PigFish

Recommended Posts

I have begun to delve a little into what I call micro analysis of humidor performance. I can't expect it to really excite anyone but me, but there is a certain amount of pleasure derived from making something work, making it work better and trying achieve perfection. It is interesting to see if an idea that you have will really turn into something that you will consider as a standardized design or whether it is a failure.

This is the same humidor on two different days with similar temperatures (outside). My goal was to try to shallow the dehydration during cooling above the 80 dF ambient mark and see if I can decrease the frequency or the amplitude of the cycle. Every improvement must do one or both of those things or it is not an improvement!

These are high magnification pics of the humidity oscillation due to refrigeration at a high temperature (ambient outside). Any of these very small dips are acceptable to me, but if I could pull them down to as little as 3 rH I would really be pleased. At 80 dF outside, I don't really think it is possible but it does not stop me from trying!

Test A is a native process. Cooling cycle begins and hydration follows. The cooling cycle is run until completion and there is residual drying that initiates subsequent hydration cycles. Pretty straight forward.

Test B on the other hand was an idea I had about interrupting the cooling cycle when the system called for hydration and then once the system was hydrated again the cooling cycle would resume. This would drag out the cooling cycle, and perhaps over cool, but it would theoretically reduce the large dips in the rH to smaller ones. In cooler weather the system would be unaffected by the design.

I am intrigued as to how it works and if I distort the chart to show the dehydration cycles highly compressed with the amplitudes exaggerated it looks like it does work quite well. I will crop some charts and run some analysis on them to look at the average rH, as well as average highs and lows to be sure.

post-79-0-57010800-1382123043_thumb.png

post-79-0-16044600-1382123054_thumb.jpg

Thanks for following. -R

Link to comment
Share on other sites

Not sure how much you know about statistics but what you are trying to achieve here is a lower standard deviation. If you have your data in a numerical form excel will easily be able to tell you if you're achieving levels closer to the mark. If you only have a graph then it can't help you though.

I often wonder about my winador, i have no control over anything, but i was wondering if it would be better to instead of have a temperature set point, have it on a timed cooling cycle. The idea in the grandest scale would be have a temperature gauge outside the unit, and based on that allow the fridge to cool for say 20 seconds and turn off for 1 minute, and repeat. The hotter it was outside, the longer the cooling cycle so to speak. Mechanically im sure it's terrible for the compressor etc but anyway. I think the temp would hold steady, but also steadily vary in small increments if you have wild swings in ambient temperature difference. I guess the same thing could be achieved if you could setup a controller that turned on heating at 19.9 degrees and turned off at 20.0, then have the cooling turn on at 20.0 and turn off at 19.9. Guarantee rock steady until your temp gauge dies and your cigars get ovened or frozen solid :P

Link to comment
Share on other sites

I understand and use standard deviation as an assessment all the time. You are correct in that it is a valuable tool.

Standard deviation (from the mean) alone is not the only answer however. Standard deviation of the mean does not solve our problem of controlling the mean itself, which should not simply be derived from the data collected but driven by the system itself.

Standard deviation from a set point, rather than a derived mean therefore is likely a better representation of data.

Still, standard deviation does nothing to differentiate frequency. It can potentially reflect high frequency low amplitude states as an equivalent value to low frequency high amplitude states. I prefer the former, but I am willing to modify my belief if I can find further statistical tools that would prove otherwise. One must of course look somewhat beyond pure statists however, in that tobacco is a physical substance with real hygroscopic properties and real hysteresis characteristics driven by both temperature and rH.

I must admit that I am no longer an expert in statistically analysis. It has been many years since I was fluent in statistics and perhaps I should again study up on the topic.

If you have suggestions on bearing out the quality of a waveform measured against a constant I would love to hear it. I am never beyond a good suggestion!

As far as your humidor is concerned, while there may be some merit to your ideas of external sensors and timers, I would suggest that assessing your internal environment, and controlling your compressor and forcing it into a cycle based on actual changes that will be reflected in its internal temperature is a better method.

While a more complex system certainly could gain from factoring the outside condtions, including establishing a trend line, your timer method would not accomplish that. While your system might compensate for outside temperature, you would need to create an algorithm that would adjust your clock based on it and presumably the acceleration of heat transfer based on increasing temperatures. On the other hand, using the hysteresis of your humidor to follow the conditions outside will do exactly what you want without the algorithm. You may suggest that one is a leading function while one if following but whether temperature data is collected either in or out, a change must be present to initiate the function of change of cooling.

