Thermostats hard on keezer?
- Thread starter evanos
- Start date
Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:
samc
Well-Known Member
You just need to have a thermostat that has a short cycle delay, which prevents it from cycling on and off too quickly. Mine is set at the max which is 12 minutes, although in practice it would be rare for it to want to go on/off so fast unless you were constantly opening it for some reason.
Duty cycle can be hard on compressor/condensers. Iknow that the johnson digital controls have a function which allows the cooler to go a few degrees above the desired temp before turning on the unit. I'm sure you can look for this in your temp controller.
I had an old (really old) freezer that died not long after using it as a keezer with external temp control. Got a new freezer and it's still going. Turning on and off the compressor too frequently can definitely be hard on it but if it is set right and doesn't do it too frequently, the freezer is off which means the compressor isn't running at all, particularly in cold months (assuming ambient temp doesn't drop way below freezer level). I think of it as cancelling each other out. Compressor on/off frequency = compressor not running time. I'll let you know when my keezer dies but it's been going about a year now with no probs.
ajwillys
Well-Known Member
Yeah, I use a Ranco and it has the ability to set a range. The wider you set the range, the easier it is on the compressor. I think I have my keezer with a four degree range, which takes a long time to warm up to that when it's full.
As far as I know compressors are always running "full bore". The only thing you are changing is what is triggering the compressor to turn on. It is operating at a higher temp than it would as a freezer which means it has to be on less often(this may not be the case with a collar and more leakage).
Preventing short cycling will be gentler on the compressor. Besides buying a controller with a short cycle delay setting, putting the sensor in/on a container of liquid to dampen its response will prevent short cycling without having the large temperature swings a large offset can cause (edit- For clarity, I should have added that this was in combination with reducing the "temp differential" parameter available on most controllers. It make no sense otherwise, as the liquid temps will vary at least as greatly as the "temp diff" range.- end edit). Sensor response varies based on the type, but the way we use them, most will benefit from some damping. A combination of sensor response dampening and short cycle delay (for insurance) is one of the better approaches, since it maintains a tight temperature range, with guaranteed short cycle protection.
Damping the sensor response will also prevent cycling due to lid openings, which will happen if the sensor is a fast responding type, and is in air only.
For fermenting in a keezer, putting the probe directly on/in the active vessel will give the tightest control over ferm temperature. Short cycling could occur for a vigorous fermentation, but as long the controller has a cycle delay, it isn't an issue.
Damping the sensor response will also prevent cycling due to lid openings, which will happen if the sensor is a fast responding type, and is in air only.
For fermenting in a keezer, putting the probe directly on/in the active vessel will give the tightest control over ferm temperature. Short cycling could occur for a vigorous fermentation, but as long the controller has a cycle delay, it isn't an issue.
"Besides buying a controller with a short cycle delay setting, putting the sensor in/on a container of liquid to dampen its response will prevent short cycling without having the large temperature swings a large offset can cause."
Not true. You will have even larger temperature swings with a submerged probe. That's what the differential setting on the controller is for.
"Dampening the sensor response will also prevent cycling due to lid openings, which will happen if the sensor is a fast responding type, and is in air only."
So what? That's what the anti-short cycling feature is for. Mostly this doesn't happen much anyway as the cold air stays put in a chest freezer. IMO, fast probe response is advantageous and there is no need to dampen it.
"For fermenting in a keezer, putting the probe directly on/in the active vessel will give the tightest control over ferm temperature. Short cycling could occur for a vigorous fermentation, but as long the controller has a cycle delay, it isn't an issue."
Except that IMO, it doesn't. Your freezer will become an ice box by the time the controller responds when cooling and the reverse on temperature rise.
I've found that the positioning the controller probe in the air near the bottom of the freezer works best (when using a fan, where you place the probe makes little difference). Then, a separate digital thermometer is used to monitor the fermenter temperature with the probe attached to the side and covered with some bubble foil (or any other) insulation. I adjust the controller based on the thermometer reading. It holds within about 1*F. The thermometer probe could also be used with a thermowell in the fermenter, but not the controller probe.
At one time, I thought that putting the probe in a container of water would be a good idea. The problem is that it is difficult to use the controller as both a controller and a thermometer. We need to know the temperature in two different locations at the same time. Tough to do with a single probe.
Not true. You will have even larger temperature swings with a submerged probe. That's what the differential setting on the controller is for.
"Dampening the sensor response will also prevent cycling due to lid openings, which will happen if the sensor is a fast responding type, and is in air only."
So what? That's what the anti-short cycling feature is for. Mostly this doesn't happen much anyway as the cold air stays put in a chest freezer. IMO, fast probe response is advantageous and there is no need to dampen it.
"For fermenting in a keezer, putting the probe directly on/in the active vessel will give the tightest control over ferm temperature. Short cycling could occur for a vigorous fermentation, but as long the controller has a cycle delay, it isn't an issue."
Except that IMO, it doesn't. Your freezer will become an ice box by the time the controller responds when cooling and the reverse on temperature rise.
I've found that the positioning the controller probe in the air near the bottom of the freezer works best (when using a fan, where you place the probe makes little difference). Then, a separate digital thermometer is used to monitor the fermenter temperature with the probe attached to the side and covered with some bubble foil (or any other) insulation. I adjust the controller based on the thermometer reading. It holds within about 1*F. The thermometer probe could also be used with a thermowell in the fermenter, but not the controller probe.
At one time, I thought that putting the probe in a container of water would be a good idea. The problem is that it is difficult to use the controller as both a controller and a thermometer. We need to know the temperature in two different locations at the same time. Tough to do with a single probe.
You can set ASD of about 10 minutes OR damp the sensor by taping it to the fermenter and also putting a square of bubble wrap over the probe. This is required with an analog controller.
motobrewer
I'm no atheist scientist, but...
catt speaks the truth.
My fridge controller is able to log data. here is some:
https://spreadsheets.google.com/ccc?key=0AriPJSz0jUx6dHBLS05ZUnIzVlplZmdKUXdUNk0wU0E&hl=en
note that i didn't set up the logs to record at the same time intervals. i can get better data when i get home. but there are two time columns so you'll have to match up accordingly.
my probe is sitting in air. just hangin' in the air. i have a big, 27cuft, 6x2.5x2 (ish) freezer.
as you can see in the data, my compressor runs for 2-4 minutes every hour.
catt is dead on about fermentor control. if you send a probe down the thermowell, the freezer will get very cold by the time the setpoint is satisfied. if you really wanted to go that route put the probe in air for the first 24 hours or so to let the beer temp get down, then you can send it down the thermowell
note, i just lagered at 35F with the probe in air temp, but i had another probe down the thermowell just to see. the probe down the well never changed temp.
