By continuing to use this site, you agree to our use of cookies. Find out more
Forum sponsored by:
Forum sponsored by Allendale Dec 6th

Dehumidifier project

A small dehumidifier for eliminating condensation on my machines .

All Topics | Latest Posts

Search for:  in Thread Title in  
XD 35120/07/2019 06:20:58
1367 forum posts
118 photos

I have recently been working on a small dehumidifier for each of my machines and thought i might post some photos and general info on this project for anyone who is interested .

My workshop is not sealed at all - basically a wooden frame with fibro walls and roof and the eves are open so using a large dehumidifier to control condensation is out of the question .

I needed something that can be placed under a plastic cover to create a dry zone regardless of the humidity level inside the workshop . The unit needs to be automatic , easy to set up and economical to run - i will need 3 of these in total so i don't want something that chews power !

The original idea came to me after reading an excellent article by Silly old duffer in MEW issues 263 &265 and how he sorted out a condensation issue in his house . This article started me thinking about dehumidifiers and i started to search online for anything small enough to fit the bill but everything seemed to be either too big or expensive .

While looking around i noticed some cheap peltier cooling units like what is used in small coolers or food warmers like what you would use in your car but these were not exactly what i wanted but they did give me an idea !

The pictures below show one of the completed units , power supply and general set up.


A modified 550W ATX power supply from an old PC will supply enough power for all 3 units and cost nothing . A new unit is around $50 AU here in oz .

970cc01c-c286-4f4e-abcc-96780464d44b.jpeg The completed controller , 3D printed condenser housing and condenser .

On the controller the temp /humidity sensor is mounted on the left side of the case , the condenser temp sensor is the silver probe in the centre of the photo and the bed temp sensor is the black block with the white lead attached. Below is a shot of the controller screen showing the various readings and unit status . More to follow .d59a21c7-5778-43aa-a5df-21f95145944f.jpeg

XD 35120/07/2019 06:53:55
1367 forum posts
118 photos

Ok part two !

The controller uses an Arduino nano which does the grunt work and switches a Mosfet to power the peltier unit and a small fan on the condenser unit .6b8a7fb8-d6e5-4cc0-b4ba-b32f51c477f5.jpeg This shot shows the completed condenser - The general idea is the small fan (on the bottom of the unit under my hand ) forces air to flow across the cooled finned condenser so condensation forms on it and the water drips down into the hopper and is carried away via the clear plastic tube to a container on the ground. The alloy plate on the top is a heat sink and transfers the heat generated by the hot side of the peltier module into the lathe bed .


This photo is of the heat sink block and one with the peltier module sandwiched between the heat sink block and condenser (which is just an alloy heat sink scavenged from some old electronic device ) and the silicon heat transfer compound used to transfer the heat and cold from the peltier module.398e8c2d-f799-47fb-9641-9ee4ac8dbf0b.jpeg

This photo is of the unit set up on my lathe and the white cable is the bed temp sensor - The reason for transferring the heat into the bed is i just don't like wasting energy and the peltier module has to be kept under a certain temperature else it will die !68d018f9-a3dd-4431-861d-6bc397af1c77.jpeg

I used some cheap plastic painters drip sheets to create a cocoon but will make a better fitting cover from heavier plastic .5a856e8a-d848-4c97-a23a-630adbc5218e.jpeg

3D printing one of the condenser housings .


I will be using two of the data logging humidity/temperature sensors to test the units ability to lower the humidity inside the cocoon and i will post the graphs when i have them.

I must thank Silly old duffer - aka Dave for his assistance in getting the sensors i wanted to use up and running as my ability to code is still only amateurish at best !

AdrianR20/07/2019 11:50:35
272 forum posts
20 photos

Hi, I remember that article by Dave it was very interesting. It got me to thinking about my experiences living on a canal boat and trying to run a dehumidifier. In a nut shell the dehumidifier I had just wouldn't work when it was cold and damp. If it was above 20C it worked OK, my solution in the end was to stand the dehumidifier in front of the electric blow heater.

As my dehumidifier was a compressor based one, it had the issue that when the air temperature was too low, the evaporator would drop bellow 0C and would form ice. The dehumidifier would detect this and shut off the compressor till the ice melted, which meant most of the time the compressor was not running.