Timing functions as they relate to humidor functions are quite useful, but I am not going to go into that here. Using your systems hysteresis and measuring changes inside saves you the work of attempting to derive the logic of heat gain from the outside world and then adjusting the clock on your timer accordingly. You will still posses the same problems by the way! You will need to decide on amplitude over frequency or vise versa.

While I won’t say that it won’t work, I will say that your theory, even if only whimsical, has eliminated most of the available tools currently available to you. There are proven tools made to keep temperatures constant and they work. It therefore does not behoove you really to choose a system more complex that is based on a theory rather than pragmatism!

Thinking provides for innovation and I understand the motivation. Thanks for sharing your thoughts.

Cheers! –the Pig

Link to comment
Share on other sites

Oh, by the way, studying time states of appliances in your humidor is very important (as I see it). While observation can provide some of the data there is nothing like the visual representation of the actual affects of your humidor appliances in action.

I did not feel like deriving a new chart but this is one used to evaluate the dehydration process. I won't go into all that I was studying here but the block cycles show the operation of the cooler and the humidifier to stabilize the system.

post-79-0-91682400-1382293182_thumb.jpg

If you look closely you will see that the on/off tags are inverted, that is a software issue!

Cheers! -pig.gif

Link to comment
Share on other sites

I dont know i understand the standard deviation from the setpoint thing. If your mean is not the same as your setpoint, then does that not suggest your controller isn't actually controlling to the setpoint you desire? Or is that what you were implying? That you use it to see if first your controller is averaging out to the setpoint and second how much it is deviating from that setpoint?

Unfortunately i use statisitcs primarily for comparative testing purposes, setting up randomised plant performance trials and analysing results etc. I don't do anything with cyclical analysis so can't help you out with figuring out waveform functions, sorry bro! Besides, i failed maths 4 times at uni so you probably don't want to take maths advice from me tongue.png

Another thing i was thinking of, while im blurting out stupid ideas. Have you ever ran a system in which the humidifier and the dehumidfier run non stop? Ie, the humidifier keep on pumping out humidity, and the dehumidifier keeps on removing it... that system should reach a steady state at some point should it not? Yes the cooling cycle will effect the humidity as well and i have no idea how it can all integrate, but just a crazy idea. For the record im an extractive metallurgist, im all about process plants and f#@$ing with **** to increase production. I have certainly had my fair share of epic failures though potty.gif

Link to comment
Share on other sites

Oh Ray, you had me at micro analysis.......love.gifwub.png

Link to comment
Share on other sites

I dont know i understand the standard deviation from the setpoint thing. If your mean is not the same as your setpoint, then does that not suggest your controller isn't actually controlling to the setpoint you desire? Or is that what you were implying? That you use it to see if first your controller is averaging out to the setpoint and second how much it is deviating from that setpoint?

Unfortunately i use statisitcs primarily for comparative testing purposes, setting up randomised plant performance trials and analysing results etc. I don't do anything with cyclical analysis so can't help you out with figuring out waveform functions, sorry bro! Besides, i failed maths 4 times at uni so you probably don't want to take maths advice from me tongue.png

Another thing i was thinking of, while im blurting out stupid ideas. Have you ever ran a system in which the humidifier and the dehumidfier run non stop? Ie, the humidifier keep on pumping out humidity, and the dehumidifier keeps on removing it... that system should reach a steady state at some point should it not? Yes the cooling cycle will effect the humidity as well and i have no idea how it can all integrate, but just a crazy idea. For the record im an extractive metallurgist, im all about process plants and f#@$ing with **** to increase production. I have certainly had my fair share of epic failures though potty.gif

There are many ways to skin a cat, but these days I don't chase too many of the stray ones!

There are several items worth considering while testing. One, does it work! Two, how well does it work and three, how accurate is it?

The data logger and the controller are two separate systems. Most sensitive sensors, placed even inches apart will have different readings. So, some of what is demonstrated here is sensor placement, and variation of instrumentation and set point analysis. This is ancillary and I got dragged into it a little by your question and comments about std. dev. This exercise is about cyclical function and control of the wave length and the amplitude. I brought up the set point because a lot of people read these threads and I did not want to leave that detail on the table without talking some about it. The set point is your desired state. I try to pull that to the mean but frankly I am not really working on that here. This is a quality of waveform analysis. I can push the waveform up or down the scale with the set point and with activation and differential logic adjustments. Those will (or can) also have an effect on the quality of the wave form. On the other hand the wave form must not really stray way off the set point! Otherwise, like you say, you are not really controlling off of one. That was my original point in bringing it up. I cannot just settle of the lowest standard deviation. There are other factors, and that closes the loop on the discussion for me.