My fridge controller is able to log data. here is some:
https://spreadsheets.google.com/ccc?key=0AriPJSz0jUx6dHBLS05ZUnIzVlplZmdKUXdUNk0wU0E&hl=en
note that i didn't set up the logs to record at the same time intervals. i can get better data when i get home. but there are two time columns so you'll have to match up accordingly.
my probe is sitting in air. just hangin' in the air. i have a big, 27cuft, 6x2.5x2 (ish) freezer.
as you can see in the data, my compressor runs for 2-4 minutes every hour.
catt is dead on about fermentor control. if you send a probe down the thermowell, the freezer will get very cold by the time the setpoint is satisfied. if you really wanted to go that route put the probe in air for the first 24 hours or so to let the beer temp get down, then you can send it down the thermowell
note, i just lagered at 35F with the probe in air temp, but i had another probe down the thermowell just to see. the probe down the well never changed temp.
No, you won't. Unless you are talking about the air temp, which I could care less about. The temp of the important stuff, the beer, will be more stable because you can set the differential lower without short cycling- thermal mass is your friend. Without putting the probe in/on something with thermal mass, the temp of finished beer will be the same as the average freezer air temp, which varies as a result of external ambient air. In reality, the temp differences between putting the probe in air or in a thermal buffer won't be that great. They will just be more consistent with it in a thermal buffer/dampener because you can set the actual temp, not some offset of avg freezer air temp, which itself varies. Also, the compressor will cycle less because it won't stop until the larger thermal mass reaches the set temp.
If the probe response isn't dampened, the short cycle isn't in play (the compressor hasn't run recently), and you decrease the interior air temp below the setpoint by opening the lid, the compressor will cycle on. Whereas with a dampened sensor, the thermal mass of the contents (beer) would likely have brought the internal air/chassis temp back down within bounds without needing the compressor to extract any heat.
Again, I couldn't care less about what the air or internal freezer wall temperature is. I would rather have the temperature of my beer, fermenting or otherwise, stay within bounds, which can be set much tighter when sensor response is dampened. The air and freezer walls/freon loops have very little thermal mass compared to the beer. If not, add some jugs of water. Full freezers cycle less, and are more temp stable regardless of what you are using them for.
All of your issues arise from not using the controller to control the temp you want controlled. There are 2 controllers in your loop, you being one of them. You are manually adjusting the temp based on reading 2 thermometers, instead of letting the controller do its job.
There is no way that one steady air temp during an entire ferm cycle will keep the beer at a steady temp, even with a fan, which does help with removing heat from the ferm vessel. The heat produced during ferm changes too greatly.
Again, if you quit inserting yourself into the loop as a controller, you don't NEED to know any temps, other than the temp set point for your beer. Attach the probe to the active ferm vessel, or for finished kegs put it in a jug (or on a keg), and let the controller do what it does best- control temps. If you WANT to know both the temp of the ferm vessel (or kegs) and the temp in the keezer, which is a good backup/sanity check, a separate thermometer in the freezer will do that.
I've been meaning to try to model this with an electric circuit analog, but so far haven't found the time. Anyway, here's my opinion.
The average temperature of your beer will stabilize at the average temperature of the air in the freezer. If your probe is in the air, the beer temperature will not vary much around its average level, but the freezer comes on and off more often, but is on for a shorter period. But the average beer temperature may be several degrees off from the controller setting - not a problem if you know what the difference is.
If you have your probe in a thermowell in the carboy, you will maintain tighter control between the controller set temperature and the average beer temperature, but the beer temperature will fluctuate more about the average level, and the freezer air temp. will get very cold. Having the probe in a container of liquid is a compromise between these two extremes.
The only problem I can see with having the probe in the air, assuming your freezer isn't short cycling, is that the difference between the temperature set on the controller and the actual beer temperature may vary depending on how active the fermentation is - a problem that is alleviated by having the probe in a thermowell, but at the expense of greater beer temp. fluctuation about the average.
The average temperature of your beer will stabilize at the average temperature of the air in the freezer. If your probe is in the air, the beer temperature will not vary much around its average level, but the freezer comes on and off more often, but is on for a shorter period. But the average beer temperature may be several degrees off from the controller setting - not a problem if you know what the difference is.
If you have your probe in a thermowell in the carboy, you will maintain tighter control between the controller set temperature and the average beer temperature, but the beer temperature will fluctuate more about the average level, and the freezer air temp. will get very cold. Having the probe in a container of liquid is a compromise between these two extremes.
The only problem I can see with having the probe in the air, assuming your freezer isn't short cycling, is that the difference between the temperature set on the controller and the actual beer temperature may vary depending on how active the fermentation is - a problem that is alleviated by having the probe in a thermowell, but at the expense of greater beer temp. fluctuation about the average.
Your sheets headings make no sense, to me at least, unless "basement temps" is mislabeled, and is really "beer temp" or ?. Otherwise, what is your chart showing exactly?
And again, who cares what temp the freezer gets to? There should be very little thermal inertia overshoot due the air/walls/freon loops, unless all you are chilling is a growler. In that case, put some jugs of water in there.
By lagering, do you mean your ferm was already over? Either way, an active lager ferm (at the lower lager ferm temps) produces less heat (per time) than ales (at ale ferm temps). So, the temp differential between freezer air and ferm vessel will be less, and temp variance will be less, than with ales.
For controlling fermentation temps, where there is a heat source within the freezer (ferm'ing beer), and a heat/cold source outside (ambient air), placing the probe on/in the active ferm vessel will give more stable temps. Especially since the temp differential can be set much smaller without worrying about unnecessary cycling.
For controlling finished keg temps, putting the probe in air vs. in a jug just means more cycling, or more temp variance, depending on how much you like your compressor, or not.
I assume you are talking about active fermentations in your post.
The problem is different for keeping kegs cold, vs. maintaining active exothermic ferm temps, so keep that in mind in your analysis.
One advantage to dampening the probe response, is that a tighter differential can be set (for controllers that allow it) without worrying about short cycling. The thermal mass of the beer will hold its temp long after the freezer air/walls have risen above the set point. The same thermal mass will absorb the thermal inertia/carryover from the air/walls/freon loops, etc. without overshooting by a significant amount.
The OP didn't say whether he was using the freezer for fermentation or as a keezer, so I assumed for fermentation, and assume that active fermentation will release heat.
Here's my reasoning, in more detail. Both the air temperature and the beer temperature will fluctuate about some average value as the freezer cycles on and off, and the average beer temperature is the same as the average air temperature. The beer temperature does not fluctuate around the average as much as the air temperature does because of its large thermal mass. When the freezer turns on, the air temperature falls rapidly, then when the freezer turns off, the air temperature begins a long slow rise back to the controller set point, where the freezer turns on again.
1. With the probe in the air - Maximum air temperature is at the controller set point, but the average air temp (and beer temp) is several degrees colder than the set point. The freezer does not stay on very long, so the air temp. does not get super cold and the beer temp. does not cool down much below its average value.
