Compressor winador/humidor....feedback.

[Tuan Huynh]

Hello, 

so a few weeks ago I bought a compressor unit at a fairly cheap price. And initial ran unit empty, I was getting 62-66 dF and rH 80-20%.

when the compressor kicks on the rH drop dramatically. This data is from the original controller.

after I took it apart and add in my own controller.  I was able to get the temperature to 68.5-69 dF and the rH right now on average of 58%. Its fluctuating from 59-53% and once every 12 hours it would drop down to 51%.

note: the unit is still empty so there’s no packs or anything to raise the rH to correct level yet. 
 
so would  the fluctuation Ruin my cigars? 
i know ideally we don’t want any fluctuation but so far this is the closes i got to keep the rH from dropping dramatically. 
 

let me know what y’all think. 

[Subcomandante]

I would put some cigars in there and some bovedas or your preference and see  what happens. The cigars should help stabilize it as well as the bovedas. I have wineador and ran it empty and humidity was in the 80’s. It went down to 62-67 RH. The higher number on the top shelf and the lower number on the bottom. So I keep my NC’s on top and CC’s on the bottom with lower humidity. 

[PigFish]

If you want to really know what it takes to convert a compressor wine cooler to an actual, accurate and reliable, functioning precision humidor, find me on YouTube. CigarClimatology...

[Tuan Huynh]

30 minutes ago, PigFish said:

If you want to really know what it takes to convert a compressor wine cooler to an actual, accurate and reliable, functioning precision humidor, find me on YouTube. CigarClimatology...

Yes sir I’ve been watching all your videos. Still working out some tweeks in my setup so far. I just finish lining my humidor and seasoning the humidor to get it to 68-69rh. Once I get that done I’ll turn the system on and see what the fluctuation looks like. Before the cedar lining and the shelves, the rH would drop 6-8rh per hour due to the compressor kicking on. Any advice to fix that issue let me know if you can. 
 

thanks. 

[PigFish]

Post up some pics of your project.

If you have any form of circulation in your project, and your sensor is sensitive, you are only seeing a small part of the actual rH flux in your humidor. This is my guess anyway. If you want, I can go to my test box, turn off the humidifier, and give you a little data log of what happens inside, even a well designed humidor if you use active cooling.

If you are really only getting an 8rH flux with a homemade box, you are doing damn well. Unfortunately, you are not likely really seeing the whole picture. And to do that, you have to look more closely for faults or you are just not going to find them with poor instruments.

(post has all sorts of assumptions based on my previous experience with helping others debug humidors, no offense intended!)

Cheers! -Ray

[Tuan Huynh]

Ray, 

thanks for the reply,  here is what my unit looks like. As for controller I’m using a plc base. Have a type j thermocouple for temperature measurement. I have the temp probe mounted at the top Inside of the unit. For humidity monitor, I’m using a sensor push to get some kind of rH monitor/log. I can’t find a resonable rh sensor that puts out 4-20ma for my system. Let me know what more information you need. I also have a 4” dc controller fan for air circulation. 

[Tuan Huynh]

Here’s anotber

[Tuan Huynh]

Just now, Tuan Huynh said:

Here’s anotber

 

[mrmessy]

Nice. All of this craziness is why I just unplugged mine. I don't have problem with the ambient temperature and couldn't spend time messing around with the rh changes. So Boveda and a full humidor in an unplugged unit gives me stability.

Sent from my SM-N950U using Tapatalk

[Tuan Huynh]

48 minutes ago, mrmessy said:

Down here in New Orleans, during summer the ambient temp gets hot. 80-90s so around that time the cool air would be a bonus and keep the winador at a stable temperature. 

[PigFish]

Did you fab the false back? Looks like you have been watching my videos!

What you have here is very similar to early experiments that I did with making refrigerated humidors. You are on the right track.

When does the fan in the back run?

What do you do with the condensate?

What do you mean 'reasonable' for the sensor? I have a lot of old sensors around the place. Tell me what you think is reasonable and I may part with one.

Let me see what you are using for a logic in the controller. While I won't write it for you, I may be able to point you in the right direction.