I think XD351 will not have too much problem with ice formation on the peltier in NSW unless it is unseasonably cold and the heat from the peltier will also help keep it all warm. Us in the UK wont be so lucky.

Since then I have learnt about a different dehumidifier type called adsorption or desiccant dehumidifier. These use zeolite which have the property that they adsorb water when cold and release it when hot. The dehumidifiers have a wheel that is coated in zeolite that slowly rotates. A segment of the wheel has hot air blown across it which dries the zeolite. The air is then recycled through a heat exchanger where the water condenses. The rest of the wheel has a fan that blows ambient air through it, where the water is adsorbed.

I just found a new (to me) item that may be of interest for the "under a cover" problem. It is a desiccant dehumidifier. that has to be recharged by plugging it in. £15 for two electriQ MD100

Could also save on the SO buying those disposable dehumidifiers.


SillyOldDuffer20/07/2019 17:38:35
4867 forum posts
1021 photos

Interesting development by XD351 of the software. My MEW article was about using an Arduino to measure and record temperature, humidity and air pressure with a view to tracing the causes of a condensation problem. As a side-note I mentioned the possibility that the same code could also control a dehumidifier, basically turning it on and off only when conditions made condensation likely.

Ian's project goes a couple of steps further. He uses his lathe to heat-sink the hot side of a Peltier plate so the warmth will reduce the chance of condensation on cool metal while the Peltier's cold side works to the same end by chilling the air whilst it removes water. The amount of energy needed is much reduced by putting the whole under a cover and developing the Arduino as an automatic controller. Ian's solution is a bit like the cold-aisle containment systems used in modern data centres: rather than air-condition entire computer rooms, money is saved by managing heat and cold intelligently.

This approach is particularly suited to places where local climate couples high humidity with large temperature swings. In some humid locations overnight temperature changes are severe enough to cause water to run inside buildings. Very difficult to protect machine tools from rapid rusting when naturally wet air condenses inside a workshop, but Ian has come up with a way of doing it without breaking the bank!


Samsaranda20/07/2019 18:18:07
794 forum posts
5 photos

Surprising how heat and condensation occur together, many years ago when I was working in Trucial Oman, now known as United Arab Emirates, at night during the summer the temperature could be 100 degrees Fahrenheit, this was at midnight, our accommodation had metal roofs and the metal caused water to condense out of the air and it used to run off the roofs as though it was raining. The temperature differential between the air and the metal of the roof could have only been a few degrees but enough to cause condensation to occur. XD 351 I like your approach to the problem and your fabrications are impressive.

Dave W

Neil Wyatt20/07/2019 19:47:23
16757 forum posts
689 photos
76 articles

Posted by XD 351 on 20/07/2019 06:53:55:

This shot shows the completed condenser - The general idea is the small fan (on the bottom of the unit under my hand ) forces air to flow across the cooled finned condenser so condensation forms on it and the water drips down into the hopper and is carried away via the clear plastic tube to a container on the ground. The alloy plate on the top is a heat sink and transfers the heat generated by the hot side of the peltier module into the lathe bed .


Now that's a good idea, a mini-dehumidifier using a peltier!

XD 35121/07/2019 01:02:03
1367 forum posts
118 photos

I have been using desiccant units for a few years now and while they do work any draughts will overpower them and i found the moisture indicator seemed to fail pretty quickly , one really annoying thing with the units i have is they fall over easily and i have destroyed a few of them when removing the cover from the machine as they toppled off and smashed on the ground - a thousand little desiccant balls went everywhere !

The dehumidifier unit has three seperate probes :

A SHT10 temperature and humidity sensor and two DS18B20 temperature sensors , one is for the bed temp and the other is in contact with the finned condenser . The software can read the temperature of the condenser and it has two parameters to stop the condenser from freezing , if the temperature of the condenser is below 5 deg centigrade it won’t turn on and if the unit is running and the condenser temperature drops to or below 1 deg centigrade it will turn off until the temperature is over 5deg centigrade.

The basic operation is if the dew point gets to 2 deg centigrade below the bed temp the unit will start and will continue to run until it has made the dew point drop to 6 deg centigrade below the bed temp , this gives a reasonable duty cycle and stops the unit cycling constantly if the dew point hovers around the threshold level of 2 deg below the bed temp.