2. With the probe in a thermowell in the beer - The freezer turns on when the beer temp rises to the controller set point and stays on until the beer cools enough to turn the freezer off. At this point, the air is super cold and the beer will continue to cool, so the beer temperature varies more about its average than in situation 1. above. Nevertheless, the difference between the maximum beer temp. (controller set point) and the average beer temp is less than the difference between the max air temp and average air temp in 1. above, so even though you have more beer temp variation about the average, the average is closer to the set point.
3. With the probe immersed in a liquid (volume much less than the volume of beer) - a compromise between 1. and 2.
With active fermentation releasing heat into the beer (which in turn is released into the air) and the probe in the air, due to the thermal resistance between the beer and the air, the beer temperature will not be at the average air temperature, as was assumed in 1. above, but will be at a higher temperature, maybe several degrees higher. With no fermentation, the assumption that the average beer temp. is the same as the average air temp. holds.
It should be fairly obvious, at least to you and me it seems it is, that having the probe in the air will not control the temp of a fermenting beer very well (without inserting yourself into the loop to adjust the offset).
Similarly, for fermenting beer, not having the probe directly on/in the ferm vessel (i.e. in air or a water jug) will require calculating an setpoint for the freezer that will compensate for the heat being generated in the ferm vessel (based on educated guesses from empirical info) to achieve the desired ferm temp. This setpoint will change during the fermentation cycle.
Another issue is short cycling, or excessive/extra cycling, of the compressor. That is where placing the probe in/on a thermal buffer will result in less cycling (higher duty cycle, longer run time per cycle). If the controller does not allow for temp differential adjustment, this will result in a larger temp swing (equal to the fixed temp differential) than may be possible with the probe in air. It will still cycle less, though. It is a really a moot point, since most controllers allow for differential adjustment.
Your analysis is mostly right. The only part that may be off a bit is the amount of effect the thermal inertia/hysteresis of the freezer air/walls/freon loops will have on the temp of the beer. Their thermal mass is much smaller than the beer, especially if there are multiple vessels in the freezer.
"At this point, the air is super cold and the beer will continue to cool, so the beer temperature varies more about its average than in situation 1."
This would only be true for very small quantities of beer in a large freezer. A situtation easily remedied by adding some containers of liquid.
Putting the probe in a jug of water instead of on a keg, is more about convenience. The gains for reduced cycling and tighter temp control aren't worth the hassle of moving the probe each time the keg floated/moved/etc. At least not for me. Putting the probe on/in a ferm vessel during active fermentation is worth the added temp control, for me at least.
If someone (not you DeafSmith) doesn't like the idea of putting the probe in a jug or on/in a ferm vessel, that is their choice. It doesn't change the laws of physics, though.
Cat22 has posted his opinion about not dampening response of the temp probe and not placing it on/in the ferm vessel in several threads, and continually gets questioned about his questionable logic/science. It hasn't stopped him yet, and I guess it won't now either.
Yeah, I was aware of that, just trying to keep things simple.
Or if the thermal mass of the air, freezer walls, and cooling coils was a significant fraction of the thermal mass of the beer. The thermal mass of the beer is easily calculated, but I really don't have a good feel for the others.
What I've been doing with my fermentation freezer is to put the probe inside of an empty White Labs yeast vial, surrounded only by air. This dampens the response of the probe some without the risk of shorting something out by putting it in liquid. This setup, in combination with a fermometer stuck to the side of the carboy for manual controller setpoint adjustment, has been working well enough for me, but I admit it is not ideal. I'm debating whether or not get a thermowell. The yeast vial may, however, work fine for my keezer when I start kegging (soon - I'm accumulating all the pieces).
Why not try taping the controller probe to the side of the ferm vessel, topped with some insulation. That will give you a close idea of how a thermowell will perform, and might be good enough for your tastes.
I am not attempting to speak with a tone of authority, it may only sounds that way because the logic and science are sound. In your opinion, they- the laws of physics- might be BS, and that is fine with me. Just don't expect me, or the 5 physicists in the other thread that were doubting your theory, to share your opinion.
Did you adjust the temp differential of your controller to a smaller value (1-1.5F for starters), or is it even adjustable on your controller?
If your controller does not have an adjustable temp differential, using an un-damped or lightly damped (like you have done with the yeast vial) sensor, then figuring out what temp to set the controller to so your desired vessel temp is maintained, is about as good as you can get. It isn't difficult to do for finished kegs, but can be for fermenting since the offset will vary by day.
Using a fixed differential controller with a thermowell in a large thermal mass, or with a highly damped sensor will guarantee a temperature swing equal to the fixed temp differential of the controller which can be as high a 5F.
Thanks for the reply! What probe/software combination do you use?
motobrewer
I'm no atheist scientist, but...
evanos, it's kind of a custom/repurposed controller.
cwi, that data was just fridge air temp and basement temp. this is for my keezer which I occasionally lager in. I was just trying to answer the OP's question about how hard the controller runs the keezer.
by lager, yeah fermentation was over (it was actually an ale, a tripel) so there wasn't any heat. i certainly agree with you that controlling to the thermowell is more accurate. it's probably the way to go IF your starting wort temperature is at or close to your setpoint, otherwise the keezer will get very cold before your wort does. if you have a dedicated keezer for fermentation control, you're right, you really don't give a crap what happens aside from beer temp. but i have kegs/bottles in mine that i don't want to freeze
i still can't get on board with your idea that controlling to the thermowell will cause less overall temperature swing. right now, controlling to air, the compressor runs 2-3 minutes and swings (sp is 41F) between 39.4326F and 42.1447. there's 3 5g kegs, a 5g carboy, and a bunch of bottled beer in there. not packed full but decent mass.
what do you think the time constant is for the fermenting beer? you think it will cool down to setpoint within 3 minutes? any longer and I would have larger temp swings.
note, my control strategy is just "on at >41F, off at <41F". i have a 2 minute "min/on min off" restriction on that output.
cwi, that data was just fridge air temp and basement temp. this is for my keezer which I occasionally lager in. I was just trying to answer the OP's question about how hard the controller runs the keezer.
by lager, yeah fermentation was over (it was actually an ale, a tripel) so there wasn't any heat. i certainly agree with you that controlling to the thermowell is more accurate. it's probably the way to go IF your starting wort temperature is at or close to your setpoint, otherwise the keezer will get very cold before your wort does. if you have a dedicated keezer for fermentation control, you're right, you really don't give a crap what happens aside from beer temp. but i have kegs/bottles in mine that i don't want to freeze
i still can't get on board with your idea that controlling to the thermowell will cause less overall temperature swing. right now, controlling to air, the compressor runs 2-3 minutes and swings (sp is 41F) between 39.4326F and 42.1447. there's 3 5g kegs, a 5g carboy, and a bunch of bottled beer in there. not packed full but decent mass.
what do you think the time constant is for the fermenting beer? you think it will cool down to setpoint within 3 minutes? any longer and I would have larger temp swings.
note, my control strategy is just "on at >41F, off at <41F". i have a 2 minute "min/on min off" restriction on that output.