Frankly all that shelving is not a very good idea, sorry! Circulation is more important than shelving. I would consider pulling it all out but one on the bottom, unless you have set a pattern of cyclical circulation around the inside of the box as a whole.

Here are some hints:

Regardless the cost (within reason) you need a fast acting sensor for rH. These will also generally respond faster than the TC, and you can get the temp data from the same sensor. Get the a good 4 to 20 or 0 to 10 whatever your system requires.

Stage the bottom fan that runs air to isolate the 'cold space.' Write code to run the fan intermittently as the rH drops. This will let you cooling system run longer and pull heat from the box but stop running based on an rH set point. This will slow the cooling process, which is just fine, and allow you to support the loss of water by better controlling your cooling via the cooler fan.

Trimming is what you want. This kills the big drops, big swings... Here is a data log from a few years ago that has been converted to a png. Small corrections mean smaller peaks and valleys.

Got your work ahead of you. Get to it...! Best of luck on your project. -R

Cheers! -R

[Tuan Huynh]

1 hour ago, PigFish said:

Did you fab the false back? Looks like you have been watching my videos!

What you have here is very similar to early experiments that I did with making refrigerated humidors. You are on the right track.

When does the fan in the back run?

What do you do with the condensate?

What do you mean 'reasonable' for the sensor? I have a lot of old sensors around the place. Tell me what you think is reasonable and I may part with one.

Let me see what you are using for a logic in the controller. While I won't write it for you, I may be able to point you in the right direction.

Frankly all that shelving is not a very good idea, sorry! Circulation is more important than shelving. I would consider pulling it all out but one on the bottom, unless you have set a pattern of cyclical circulation around the inside of the box as a whole.

Here are some hints:

Regardless the cost (within reason) you need a fast acting sensor for rH. These will also generally respond faster than the TC, and you can get the temp data from the same sensor. Get the a good 4 to 20 or 0 to 10 whatever your system requires.

Stage the bottom fan that runs air to isolate the 'cold space.' Write code to run the fan intermittently as the rH drops. This will let you cooling system run longer and pull heat from the box but stop running based on an rH set point. This will slow the cooling process, which is just fine, and allow you to support the loss of water by better controlling your cooling via the cooler fan.

Trimming is what you want. This kills the big drops, big swings... Here is a data log from a few years ago that has been converted to a png. Small corrections mean smaller peaks and valleys.

Got your work ahead of you. Get to it...! Best of luck on your project. -R

Cheers! -R

Great data you got there. Something I would like to achieve before putting cigars in the box. 

yes I’ve been watching your videos and use your idea of the back. Initial thought was to block the cold plate from drawing out all the rH but it didn’t matter. So I just left that there. 
 

the fan I haven’t program for it to run yet. 
I am using a Idec plc. Right now in short version, I’m only controlling the unit with tc temp reading. When hit hi set point (69df) the system turn the compressor relay on. After 1min and 30 sec the system turn the compressor relay off, have a 2 min grace period, if the temp still high then the system will kick on again for another 1.5min. And repeat this process. So far I’m able to achieve a 6-8 rH drop with this. I would love to get my hands on a 4-20ma temp/rH sensor. Let me know what it’s running for. I will turn on the fan while the compressor is on. Intermittently like you suggested and see how it go.  With this program the temp stays perfect within .2 dF difference. So I like that. Just the rh I’m worried about. Not sure how the sticks will act with that fluctuation. 

[PigFish]

The problem with what you are doing is that the compressor needs a dwell period after running. This is what is referred to as anti-short cycle. If you run the system, and don't let it sit long enough, and there is still head pressure on the compressor you will just continue to trip the thermal breaker on the compressor system. The thermal breaker takes longer to reset than a programmed anti-short cycle. Additionally, you will eventually wear out the thermal breaker and then wonder why the compressor does not start.

If you are ladder logic programming, you can look up a logic sequence for a simple logic delay on make breaker. if time since off is greater than 4 minutes, the compressor comes on, or it waits for the 4 minutes if a demand for cooling still exists. 4 minutes should be just fine. You can tinker with this and slowly shorten the 4 minute time. 4 minutes is pretty good for these compressors.