When the unit is idle it draws 70 ma and when it is running 6 - 7 amps

not done it yet21/07/2019 07:55:24
3584 forum posts
15 photos
Posted by AdrianR on 20/07/2019 11:50:35:

I just found a new (to me) item that may be of interest for the "under a cover" problem. It is a desiccant dehumidifier. that has to be recharged by plugging it in. £15 for two electriQ MD100


This type of dehumidifier might be adequate for closed containers, but the vendors are clearly idiots or trying to ‘bull excrement’ any unsuspecting readers.

No such thing as “25W per hour”.

Only the maximum rate of water collection is quoted and no units indicated.

The dehumidifier weighs only 600g so I expect it actually contains 480ml of desiccant balls - with a density of rather less than unity. Presumably this is all silica gel and not some material coated in silica gel? Completely dry silica gel is able to absorb up to 40% of its weight, but, by then is totally ineffective as a desiccant.

They require the surrounding air to percolate the desiccant - not too efficient, but OK for a tightly enclosed volume, once dried, of course.

They quote the recharge time as about 10 hours (it may be rather more under less than ‘ideal’ conditions).  They may never be fully ‘recharged’.

Reading the ‘questions and answers’ should be rather enlightening, for that item.

Most drying instructions quote that heating to 110 degrees celsius is required to fully regenerate the silica gel. Amazon quote 120 Fahrenheit! Which do you believe?

Silica gel has been used for umpteen years as a desiccant in laboratory desiccators. These had lightly-greased ground-glass lids to effectively seal the desiccator and would withstand being evacuated of air to increase the effectiveness (no air, no water vapour!). They were rather slow to work, actually drying items, at atmospheric pressure but were OK, given time. We nearly always used them under vacuum conditions at work (covered with a cloth - just in case of implosion).

All in all, not a particularly satisfactory item. Only OK for totally sealed environments, IMO.

Nothing wrong with silica gel as a desiccant, but the item advertised is of very limited in use. Standing it on a shelf in an open room would, IMO, be a totally useless occupation - it may well soon weigh a bit more (easily assessed but not quoted in the specifications, of course), but would then stop working.

Not a patch on my dessicant dehumidifiers, but mine do consume rather more energy (the particular model I use can collect up to 8 litres of water per day).

I would suggest these items are of no more use than a chocolate teapot in most scenarios encountered/expected by unsuspecting purchasers - cheap trays of the the material, regularly re-activated by heating in an oven, would be a better option.

Edited By not done it yet on 21/07/2019 08:02:31

XD 35122/07/2019 04:27:27
1367 forum posts
118 photos

A few more pics for anyone who is interested .

Sub assembley

Peltier module fitted to the heat transfer block and condenser fitted into the housing.

Shot of fins

end view of housing showing the fins of the condenser and the hole for the DS18B20 temp probe for the condenser .

DS18b20 cable routing

DS18B20 probe fitted to the housing.


Some insulation to minimise heat transfer from the hot side to the cold side of the unit .

o rings

O rings fitted to the underside of the screw heads to give some elasticity so the peltier unit is not crushed , i just nip these up until the O ring starts to distort .

T nut

A crude T nut arrangement will facilitate fitting to the C2 lathe bed .

Fan hole

The base of the housing showing the hopper/funnel and hole where the fan (50mm ) is to be fitted.

Mounting to c2

The T nut slots int the bed and is rotated 90deg and the even cruder wing nut is tightened .

Fitted to bed

Fitted to the bed - the top steel piece is aligned to the bottom T nut so i can tell when it is in the correct position and then tighten the wing nut -that thing gets uglier every time i look at it !

2 data loggers and controller set

Two data loggers set up to track the humidity and temperature inside and outside of the plastic cover , i let all three stabilise before starting the recording .

I don't know why the last two images decided to rotate 90 deg - one day i will work out how to rotate them back !

On the top of the toolpost under the nut there is a set of needle roller bearings , this is one of the best modifications i could recommend someone does to their lathe - no more toolpost turning as you tighten it and the extra amount of clamping force you get because of the rolling element is exceptional .