Had you placed the sensor in a jug of water, a normal/standard controller would have cycled even fewer times per day. It may have cycled less with your controller, too; but your controller doesn't sound normal, and you may not be setting it up properly. See later discussion for more info on that.
That is where some people don't, and defend putting the sensor in air as a more accurate method, even though it defies the laws of physics. If I proposed that the best way of controlling the temp of the mash was by placing a sensor in the burner flame/heat, and, based on past empirical/anecdotal info, maintaining some temp offset from what the sensor reads to what the mash should be, no one (hopefully) would agree with me. Yet, that is exactly the same approach as putting the sensor in the air to control a ferment.
There is no such thing as a free lunch. It is hard to ramp, or maintain ferm temps for one vessel, without affecting its neighbors. For most ale ferments, the initial wort temp IS very close to the desired setpoint, temps are fairly far from freezing, and the heat generated/extracted is small enough, that the temps of surrounding vessels freeze, or even change that much.
If you "warm pitch" lager ferment in a chamber, because of the initial ramp phase, it has good chance of freezing any other vessels that are too small to endure the heat loss of the initial ramp. Some lager ferm temps are too close to freezing, and ramping more than a few degrees may cause small vessels to freeze. Post-ramp from pitch to ferm temp, lager ferments will generate much less heat power (heat/time), and neighboring vessels will remain closer to the setpoint than with ales. The ramp to ferm temps also needs to happen in a controlled fashion, not "as quick as possible". That can't be done reliably by setting the controller to the ferm temp, and walking away. If you have empirical/anecdotal info from many past ferments, you could come up with a method that matches the ramp slope, and it may involve putting the sensor in the air. However, any changes to the system will require adjusting the method. The ramp rate is more repeatable/predictable if the setpoint is changed over time, either manually, or with a controller that has stepping/ramping capabilities.
The slope/speed of the desired temp ramp will determine what the temp of the chamber will be during the transition. If that causes freezing of other vessels, but the vessels must remain in the chamber, something has to give. That gets into personal choice for what is more important. If it were me, I would just remove the small vessels (bottles), and put them in a cooler/refrig for the day/days.
I think a lot of the confusion around temp control is caused by hidden constraints and/or setups/uses that are out of the norm. If the wort is not close the desired setpoint, that is not simply maintaining temp, that is ramping/crashing. When ramping with the probe placed on/in the vessel, there will be some temp overshoot (for each initial ramp cycle), and possible collateral freezing damage if other vessels' thermal mass isn't great enough to endure the initial heat loss without freezing. Depending on conditions, the other vessels may cause additional temp oscillations to the ferm vessel as they absorb heat getting back to the keezer's setpoint/internal temp.
I don't know what you mean by "overall temp swing". The important temp swing/avg is the temp swing/avg of the beer, and most importantly the fermenting beer, if there is any. How could it not be more accurate with the sensor in the beer or ferm'ing wort? If overshoot and temp variance between vessels of different size is the issue, see below about "thermal inertia" and vessels of disparate size.
Are you are talking about overshoot caused by the keezer's thermal inertia? That is an issue caused by an improperly sized chiller, but can be mitigated. In general, the keezer's inner air/walls/foam/freon thermal mass is small enough relative to the probed ferm vessel and other vessels that the thermal inertia has a negligible impact. I think you said your freezer is >25cf. That is above the norm, and has a larger thermal mass than normal. As long as you start with all temps close to the setpoint, and you add enough additional passive thermal mass, it should not be a problem. The additional passive thermal mass should actually keep temps more stable, with less compressor cycling. The passive vessels' temp will be some offset below the setpoint during exothermic ferm, but it should not be a large amount.
If you have very large thermal mass actively fermenting, and very small passive thermal masses (bottles) in your keezer, the smaller ones should not be placed close to freon loops, metal walls, or in the bottom of the freezer if no fan is used. A good strategy is to keep the small vessels in an insulated box inside the keezer, or very close to a much larger thermal mass.
If you are trying to both maintain an active ferm temp as accurately/tightly as possible, and keep the temp of some other vessel at the same ferm temp, that is impossible. You might as well ask the controller to chew gum and pat its head at the same time. You would have to personally choose some balance point between keeping the active ferm temp stable and keeping the other vessel within temp bounds.
I am not sure exactly what the on/off parameters you refer to do, but it sounds like you may be placing too much of yourself in the control loop. Let the controller do the controlling. Setting the temp differential to "0" (>41F & <41F) is not a normal approach, and is restricted on most controllers to ~1F - even higher for ones designed for use with the probe in air. The min off time is referred to as the "short cycle prevention delay". That is not (normally) used as any kind of temp control, but as a protective measure for the compressor.
Your controller has an additional min "on" parameter. For normal temp maintenance applications, the min on time should be used for compressor protection more than anything. With the sensor in/on the vessel, it can be set to 5-10 minutes, and should not normally come into play. For sensor in air method, it , dictating a min "on" time may be necessary for undamped fast response sensors, which is why it is not a preferred method aside from temp control issues.
If I were using your controller, from what limited knowledge I know of it, to maintain (not ramp from the "warm pitch") a lager ferm, I would do this-
Place the probe on/in the ferm vessel. Set the temp diff to at least .5F (41F, 41.5F?), set the min on time to a "sanity check" type value (5 min?); set the min off time to a standard compressor protection value (10 min?); and then let the controller do all the work. If bottle freezing occurs, move them higher and away from walls. From your data,your compressor appears to be cycling much more than I would like mine too, but that may be a tradeoff you can live with to be able keep bottles wherever you want in the keezer, ramp/crash, ferment, etc, and not worry about freezing anything.
Note- all of the following assumes normal conditions- a controller with a set point and temp diff adjustment, and short cycle delay protection; a fast response sensor; relatively high thermal mass of vessels to keezer; fairly consistent size/mass of all vessels; no small vessels (bottles) placed where they experience large unbuffered cold temp swings (walls, bottom of keezer, etc.).
Damping the sensor will allow the temp differential to be set to a smaller variance than is possible with an undamped sensor, without adversely affecting compressor cycling.
For maintaining ferm temps, placing the sensor directly on/in the ferm vessel is the simplest, most straight forward, repeatably accurate, strategy.
For maintaining finished beer, there are a few strategies that work equally as well for temp control. What changes is usually the frequency of compressor cycling, and internal temp variance when ambient temps vary greatly. Again, placing the probe on/in an avg size vessel in the keezer will maintain more stable temps, more simply/easily, and cycle the compressor less.
For ramping temps with other vessels in the same chamber, that is a balancing act. You just need to know the options, and set priorities.
Wow, sorry guys, maybe I should have been more specific. This freezer will not be for lagering, so all the argumentation about such is frivolous. Thanks for the concise answers that are gonna help with my consideration of controlling a keezer with kegs of finished beer.
Again, thanks.
Again, thanks.