Timers for running the compressor is not a good idea. This is because you will work on an algorithm based on a certain ambient temp rather than feed back from the sensor. In other words, you will tune the system for an ambient, a set amount of heat gain in the cooler per unit of time. When the temperature changes, your tuned algorithm will show faults.

What will happen here is that if you are in a high enough ambient, this compressor cycle will likely never end. If your ambient is not far from the set point, and is never greater, your system will work until something changes. This is not an automated system. This is an accidental system, tuned for only one set of ambient conditions.

Cheers! -R

[Tuan Huynh]

28 minutes ago, PigFish said:

The problem with what you are doing is that the compressor needs a dwell period after running. This is what is referred to as anti-short cycle. If you run the system, and don't let it sit long enough, and there is still head pressure on the compressor you will just continue to trip the thermal breaker on the compressor system. The thermal breaker takes longer to reset than a programmed anti-short cycle. Additionally, you will eventually wear out the thermal breaker and then wonder why the compressor does not start.

If you are ladder logic programming, you can look up a logic sequence for a simple logic delay on make breaker. if time since off is greater than 4 minutes, the compressor comes on, or it waits for the 4 minutes if a demand for cooling still exists. 4 minutes should be just fine. You can tinker with this and slowly shorten the 4 minute time. 4 minutes is pretty good for these compressors.

Timers for running the compressor is not a good idea. This is because you will work on an algorithm based on a certain ambient temp rather than feed back from the sensor. In other words, you will tune the system for an ambient, a set amount of heat gain in the cooler per unit of time. When the temperature changes, your tuned algorithm will show faults.

What will happen here is that if you are in a high enough ambient, this compressor cycle will likely never end. If your ambient is not far from the set point, and is never greater, your system will work until something changes. This is not an automated system. This is an accidental system, tuned for only one set of ambient conditions.

Cheers! -R

So In short I should look into a logic, when the internal temp reach the high setting, I should energize the relay to turn the compressor on. And you state 4 min is roughly on time for the compressor? 
 

I think last time I had it set to 3 mins on time the compressor drop the rH by 20 percent. That’s why I figured shorten the time in theory shorten the rH lost.  But my theory on this could be way off. 

[Tuan Huynh]

Hey ray, what sensor you suggest to use for temp and humidity that qualifies as fast acting. The type j tc I’m use is pretty fast but not sure if any humidity. If you have some laying around that have 4-20ma let me know how much it would be for it. 
 

thanks 

[PigFish]

Where are you at Tuan?

No, you missed my point, however you have come up with another.

I short cycled compressors for many years for dehydration and ultimately don't think that it hurts them. I still have one of my original projects, it looks some like your project, I put it about generation 12. It is my test box that I have in a lot of videos. It short cycles for dehydration and that cooler is almost 15 years old. It has been through the ringer! It still goes strong.

I have burned up a few parts, mostly cooler relays... I digress.

No, what I mean is that the cooler must remain OFF for a certain period. You can not turn compressor bases systems on and off like a strobe. The head pressure in the system must fall so the compressor motor can start with no pressure on it. This is why AC systems can actually run so long and use little energy. If you needed a compressor that would start against head pressure, you would need a considerably larger motor. The motor and compressor in your system is no larger than your fist.

The reason you have so much rH loss is due to lack of isolation of your cold plate. You put a shield in front of it, that is the start, but better isolation is necessary to help you with the rH dip. This is not a controls but an engineering problem.

You can add an active humidifier to run while the cooling runs and battle against the low rH problem. That is one aspect of the way these systems work. They will not ever work well with passive systems. You will remove water much faster than you can put it back without an active humidifier. However once you have an active humidifier, you now need to be able to dehydrate or your system will always creep into a much higher rH domain.

Prepare yourself for finding pitfalls around each corner.... For now I will leave you be and let you have some fun experimenting. 

I will leave you with one more hint about all this. These people who sell these types of systems to people run them at 65F. There is a reason for that. It is because they expect that no one lives in a 65F home and they will always keep it heated above 65F. Why? Because in that way, they are assured that the system will always cycle... Even in winter time. With the cycle they get the dehydration that they need to call there system working. They are not working actually. They are accidental. A working system allows the owner to work around various temps and rH levels to suit himself. They cannot do it... Their systems would go over rH.

Have fun with your project!

Cheers! -the Pig