Now i need to finish the units off and set up a power meter to check power usage while i let the data loggers do their job .

Circlip22/07/2019 09:38:41
985 forum posts

'Bout time the soft/hard ware sorted the upright version of picture posting angry 2

 Not sure if it is electronickery problems or laziness.


Regards Ian

Edited By Circlip on 22/07/2019 09:42:29

Edited By Circlip on 22/07/2019 09:43:12

XD 35130/07/2019 06:21:02
1367 forum posts
118 photos

Here are a few graphs from the two data loggers:

humidity  inside.jpeg Humidity under the cover above and outside below , the vertical dotted lines are 48 hrs apart .

humidity outside.jpeg

These are the ambient temperatures and as above the top one is from under the cover and the bottom one is outside .

temp inside.jpeg

temp outside.jpeg

Probably not the dramatic effect i was expecting but the wave form under the cover is a little more stable , i believe the sudden drop in humidity on the peaks of the top chart (almost like a saw tooth ) are from the dehumidifier running and this coincides with the highest humidity and lowest temperatures of each day . There is only a few degrees difference in temperature between the two charts but under the cover doesn't get as cold or as hot and the rate of rise is a little less . The humidity also stays lower at the peak (except the last day and as i have been checking the units every few days it may have been a sealing issue ). The dehumidifiers are designed to keep the environment under the cover at a level where condensation will not form - not to dry the air out completely and as the cover is only a plastic sheet draped over the machine i expected there to be some leakage.

The power usage looks to be around 0.4 KWh a day and at 13.2 cents a KWh it is costing 7 cents a day But i feel a large chunk of this cost is from the power supply fan as it is always running all be it only slowly so i will hook up another supply that has no fan and see what it yields, the power meter says it is using 21w of power at idle and considering the dehumidifier unit only draws 70 ma at idle or 8.4 w if i can get the idle time usage down to below 10w it should lessen the cost i bit .

Another modification i have made to the condenser housing is to add a duct on the exhaust end to carry the cold air coming out of the exhaust onto the heat transfer plate to aid in cooling .


Michael Gilligan30/07/2019 06:54:31
14283 forum posts
628 photos

That looks very encouaging ... A useful project; thanks for sharing it.

At the risk of stating the obvious; Your un-sealed cover is probably the biggest obstacle to success ... If there is any leakage path your little unit will be trying to de-humidify the World.

May I suggest that you run another test, on a completely sealed environment [maybe a big plastic storage box, with Plasticine as gasket under the lid]


Alan Charleston30/07/2019 07:02:00
76 forum posts
19 photos


I assume you want to dehumidify the atmosphere around your machines to prevent condensation and corrosion. If that is the case all you need do is to raise the temperature a few degrees hotter than the surrounding air to keep the relative humidity less tan 100%.

I used to run a steam generator which was sometimes hard to start due to condensation in the electrical control gear. I fixed it by putting a heated towel rail on the ground underneath it. It didn't take much electricity to run, was designed to run continuously and stopped water from condensing on the generator.



Blue Heeler30/07/2019 07:12:39
189 forum posts

Interesting project.

Douglas Johnston30/07/2019 08:49:44
642 forum posts
32 photos
Posted by AdrianR on 20/07/2019 11:50:35:

I just found a new (to me) item that may be of interest for the "under a cover" problem. It is a desiccant dehumidifier. that has to be recharged by plugging it in. £15 for two electriQ MD100

Could also save on the SO buying those disposable dehumidifiers.


I have been using a couple of these units for the past 5 years over the winter months in Scotland. I put one under a polythene cover on my lathe and another under the mill. I try to make the covers provide a reasonable seal at the bottom and find that they last about a week before needing to be recharged overnight. This costs very little to run and seems to work quite well.

As an aside you can get a large quantity of cheap silica gel for making your own units by buying silica gel cat litter.


not done it yet30/07/2019 09:43:35
3584 forum posts
15 photos

My question is: How much water did you collect for your 7 cents?

I use 0.375kWh for one hours running of my dehumidifier and collect a bit under 2 litres/week if run for one hour each day - say 250ml/hour of running time. Obviously much less when the workshop humidity is lower. At around a cost of 4p/day,

I usually run it for 2-4 hours per day in the winter - more to keep the temperature of the machines up a bit in the colder winter months. That keeps my approx 10-11square metre workshop (crammed with machines, cupboards, etc) condensation-free and just about at a workable temperature (a fan heater soon raises the air temperature in the winter) without need to cover anything.