Because my systems work in 4 control states, I must work within an envelope of performance. Every system works within an envelope but some allow the bottom or the top to vary with environment. Your typical active humidification system controls only low humidity state. Your typical active cooled and humidified humidor controls only two states, low humidity and high temp, and so on! But once temperature control is attempted, if the cold state, the heat sink is introduced into the humidor (a practical location for it) then it will also affect the state of humidification as a matter of design convenience. You can of course heat and cool your whole room! This is outside humidor control and it is a valid and very stable method. The best method actually! If people had the space and the ability and the budget to arrange it, then they don't need a sophisticated enclosure that manages it for them.

Once you establish that you wish to "envelope" your environment, you need to decide how small the envelope should be and can be. The size of it depends on equipment, design, programming and of course your outside environment. Insulation is a key part of the puzzle. Most people want to store cigars, not insulation and have a reasonable home environment. There has to be assumptions made to make the system work the best it can within a finite realm of exterior conditions.

I like to teach and share my analysis, I like to explain it, to demystify the process and introduce reality and dispel myths. Thanks for following and commenting!

Cheers! -R

Link to comment
Share on other sites

I have begun to delve a little into what I call micro analysis of humidor performance. I can't expect it to really excite anyone but me, but there is a certain amount of pleasure derived from making something work, making it work better and trying achieve perfection. It is interesting to see if an idea that you have will really turn into something that you will consider as a standardized design or whether it is a failure.

This is the same humidor on two different days with similar temperatures (outside). My goal was to try to shallow the dehydration during cooling above the 80 dF ambient mark and see if I can decrease the frequency or the amplitude of the cycle. Every improvement must do one or both of those things or it is not an improvement!

These are high magnification pics of the humidity oscillation due to refrigeration at a high temperature (ambient outside). Any of these very small dips are acceptable to me, but if I could pull them down to as little as 3 rH I would really be pleased. At 80 dF outside, I don't really think it is possible but it does not stop me from trying!

Test A is a native process. Cooling cycle begins and hydration follows. The cooling cycle is run until completion and there is residual drying that initiates subsequent hydration cycles. Pretty straight forward.

Test B on the other hand was an idea I had about interrupting the cooling cycle when the system called for hydration and then once the system was hydrated again the cooling cycle would resume. This would drag out the cooling cycle, and perhaps over cool, but it would theoretically reduce the large dips in the rH to smaller ones. In cooler weather the system would be unaffected by the design.

I am intrigued as to how it works and if I distort the chart to show the dehydration cycles highly compressed with the amplitudes exaggerated it looks like it does work quite well. I will crop some charts and run some analysis on them to look at the average rH, as well as average highs and lows to be sure.

post-79-0-57010800-1382123043_thumb.png

post-79-0-16044600-1382123054_thumb.jpg

Thanks for following. -R

Haven't had time to give this a good perusal ,but I will when I get 10 mins .

But more to the point ,it's good to see you back posting my friend

Cheers

Steve

Link to comment
Share on other sites

  • 1 month later...

Any idea on if these micro variations are propagating themselves across the entire Winador environment? If you store your cigars untubed/unwrapped on shelves in the humi then they will obviously feel the micro variations. If you store your cigars in the humi in their tubes or boxes, then that packaging is likely substantial enough to blunt this variability: much like your example of cooling the entire room in order to make the inside of the humi more stable. In this case it's the addition of boxes within the humi that will likely temper the variations. Got something small enough to put inside a cigar box surrounded by cigars?

And, besides trying to solve this by controlling the cooling cycle, might it be possible to make the humidification cycle more adaptive? For instance, using Boveda packs may only give you a constant rate of absorption or emission (I don't know that for a fact, maybe Boveda's absorb and emit on a curve), whereas what would be useful here is something that emits at a faster rate at lower humidification levels. If the dehumidification upon cooling cycle was predictable, might it be possible to modify a cigar oasis or such to turn on for a certain amount every time the cooling cycle triggered? The intent would not be to run the modified oasis to get back to target Rh, but just to 'take the edge off' the deepest part of the dehydration cycle.