From another thread. From cat22's own words, it is very apparent that putting the sensor in the air to control fermentation temps does not work. At least not without a second controller, himself, continually adjusting the setpoint, which has to have an offset to account for the heat of the fermentation, which varies.
With the probe on/in the fermenter, and the temp diff set to 1F, the beer will stay very close, ~1F, to the setpoint, without having your in-laws come over daily and adjust your keezer for you when you want to go away for the weekend. Your compressor will also cycle less and last longer.
Same goes for maintaining finished beer. You just can't expect to hoist in a room temp 1/2 barrel, slap the probe on it, and not expect the other things to get colder than the setpoint. That is why, for finished beer, placing the probe in a smaller jug of water strikes a balance between chill time for new additions, compressor cycles, and freezing existing kegs.
Most of this was all a bit off-topic. Motobrewer and I were having a back and forth. Cat22 bowed out, as usual with his "my way works, your fancy book learnin' don't mean squat" exit, once someone questioned his method.
Specific to your sticking it to your friend, most controllers have an anti-short cycle delay (ASD), which prevents it turning on shortly after it just ran. By putting the probe in/on something thermally massive, and keeping the freezer relatively full, the ASD shouldn't come into play. The compressor will only turn on a few times a day, and run for much longer than it would with it in air.
There is a very complicated way to maintain keg temps more tightly than the lowest temp diff setting (~1F) will allow. That way would indeed involve putting the probe in air, and calculating an offset from avg air temp for the set point, which would vary based on the ambient (external) air temp effects on internal temp. It would be difficult to do, and would be guaranteed to cycle the compressor much more.
That way is similar to what cat22 proposes for maintaining ferm temps. He just leaves out the part about "I manually adjust the temps constantly to keep the temps in range." when he foists his method.
motobrewer
I'm no atheist scientist, but...
holy cats thats a long post.
can't, which is why i'm saying this
exactly, which is why i do what I do
I already agreed with you that it isn't more accurate. But i'm talking about overall temp swing of the keezer air, most importantly, not freezing bottles.
lol, well i didn't size the chiller. yes, my keezer is big...again, which is why i do what i do.
well again.....it's why I'm doing what I'm doing. when i lagered the tripel, i set the controller to control to the air temp sensor at 36F, knowing it would average out to 35F. I did this while the beer was coming down from a ferm temp of 79F. If i were to put the probe in the thermowell, at 79F, and set it to 36F, i would have a keezer full of glass and beer. After the beer got down to 35F, I thought about switching the controller to control to the thermowell probe, but the thermowell probe hadn't moved 0.1F so i said fuggit and left it alone.
not really. i set the setpoint and walk away. on/off means "energize the relay at <41F, de-energize at >41F". Two minute timers were to protect the compressor.
really? two minutes every hour? i thought that was pretty good myself. how often does yours cycle? again, it's a 27cuft keezer with a collar that isn't exactly sealed perfectly
wait, i thought "inserting yourself into the control loop" was one of the 7 deadly sins??
thermal mass works both ways. when i lagered my tripel, the air temp swung +- 2 degrees. the beer had enough thermal mass to dampen that response to zero. if i were to send the probe down the well, and set the setpoint to 35F, are you saying there will be ZERO temperature swing in the beer? I don't think so, it would probably vary +-1 degree F.
min on time 5 minutes? My keezer would get very cold after 5 minutes
can't, which is why i'm saying this
exactly, which is why i do what I do
I already agreed with you that it isn't more accurate. But i'm talking about overall temp swing of the keezer air, most importantly, not freezing bottles.
lol, well i didn't size the chiller. yes, my keezer is big...again, which is why i do what i do.
well again.....it's why I'm doing what I'm doing. when i lagered the tripel, i set the controller to control to the air temp sensor at 36F, knowing it would average out to 35F. I did this while the beer was coming down from a ferm temp of 79F. If i were to put the probe in the thermowell, at 79F, and set it to 36F, i would have a keezer full of glass and beer. After the beer got down to 35F, I thought about switching the controller to control to the thermowell probe, but the thermowell probe hadn't moved 0.1F so i said fuggit and left it alone.
not really. i set the setpoint and walk away. on/off means "energize the relay at <41F, de-energize at >41F". Two minute timers were to protect the compressor.
really? two minutes every hour? i thought that was pretty good myself. how often does yours cycle? again, it's a 27cuft keezer with a collar that isn't exactly sealed perfectly
wait, i thought "inserting yourself into the control loop" was one of the 7 deadly sins??
thermal mass works both ways. when i lagered my tripel, the air temp swung +- 2 degrees. the beer had enough thermal mass to dampen that response to zero. if i were to send the probe down the well, and set the setpoint to 35F, are you saying there will be ZERO temperature swing in the beer? I don't think so, it would probably vary +-1 degree F.
min on time 5 minutes? My keezer would get very cold after 5 minutes
cwi,
IMO, your pseudo science is junk science. The only thing you have accomplished is thoroughly muddying the waters for anyone trying to sort this stuff out. The only way to know for sure is to do some actual unbiased testing. The key word there being unbiased. So, have a nice day and while your at it, insert this.
IMO, your pseudo science is junk science. The only thing you have accomplished is thoroughly muddying the waters for anyone trying to sort this stuff out. The only way to know for sure is to do some actual unbiased testing. The key word there being unbiased. So, have a nice day and while your at it, insert this.
I am trying to clarify the science, instead of giving incomplete anecdotal information. I don't need to do any testing to know that I don't want to have to babysit my ferm vessels to continually adjust the keezer temp to account for the offset caused by using a secondary indicator. That is what I bought a controller for.
You have not designed/engineered a control system, you ARE the control system.
You get questioned/corrected by an engineer/physicist every time you foist your "probe in air" method of "controlling" ferm temps. You may not have a problem with continually adjusting the keezer temp to account for ferm heat changes. Most people would, and would want to know that is mandatory housekeeping with the "probe in air" approach. I quoted your own post where you admit to having to do that. Essentially inserting yourself as the 2nd controller in the system.
Adding passive thermal mass to the system will mitigate overshoot, or PID it if you want more control, but don't want to sleep on a cot next to your ferm chamber.
Taping the probe to the ferm vessel wall and insulating, or placing it in a thermowell are the scientifically validated methods to use ONE controller (including humans). Every other approach should add swinging a dead chicken over your head as highly important to a successful outcome.
motobrewer
I'm no atheist scientist, but...
I believe you can, if just for overshoot. It may be the fuzzy logic or auto-tune capability available on many PID controllers. There is a guy on here that does industrial temp controls systems, and has used them.
