permanent if you need it gone and won't harm the fridge.
 
I think I am gonna go ahead and try it submersed in water. I think the pill bottle idea is pretty awesome!

Submerging the temperature probe has worked well for me. I originally had my temperature probe submerged in a small water bottle inside my keezer. My compressor would kick on for 15 minutes every 1 - 1.5 hours. The temperature would fluctuate between 39.5°F and 41°F during this time. I have modified my probe setup in order to increase the cycle time. My probe is now inside a 6 inch thermowell that sits in a large Aquafina bottle. The bottle sits at the bottom of the keezer next to my CO2 tank. The compressor kicks on for 20 - 25 minutes every 3.5 - 4 hours. The temperature fluctuates between 39°F and 41°F during the cycle.
 
The probe wrapped in bubblewrap hasn't helped as much as I needed it to.
You are either misreading the original suggestion, or miswriting your application of it.

Tape, or, even easier, strap, the probe to something massive- like a keg -, THEN insulate over the top of it. Attaching the insulation with more bungy/elastic cord is what I have found to be easiest. You should be able to reduce the temp diff to ~1F and still cycle less than 1/hour. If not, insulate the probe even more (but not between it and the keg).

Your description reads like you insulate the probe, then attach it to the keg, which will not give the desired effect of reducing cycling (or controlling the temp of the beer accurately).

The collar insulation needs to be glued or fit very tightly to the collar. That makes a big difference in cycling freq as well, since wood is a poor insulator. Adding a fan will become more important as your cycle freq decreases, since the air will stratify even more during the longer off times.
 
Decades ago, whole teams of people worked out how to do this, back when cpu cycles were really expensive. They even came up with ways to do it without software. More recently, some coder put in hours implementing the equations someone else derived. Hopefully, someone like me put in hours making sure his implementation works the way it ought. Seems a shame to ignore all that just because telling the controller to ignore temperature fluctuation until you are +3 over set point results in a less appealing user experience than mechanically buffering the input.
I admit to having a good friend with a background in factory automation.
Is that like not really being a doctor, but playing one on TV?

You don't know what you are talking about. The suggestions you made about blindly increasing 'on times' and 'off times' is a guaranteed way to increase temp variance. Those control parameters are mainly used for commercial systems where they have to put the sensor on a fixed mass in a fixed location, but have varying cooling loads. It is a 'best fit' type work around, which is a complete kluge compared to the completely modern and correct approach of placing the sensor on the primary indicator- the beer. The 'minimum off' time should be used as ASD, and set for at least 10 minutes. While increased cycling freq may take a few years off the life of a freezer, not having adequate ASD can kill your compressor in a few hours, and has.

The most commonly suggested, and correct, method for a keezer is to attach the sensor to the smallest container (or to a reasonable facsimile) in the keezer, then insulate over it. The temp diff can then be used to adjust cycling freq to anything desired, at the cost of temp variance or compressor life of course. The reason to attach the sensor to a smallish container is to prevent freezing and collateral damage that may result if the sensor is on something much more massive than the smallest container, and set close to freezing.

As for those that hang their sensor in air and guess at the temp setting and differential, I would have to be even more blunt than I already have been.
 
Edit: Guess I posted a bit to quickly without reading the whole thread. Didn't realize temp control could be such a heated topic! ;)

MY original post was a question to Revvy (sorry OP to go OT for a quick moment).

Hey Revvy great post. I know you were working on the kegerator awhile back and saw the post with the nice keg handles but must have missed the post of your final kegerator. Can you PM me the link?

Thanks!

Dan
 
I think it's not accuracy so much as normalization.

The data that you get from a thermal sensor in free air is as accurate as ever, but what it's telling you is not necessarily what you want to know. Unlike a liquid caught in a thin channel or a needle attached to a bimetallic spring, it reacts immediately to the changes around it.

Laboratory types call a thermometer "traceable" when, among other things, it averages the temperature it reads over a pre-determined duration of time, so that when they open the door of a refrigerator the number on the display gives them a useful impression of the temperature it has been reading instead of a reaction to the door being opened.

By adjusting the software parameters of a PI (Program and Interval) controller - in this case the c0 and c1 parameters to do what the OP asked, and r0 to make it less interested in minor deviations in temperature - we can make it behave in much the same way as it behaves when you swamp the sensor with a big thermal buffer. Bolting it to a large piece of metal would have a similar effect, fwiw, if you had good enough thermal contact.

But it won't make the display on the willhi controller change it's digits slower.

I referenced being a QA manager not because it means that i am smart - i am clearly a ******* for getting talked into this management gig - but because it means that i have been conditioned to complain about things that don't seem right to me.

I stumbled upon this thread and don't mean to resurrect something just to complain, but I need to clear up a few things.