Link to comment
Share on other sites

Any idea on if these micro variations are propagating themselves across the entire Winador environment? If you store your cigars untubed/unwrapped on shelves in the humi then they will obviously feel the micro variations. If you store your cigars in the humi in their tubes or boxes, then that packaging is likely substantial enough to blunt this variability: much like your example of cooling the entire room in order to make the inside of the humi more stable. In this case it's the addition of boxes within the humi that will likely temper the variations. Got something small enough to put inside a cigar box surrounded by cigars?

And, besides trying to solve this by controlling the cooling cycle, might it be possible to make the humidification cycle more adaptive? For instance, using Boveda packs may only give you a constant rate of absorption or emission (I don't know that for a fact, maybe Boveda's absorb and emit on a curve), whereas what would be useful here is something that emits at a faster rate at lower humidification levels. If the dehumidification upon cooling cycle was predictable, might it be possible to modify a cigar oasis or such to turn on for a certain amount every time the cooling cycle triggered? The intent would not be to run the modified oasis to get back to target Rh, but just to 'take the edge off' the deepest part of the dehydration cycle.

Yes, in a fast responding and well circulated humidor, the entire humidor feels even the smallest moves. This is a double edged sword. What I mean by this is a localized stable point is actually a partial if not absolute failure. Let me explain!

If 90% of your humidor was absolutely stable then it is of course a win! This is not the way it works however. In most cases, simply due to the laws of physics all humidors except the very smallest have energy/temperature and water density issues as you travel from low to high or vice versa. You can't beat physics, you can only compensate and design around your knowledge of it.

Before I get into a tailspin about this, one must understand that this is a "micro" analysis. I use super sensitive precision equipment. These small effects will have absolutely no affect on your cigars. Why? Here again we come up with the real world. The real world defines the rate at which tobacco responds to changes in nature and by empirically establishing its isothermal curve we can visualize that characteristic for the temperature in question. With the isotherm, then comes the hysteresis of the substance as it responds to the delta between the ambient and the isotherm (assuming it is at equilibrium in the first place).

Look at hysteresis as holding your breath, or perhaps better yet, like cooking a meal. Short period cycles are not really going to hurt you nor cook your meal. I mean if you leave your steak on the counter for an hour does it cook itself? Of course not. You need to apply energy to your steak in sufficient quantity and duration to affect its character and moisture content, its well-doneness and internal temperature.

These changes that I have charted are akin to taking your cold steak out of the fridge, say at 40dF and moving it to the counter at say 42dF. Are you risking your steak by doing so? Of course not. Nor are your risking your cigars buy allowing them to be exposed to a nominally lower or higher temperature or humidity. If these nominal minute changes were negative to your cigars, imagine the damage that would result from shipping cigars from location to location, state to state or country to country. If cigars were truly that anemic, there would be no online vendors, now would there? Shipping cigars represents potentially and generally huge climatic changes for your cigars. Cigars survive those types of conditions just fine.

The goal as a humidor maker is perfection. The real world gets in the way of that goal, but I get damn close!

So lets move from the micro to the macro. I have taken the data from the same chart used for the micro analysis and took it to the big picture. In the chart you will see 4 lines. The dark blue and the black are the humidor lines. The light blue and the red are lab.

post-79-0-44688600-1386300595_thumb.png

These peaks and valleys in the micro chart are minor deviations on a linear scale macro chart. The question is then, where do you want your cigars; in the lab or in the humidor? Do you even want your desktop humidor in my lab? Probably not! Further examination will show that the largest deviation are actually from dehydration and not a cooling cycle.

What you will also notice is that I am not attempting to force the lab to make the humidor look good. If I took this humidor and put it in my office at 70 to 74F range, the movement of the rH and Temp lines are reduced to white noise. The lab here ranges from the low 60's to the high 70's and the humidor takes those moves and stabilizes the cigar space to near perfection. You can't walk through your house and not have the rH range several points as you move from one room to the next.