I have no need to use one on a keezer. I could write a control loop in about 10 lines of code that would do the same thing. Accounting for a rolling avg overshoot is not that difficult. I just wouldn't want to rely on a PC or micro-controller board to stay up 100% of the time.
motobrewer
I'm no atheist scientist, but...
autotune? fuzzy logic?? wtf? what does autotune have anything to do with it? If he's using a PID in a industrial temp controller he's probably controlling a cooling valve or a fan, ie, a proportional output
ok, you can use a PI (never D) to control a multi-state output (altho not "just for overshoot", not sure what that means...) but it's never really desired or "best" (note, specifically not using the word "optimal". serious connotations with the phrase "optimal control"). you'd take your PI block and feed it to a state block, which has different linear state on/off levels mapped. Not very common, and it's almost always where there is more than a single stage (our freezer is just one stage...). Also note how you aren't controlling the output by the PI block alone.
the best control for a staged output is to define a deadband, and a safety min on/ min off time. then try to fit a sinusoidal wave function to the band using the temp feedback loop to tune. it's what our controllers (HVAC, not the A419) do when they control multi-stage outputs. Oh, I should mention I work for Johnson Controls. I'm in the energy group, not the control group, and know just enough about this stuff to break things. But I sit a few feet away from the guy that literally wrote the book (and multiple patents) about control theory.
circling back, i've already agreed with you. Controlling fermenting beer inside a keezer is probably best accomplished with the probe in the thermowell, since it's exothermic. I still think controlling finished beer is more accurately controlled by air temp, because, as i've stated before, the swing in air temp is so small it's dampened by the thermal mass of the beer, so the beer temp never changes. Obviously, it does require human intervention. If you were to control finished beer with the probe in the thermowell, it would oscillate +- a few degrees (whatever you set your deadband to). Which is usually good enough, and you don't have to keep screwing with it.
huh???
Motobrewer, you got a reply in before I could respond to your previous one. I will address this post, then post my previously written reponse next. There may be some duplicate info.
I wrote about the PID as joke for an extreme/overkill approach to temp control, but the guy was talking about a home fridge and using a "PID controller", which has turned into a generic term for a smart controller with PID, and usually some additional functionality. I should be more specific, and would if I knew what method/capability/programming of the controller was being used. I would be surprised, though, if a controller with the necessary capabilities doesn't exist. The extra functionality could be added fairly easily using a Xylinx/FPGA and its libraries. That is if you are familiar enough with them, and that into tight temps.
I had enough exposure to control theory to know I didn't want to do it professionally, especially designing hardware for it. Your cubemate would be a good guy to ask what the best approach/controller would be good for several standard scenarios. I don't think cat22 would be happy about his answer, either.
If you look at previous post, it was directed at catt22, not you.
I know you agree, at least since a few post ago, even though in your initial post it was not apparent since you agreed with catt22. His only post at that point was to contradict my post that fermentor temp was best controlled in/on the vessel, and that he felt that FERMENTOR sensor placement IN AIR was the best method for controlling ferm temps. He left out the manually adjusting the offset on a daily basis. Sounds better that way, I guess.
QFT
It was even more confusing because what you were calling fermenting was really ramping a 79F vessel down to lagering (or lager fermenting) temps, while also preventing collateral freezing. Most ale ferments, and some (true?) lager ferments, begin with the beer already at the setpoint, and don't need ramping/chilling to get to ferm temp.
I have stated before that I agree that for finished beer, a secondary indicator can achieve temp stability as good as, or with some planning tighter than, is possible using a primary indicator (the beer) in a system using a 'normal' controller and freezer. There will be a higher compressor cycle freq the farther the indicator is away (to the fast side) from representing the thermal response characteristics of the primary controllee. It is a personal choice between comp cycle freq and temp variance how "secondary" that indicator is due to sensor placement. Sensor placement will also determine the influence other variables, like changing external ambient temp, have on temp stability that need to be manually compensated for.
What part of running "code" (software) on a PC or micro-controller board, and worrying about the PC/board staying up for extended periods, do you not understand?
I wrote about the PID as joke for an extreme/overkill approach to temp control, but the guy was talking about a home fridge and using a "PID controller", which has turned into a generic term for a smart controller with PID, and usually some additional functionality. I should be more specific, and would if I knew what method/capability/programming of the controller was being used. I would be surprised, though, if a controller with the necessary capabilities doesn't exist. The extra functionality could be added fairly easily using a Xylinx/FPGA and its libraries. That is if you are familiar enough with them, and that into tight temps.
I had enough exposure to control theory to know I didn't want to do it professionally, especially designing hardware for it. Your cubemate would be a good guy to ask what the best approach/controller would be good for several standard scenarios. I don't think cat22 would be happy about his answer, either.
If you look at previous post, it was directed at catt22, not you.
I know you agree, at least since a few post ago, even though in your initial post it was not apparent since you agreed with catt22. His only post at that point was to contradict my post that fermentor temp was best controlled in/on the vessel, and that he felt that FERMENTOR sensor placement IN AIR was the best method for controlling ferm temps. He left out the manually adjusting the offset on a daily basis. Sounds better that way, I guess.
QFT
It was even more confusing because what you were calling fermenting was really ramping a 79F vessel down to lagering (or lager fermenting) temps, while also preventing collateral freezing. Most ale ferments, and some (true?) lager ferments, begin with the beer already at the setpoint, and don't need ramping/chilling to get to ferm temp.
I have stated before that I agree that for finished beer, a secondary indicator can achieve temp stability as good as, or with some planning tighter than, is possible using a primary indicator (the beer) in a system using a 'normal' controller and freezer. There will be a higher compressor cycle freq the farther the indicator is away (to the fast side) from representing the thermal response characteristics of the primary controllee. It is a personal choice between comp cycle freq and temp variance how "secondary" that indicator is due to sensor placement. Sensor placement will also determine the influence other variables, like changing external ambient temp, have on temp stability that need to be manually compensated for.
What part of running "code" (software) on a PC or micro-controller board, and worrying about the PC/board staying up for extended periods, do you not understand?
Motobrewer, it's funny that you "lol" about me suggesting that a PID controller could be used to control overshoot/temps in a keezer. If one isn't available that has the functionality to control overshoot for an on/off device, it wouldn't be that hard to add a circuit to turn whatever output the PID controller has (analog, PWM, SSR control, etc.) into something that could properly control a fridge. You talked about wanting to build/model an analog circuit to model the system, so I am sure you must have the skills to envision a few solutions, even if you had to resort to semiconductor devices for calculus operations to make things easier . You could even use the PID to control compressor speed through a VFD/VVD, although 1 phase motors don't last long that way, and a 3 phase or DC compressor could be pricy. I also saw the specs on a model of compressor that could be sped up/slowed down with a resistor change to convert it from high temp and low temp service. Lots of "possibilities" out there. Me, I would just hack out some code.
Regarding your system- If you have settled on your method as the "best fit" for your rather unique situation, that is your choice. Setting priorities and balancing the side effects is personal preference, so it is much like an opinion.