PID does indeed stand for Proportional Integral Derivative control. It's a control method that allows you to fine-tune control by adjusting the three PID parameters to account for the rate of change of your process variable as well as the historical error in the control loop. It's fantastic if you need to control a heating element, a proportional valve to control pressure or flow rate, but it can only be used on something with variable (or simulated variable) output. The parameters that you mention are in no way related to PID, the only parameters are P, I and D. But none of that matters, since a refrigerator can only be on or off PID is completely irrelevant here. We're using on/off control, not PID control.

Also, the word traceable means traceable, nothing more. A traceable thermometer is one with a traceable calibration, it says nothing about how/if it averages data.

If you're going to rant, criticize and complain, at least have some clue what you're talking about.
 
Again, I apologize. I was in a bad way. Doing poorly. Off my meds, as it were. Far more embittered than usual.

But i do have an aesthetic objection to using luddite methods to circumvent shortcomings in high-tech solutions.
 
Just to be annoying I want to point out that you can implement PID control in systems that are only 'fully on' or 'fully off', if the frequency of the switching is somewhat higher than the lag time of your system. I work with industrial furnaces that are only fully on/fully off, but they are controlled proportionally with PID because they use PWM that switches a few times per second. Your proportional variable is the duty cycle of the furnace element in this case.

This is irrelevant for freezers because you can't cycle the compressor quickly. You could use PID over very long time scales, though.

And I still fail to see any way in which changing the programming is superior or inferior to adding 'thermal mass' to the sensor. It's totally a six-of-one, half-dozen of the other situation.
 
No disagreements from me. That's why I mentioned that you can use PID for something with a variable or simulated variable output (duty cycle like you pointed out). :mug:
 
Again, I apologize. I was in a bad way. Doing poorly. Off my meds, as it were. Far more embittered than usual.

But i do have an aesthetic objection to using luddite methods to circumvent shortcomings in high-tech solutions.
Nice passive/aggressive apology.

Exactly what luddite methods are you talking about?

The high tech solution is to use the ASD provided by modern digital controllers (except for Ranco) for guaranteed hot start protection, instead of the wish and a prayer method most home freezers use.

The other modern solution is to specialize the system by using control inputs that directly sense the thing being controlled, instead of some secondary indicator. Since the thing we want to control is thermally massive, this is a win/win- tighter, easier temp control, and fewer cycles (which was the whole point of this thread). There is no amount of software in an off the shelf controller that can replace this approach.

Most digital controllers have <1F temp differential resolution. If tighter control than that is desired, it would most likely require something besides a home fridge/freezer with a fixed speed compressor.
 
Also, the word traceable means traceable, nothing more. A traceable thermometer is one with a traceable calibration, it says nothing about how/if it averages data.

If you're going to rant, criticize and complain, at least have some clue what you're talking about.
Glad you were the one to stick it to him about that one. I was tempted when I saw it, but already catch enough grief for the reaming I do.
 
Nice passive/aggressive apology.

Exactly what luddite methods are you talking about?

How is it passive/aggressive when I'm just stating my opinion?

I'm serious when i say i was not myself when i made those posts. But i am also serious that i have an aesthetic -- not technical -- objection to having to result to a physical solution to a software problem.

Sticking a thermistor inside a container of water - to use the water as a physical buffer against temperature change - can be observed and characterized and expressed as a hysteresis. I can already program a hysteresis into my $13 ebay aquarium controller.
 
How is it passive/aggressive when I'm just stating my opinion?
When you apologize for a previous comment, then make the essentially the same comment again that calls everyone a cretan that chooses to place their sensor on their beer.

I'm serious when i say i was not myself when i made those posts. But i am also serious that i have an aesthetic -- not technical -- objection to having to result to a physical solution to a software problem.

Sticking a thermistor inside a container of water - to use the water as a physical buffer against temperature change - can be observed and characterized and expressed as a hysteresis. I can already program a hysteresis into my $13 ebay aquarium controller.
I have an aesthetic AND a technical objection to resorting to a software solution for a physical problem, especially when there is no software solution for the given physical problem.

Your argument has already been debunked- several times. This is not a software/controller problem, and without also including a sensor input for the object being controlled, there is no way for any software to both decrease cycling AND maintain temps of the beer.

For a single sensor controller, the best way to reduce cycling and maintain temps is to put the sensor on the smallest mass being controlled (a bottle, a keg, or a facsimile), then use the temp diff to balance temperature control vs. cycling. They go hand in hand. This has been explained many times.

Measuring air temps, or some other mass smaller than that being controlled, it is possible to use a temp setting and temp diff that would result in the desired balance of temp control and cycling. To do this, you could either spend a couple of weeks tuning it manually; spend a few days (or more) writing some code to do it for you (with an additional sensor, and many tuning cycles for the code to converge on the proper settings); or you could simply put the probe on the thing being controlled, set your controller for your maximum temp tolerance, and have everything be automatically optimized.

RE: Hystersis- More than likely, that is not a true 'hysteresis' setting on your ebay controller, it is temp differential. Hysteresis is an oft mis-applied term as demonstrated by both of your inaccurate applications of it. The subject is too boring to go into here, especially since it is not relevant.
 
Back
Top