Observation and response are the responsibilities of the controller. None of this is passive. Your Boveda packs (your example) would allow drastic dehydration during cooling and the rehydration of the humidor would take at least an hour. I know, I have tested it! If I kill the humidifier, you would see that with the silo running (that is my bead pack, also active) along with the natural diffusion from the humidor contents, it would take about 30 minutes to move the rH back to these levels. Frankly, partial active/passive systems are a complete failure at my level! A complete air exchange from opening the door, can require a full 24 hours from your passive system to recover. Mine will recover in a few minutes…

Boveda packs, while I understand are just your example are no help. You need an active humidifier to get anywhere near my results. Frankly, if I added a second, I could get even faster response times. The point is, we need to get cigars in there somewhere and try to keep the costs low. Each additional appliance takes room and costs money. It is all a trade off.

With no offense to the cigar Oasis people, I would not use one and they would not work for me. You cannot put a sensor in your humidifier and expect it to reflect your humidor as a whole. They sell these things because they are better than a passive system. At my level they are primitive at best. Cost effective yes, precision no! They work, but they are not at all precision control or delivery systems. I doubt their differential logic settings are less than 2% rH.

The Oasis has its place based on cost and budget. I won't go on and on about them because it is not my style. Lets just say, if I removed my humidifier and used an Oasis, it would ruin my chart!!! -LOL

I have written a piece about activation logic. You should read it to understand just how control is achieved via active controllers and that includes the Oasis.

Look, for the money they are better than nothing! Put one in an active cooled humidor, one without dehydration and you have got 70-75-80rH in there in about an hour if the cooler does not cycle. That is not control… If you really want results you need a unified solution, not a grab bag of stuff that might work in a passive humidor.

Thanks for the comments and for posting your thoughts. You are using your brain and that is what these threads are about.

Cheers, Ray

Link to comment
Share on other sites

Hello Pig:

Quick question for you.

I have two hygrometer both digital. Did salt test 4hrs and both were accurate off by 1 rh.

When I put them both back into my humi their is a 5 point difference. Very confused.

Which reading souls I use?

Link to comment
Share on other sites

Hello Pig:

Quick question for you.

I have two hygrometer both digital. Did salt test 4hrs and both were accurate off by 1 rh.

When I put them both back into my humi their is a 5 point difference. Very confused.

Which reading souls I use?

What kind of salt did you use? -LOL

While this is a bit tongue in cheek, it should be known that 'salt tests' as in a means to calibrate instrumentation requires not only a pure salt, but a constant temperature. Saturated salt solution tables are available all over the net. Ammonium nitrate and sodium nitrate gets you closest to 65rH @ 20C according to my studies. If they are not pure they are not reliable.

In all seriousness, they are both likely right, or perhaps equally wrong! I have written extensive opinions on the topic of accuracy in instrumentation and how little it really means. You must understand as well, that one point calibration means little or nothing. I have written on this extensively on this forum.

No one I know, and this includes me, uses a chilled mirror hygrometer. Those are accurate! Without such accuracy we only have our limited accuracy instruments and our own tastes. While I have a number of NIST certified sensors and I use them to calibrate my instruments and seek the best accuracy I can afford, it really is your taste and judgement on the condition of your cigars that really matters.

About your hygrometers. Welcome to the world of humidor climatology. There really is no uniformly stable humidor. The laws of physics are ever-present, moving the air, water and energy around in your humidor. All you can do is push it around and blend it up again! I think I build some of the best humidors around, and they are not perfect! The second you stop mixing, just like salad dressing, water vapor and heat begin to stratify!

If I were guessing I would say that your hygrometers are separated by height. A hygrometer requires a temperature datum to reflect your humidity. So therefore you have two effects working on your difference in readings. You not only have a measurable difference in water vapor density (rH) but likely have a measurable difference in energy (temperature).

If I break out my infrared thermometer and run it up the outside of one of my humidors, or even a filing cabinet in my office it will likely read 1 to 2dF difference from the bottom to the top! That's physics for you!

Hygrometers are also sensitive to where you place them in that if you place them on or near hygroscopic substances they tend to read the water concentration nearest to their sensor. If they are not placed in a gentle draft of moving air for temperature and water vapor for rH, they are not reading the environment, they are reading the nearest hygroscopic substance, or the nearest source of water vapor.

You cannot assume that your hygrometers are at fault. This is likely a humidor design issue and that encompasses a lot of facets.

Hope that helps! -PIggy

Link to comment
Share on other sites

Thanks you.

This is driving me crazy.

They are both next to each not near the fan system.

I just don't want to much humidity. I like to keep it at 68. Ine reads 71-72 and the other reads 65-67.

Just crazy.

I an crazy for perfection.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.

Community Software by Invision Power Services, Inc.