What is strange is the OP specificially asked about temp controllers being hard on freezers (frequent cycling) when used for high(er) temp apps, and you sided with catt22 about hanging the probe in air as the best fermenter sensor placement. This approach will guarantee close to the highest compressor cycle frequency possible, when the OP specifically wanted evidence to the contrary. It is also strange that you don't mention your choice being predicated on your secondary/tertiary constraints of ramping a 79F vessel while simultaneously trying to prevent bottle freeze in the same chamber.
Sensor placement, from lowest to highest compressor cycle frequency, would be directly on the freezer wall/compressor coils with a fast response sensor for the highest cycle freq; to inside the largest thermal mass in the freezer with a slow response sensor for the lowest cycle freq. Temp control in a freezer, for thermally passive objects, generally has the exact opposite relationship with regard to sensor placement/response. This is mainly/only due to thermal inertia/overshoot.
Anything in between the extremes of sensor response/placement is a balancing act to fit whatever the priorities/preferences are for the system. Keep in mind that he more removed the indicator is from the object being controlled, the more problematic handling exceptional situations becomes. For active ferments, I can think of a few related to varying/specific ambient conditions that would cause issues.
You feel that your 1 cycle per hour freq is "pretty good". If you are happy with that, and feel/know that a lower cycle freq will cause too much temp swing (beer temps, bottle temps, etc.), then that is what is best for you. It is strange to me that your control settings function almost equally as sensor and fixed run time based (chosen by you). This is from the high number of "min on time" duration cycles in your data. Your sensor control appears to want to shut off earlier.
If you wanted to experiment, you could probably get very comparable beer temp stability, and a much lower compressor cycle freq, by placing the sensor in/on a small(ish) mass vessel. As long as your sensor vessel/damper is small enough, and the overshoot caused by your large freezer small enough (relatively), you still shouldn't get any frozen bottles.
For the multitasking you do with your freezer, like ramping (you lump that into "lagering") a 79F vessel down to mid 30sF, without freezing bottles in the same chamber, that gets into a lot of personal preference/prioritization. I would still not choose the sensor in air approach. Personally, in this case, I think there are better ways that give more control over the ramp rate, or lower cycle freq, or both. For lager ferments, the heat output is, I assume, so low that it behaves is very close to a passive object. That should allow an indirect indicator method for controlling the temp to function acceptably, and even to a tighter range than the (normal) controller allows (yours allows 0, but has to be mitigated with "min on time"). Only testing would tell.
For the average homebrewer fermenting an ale, sensor placement anywhere but in/on the ferm vessel will cause larger ferm vessel temp swings, or continual adjustments to the controller setpoint to compensate for using the ferm chamber air as the (secondary) indicator for ferm vessel temp. It will also cause a higher compressor cycle freq. For "hands off" start-to-finish ale ferment temp control, the current "best fit" sensor placement for keezer fermenters, according to consensus based on empirical validation of control theory, is on the wall of the ferm vessel.
Regarding your system- If you have settled on your method as the "best fit" for your rather unique situation, that is your choice. Setting priorities and balancing the side effects is personal preference, so it is much like an opinion.
What is strange is the OP specificially asked about temp controllers being hard on freezers (frequent cycling) when used for high(er) temp apps, and you sided with catt22 about hanging the probe in air as the best fermenter sensor placement. This approach will guarantee close to the highest compressor cycle frequency possible, when the OP specifically wanted evidence to the contrary. It is also strange that you don't mention your choice being predicated on your secondary/tertiary constraints of ramping a 79F vessel while simultaneously trying to prevent bottle freeze in the same chamber.
Sensor placement, from lowest to highest compressor cycle frequency, would be directly on the freezer wall/compressor coils with a fast response sensor for the highest cycle freq; to inside the largest thermal mass in the freezer with a slow response sensor for the lowest cycle freq. Temp control in a freezer, for thermally passive objects, generally has the exact opposite relationship with regard to sensor placement/response. This is mainly/only due to thermal inertia/overshoot.
Anything in between the extremes of sensor response/placement is a balancing act to fit whatever the priorities/preferences are for the system. Keep in mind that he more removed the indicator is from the object being controlled, the more problematic handling exceptional situations becomes. For active ferments, I can think of a few related to varying/specific ambient conditions that would cause issues.
You feel that your 1 cycle per hour freq is "pretty good". If you are happy with that, and feel/know that a lower cycle freq will cause too much temp swing (beer temps, bottle temps, etc.), then that is what is best for you. It is strange to me that your control settings function almost equally as sensor and fixed run time based (chosen by you). This is from the high number of "min on time" duration cycles in your data. Your sensor control appears to want to shut off earlier.
If you wanted to experiment, you could probably get very comparable beer temp stability, and a much lower compressor cycle freq, by placing the sensor in/on a small(ish) mass vessel. As long as your sensor vessel/damper is small enough, and the overshoot caused by your large freezer small enough (relatively), you still shouldn't get any frozen bottles.
For the multitasking you do with your freezer, like ramping (you lump that into "lagering") a 79F vessel down to mid 30sF, without freezing bottles in the same chamber, that gets into a lot of personal preference/prioritization. I would still not choose the sensor in air approach. Personally, in this case, I think there are better ways that give more control over the ramp rate, or lower cycle freq, or both. For lager ferments, the heat output is, I assume, so low that it behaves is very close to a passive object. That should allow an indirect indicator method for controlling the temp to function acceptably, and even to a tighter range than the (normal) controller allows (yours allows 0, but has to be mitigated with "min on time"). Only testing would tell.
For the average homebrewer fermenting an ale, sensor placement anywhere but in/on the ferm vessel will cause larger ferm vessel temp swings, or continual adjustments to the controller setpoint to compensate for using the ferm chamber air as the (secondary) indicator for ferm vessel temp. It will also cause a higher compressor cycle freq. For "hands off" start-to-finish ale ferment temp control, the current "best fit" sensor placement for keezer fermenters, according to consensus based on empirical validation of control theory, is on the wall of the ferm vessel.
motobrewer
I'm no atheist scientist, but...
so much text...
let me start off by saying this:
yes, my original post was misleading. What I really meant by it was, "my system probably cycles the most, and its only 2-4 minutes an hour. therefore you should be good". Also, what I said in that about ferm temps was wrong. In my head I was thinking about was ramping down a non-fermenting beer, not controlling actively fermenting beer.
i did, here:
the part where all our controllers are "micro-controller" boards (our supervisory controllers are actually small windows CE machines) and they run for years on end, in highly critical applications (air quality / smoke control for hospitals, for example).
I "lol'd" when you suggested using a PID to control a staged output. A "PID controller" isn't something you buy, it's a feedback control loop mechanism. Remember all your "fancy book learnin?" LaPlace transforms, transfer functions, etc? I'm not saying i'm good at it, control theory was a b*tch of a class. But anyway, the output of a PID calculation is an analog value, 0-100%. Period. That can get processed in a few different output signals on the electrical side, 0-10 volts, 0-1 volts, 4-20mA, etc. Send that to a valve with an actuator on it, and you have a variable flow system. Or feed that into a VFD and you have a variable fan. How are you gonna take a 0-10 volt signal and put it to a fridge compressor? How are you gonna take a 0-100% command and put it to a piece of equipment that's either on or off? The only way is to set limits - on at 75%, off at 50%, or whatever. Just like I stated before.
A PID is not the best control loop for a staged output (staged output is something that's either in one stage or the other. On/off, high/med/low, etc. Not like a valve where you can be 0% open, 100% open, 33% open, 33.42451% open, etc.). The fact that you thought "autotune" or "fuzzy logic" could make a "PID" (seems like you think a "PID" is a chip you buy) control a staged output is what makes me thing you really don't know all the "fancy book learnin" you mentioned before. maybe you do.
I think it's pretty good, do you? Is yours better? I really have no point of reference. You're right, the freezer seems to be held on at some instances by the min timer I selected. Most times it runs longer than 2 minutes tho. I felt the deadband created was adequate, and didn't want to turn down the timer because that would create more cycling. I didn't want to increase the timer because that would create a larger deadband.
oh, please lay off the '/'.
let me start off by saying this:
yes, my original post was misleading. What I really meant by it was, "my system probably cycles the most, and its only 2-4 minutes an hour. therefore you should be good". Also, what I said in that about ferm temps was wrong. In my head I was thinking about was ramping down a non-fermenting beer, not controlling actively fermenting beer.
i did, here:
moto said:
the part where all our controllers are "micro-controller" boards (our supervisory controllers are actually small windows CE machines) and they run for years on end, in highly critical applications (air quality / smoke control for hospitals, for example).
I "lol'd" when you suggested using a PID to control a staged output. A "PID controller" isn't something you buy, it's a feedback control loop mechanism. Remember all your "fancy book learnin?" LaPlace transforms, transfer functions, etc? I'm not saying i'm good at it, control theory was a b*tch of a class. But anyway, the output of a PID calculation is an analog value, 0-100%. Period. That can get processed in a few different output signals on the electrical side, 0-10 volts, 0-1 volts, 4-20mA, etc. Send that to a valve with an actuator on it, and you have a variable flow system. Or feed that into a VFD and you have a variable fan. How are you gonna take a 0-10 volt signal and put it to a fridge compressor? How are you gonna take a 0-100% command and put it to a piece of equipment that's either on or off? The only way is to set limits - on at 75%, off at 50%, or whatever. Just like I stated before.
A PID is not the best control loop for a staged output (staged output is something that's either in one stage or the other. On/off, high/med/low, etc. Not like a valve where you can be 0% open, 100% open, 33% open, 33.42451% open, etc.). The fact that you thought "autotune" or "fuzzy logic" could make a "PID" (seems like you think a "PID" is a chip you buy) control a staged output is what makes me thing you really don't know all the "fancy book learnin" you mentioned before. maybe you do.
I think it's pretty good, do you? Is yours better? I really have no point of reference. You're right, the freezer seems to be held on at some instances by the min timer I selected. Most times it runs longer than 2 minutes tho. I felt the deadband created was adequate, and didn't want to turn down the timer because that would create more cycling. I didn't want to increase the timer because that would create a larger deadband.
oh, please lay off the '/'.
deleted
So many accusations...
I figured as much after a few posts. It is unfortunate you don't exhibit the same zeal regarding cat22's post as you do mine, especially after initially appearing to validate his approach.
Your "deadband" is 0. Does that even count as a "band"? It is a band with zero width. Does it even exist? Dead-line may be a better term here. Unless you are talking about the realized deadband, and not the temp diff or "dead band" controller parameter. The compressor safety parameter for a fridge is usually for restart delay, not "on time". I am not sure if a short "on time", in and of itself, hurts a compressor in any way. Normal fridge controls have a large temp differential, so it usually won't shut off for a few minutes. edit: My understanding is that a short delay between cycles is especially hard on the compressor due to high pressure on startup. That may be the reason for the short cycle delay prevention. The wide temp diff is to prevent frequent cycling, since each start is hard on the compressor, regardless of initial pressure. I know its not good for the junky PTC start relays. They seem to have a fixed number of cycles to failure.
Those most likely have proper reset, error, and exception handling. A home baked board running freeware, or worse a PC with an MS product without all the crapware disabled, running my code with only minimal testing, validation, and fail-safing, is what I was referring to. I would have thought that would be fairly obvious, since I talked about coding my own loop.
You are getting really nit-picky here. Do I need some more /'s next time? I stated that a "PID controller" or even "PID" has become a generic term for a controller using a PID control loop, usually along with other control mechanisms as well. I also apologized, and said I should have been more specific. And, I also stated it was meant to be a joke about being to anal about temps, which I thought would have been caught the first time.
I try not to think about those, or Eigen values/vectors, Fourier, z transforms, determinants, and many others. This helps avoid flashbacks, and waking up an hour later on the floor, curled in a ball with my thumb in my mouth.
Ask your cubemate. I am sure he can fairly quickly think of a few ways to process the output to control overshoot in a fridge. I know I did. It's not like I said I could solve a 3 orbiting bodies system.
From what I have read, PIDs are good for controlling overshoot, especially when tuned for a given system. Valves are continuous, but the flow is not linear, that is why there are parameters to account for this. It may be possible to take advantage of those to control a fridge, although an intermediate circuit may be needed, and yes, it would not be linear (or is it continuous?). It has to turn something on and off at some point based on a linear (continuous?) signal. It would control overshoot, better than using a straight temp differential approach, while using a PID controller, which is all I ever claimed.
I was just using the names for the control options available on many PID based controllers. They are usually marketing terms to some extent, and usually have other methods under the hood. I am not certain.
I am not sure you CAN buy a "PID on a chip", but why wouldn't you be able to. Seems like a reasonable thing to fill a chip up with a PID's magic smoke, and sell it. Any software or microcode (and that is not just software for microcontrollers), can be burned into the chip.
I don't have DAQ capability, so do not have hard evidence. I have only used my ferm chamber for ale temps, so no comparable data to yours anyway. It may only cycle once a day when ambient is near the SP. I need to get my other controller wired up for my "new to me" and free keezer, after I get a new PTC start relay for it (casualty of excessive cycling from running empty, no doubt). Just using a standard fridge for serving now.
I guess by deadband, you mean the realized deadband, and not the deadband (temp differnetial) parameter, which is 0 in your case. You are using time based control whenever the min on time is in play. There is a name for that method, but I can't think of it.
If you read my previous post, I suggested that you might get acceptable temp control, no freezing, and less cycling, by damping the sensor. The amount of damping is limited, in your case, by the hysteresis caused by your keezer's thermal inertia, for lack of a better term. Something close to the same thermal response as the bottles of your beer that you are worried about freezing would be a good starting point. What would it hurt? You would know in a couple of hours where it is headed.
oh, please lay off the '/'.
If I don't specify everything (with caveats), you seem to call me out for generalizing or not including some piece of minutia, including in this post.