With my workshop built inside a sectional concrete garage as a cocoon, with at least 100mm of insulation around and above, the workshop temperature lags far behind the outside, so warmer in the winter and cooler in the summer - as long as the door is kept closed.

XD 35130/07/2019 19:23:03
1367 forum posts
118 photos

It is hard to compare yours V mine as the circumstances are vastly different .

The cost of energy. In Australia is horrific and I don’t know what it is like where you are .

At todays conversion rate 4p = approx 7cents AU.

I’m using 0.4 KWh a day so 0.0167 KWh an hour or 2.8KWh a week = 27 cents a week or 15p in your money .Yours is 0.375KWh an hour and at 4 hrs a day = 1.5KWh a day or 10.5KWh a week =28p in your money .

The volume of the area under the covers ( at a guess ) would be no more than 0.07cubic metres .

I’m averaging 1- 1-1/2tablespoons a night from this area so 20 - 30 ml a night so if i divide 20 ml by 24 hrs i get 0.8333 ml an hour from 0.07 cubic metres .

If i work out your work space at 10 sq metres x 2 metres hight we get 50 cubic and divide that by 3 to allow for the contents we get 16.6 cubic and 250ml an hour from this . If we we divide 16.6 by 0.07 = 237. Now 250 ml divided by 237 = 1.054 ml an hour from your unit in an area equivalent to mine so not a lot of difference .

I don’t know how long my unit is actually running per night and i suspect that most of the electricity usage is from the power supply fan .

It is almost impossible to do a comparison as there are so many variables like climate , cost of energy , workshop size and construction and the fact my unit only runs when it is needed so i can set it and forget it .

One really bizarre thing is the lighter constructed machines like my drill press don’t suffer condensation issues , i have been out to the workshop to discover my lathe and mill dripping wet but the drill press which is no more than 10 feet away is bone dry ! That is another factor that makes comparing yours V mine difficult as the rate at which my workshop heats up after sunrise is more than likely vastly different to yours and this is probably the main contributing factor in causing condensation to form only on the heavier machines .


I will try the box idea that you mentioned but i will run the unit on manual to see how long it takes it to drop from a very high level of humidity , i should be able to set something up in my 100lt esky ( cooler ) by feeding the power cables in through the drain hole and seal them off with blutac or plasticine..


DMB30/07/2019 23:02:21
936 forum posts

Hi All,

I admire your efforts with electronic gadgetry and well done. However, I have taken a different approach as follows.

They charge 19.98p a day standing charge, 15.6647p a unit or KWH. Working out total charge ÷ useage, almost 20p a unit.

Now, I cover my Super7 and hefty lump bench drill, Fobco Star, with thin plastic sheet, no heat. Only my small mill seems susceptible to condensation, even covered in thick plastic so I keep heat on under the plastic, 24/7. My home made heater is a 3-ply box made to house 4 - off 60W bulbs in batten holders fixed in round holes in the box, in line and wired in series to reduce current and cost. Opposite side of box open adjacent to top of bulbs to let the heat out. This works very effectively. Only gentle warmth, can switch off and touch a bulb and it's only a bit warm. They glow, very low light output.

I am not sure how to work out the running cost but I am certain that it's very low.

Assuming 240V and 60W, A = 1/4, so 240 ÷ 1/4 =960 ohms. Four in series. Continuing calculations resulted in such a small useage for being on for 8760 hours a year and cost apparently so low that I doubt whether I have worked it out correctly. Could somebody please advise how to do this?

The rest of the 10 × 8 feet shed unheated and covered entirely with felt for waterproofing, together with internal insulation. I do have a small humidifier which works by a moving belt and fan over a metal sheet (?) and water collects in a tank at the bottom. Dont use that very often. Tools and vice and work on bench protected by thin plastic sheet and are OK.I keep an awful lot of small tools in the house so they keep rust free with the central heating.


Edited By DMB on 30/07/2019 23:06:45

XD 35102/08/2019 09:48:01
1367 forum posts
118 photos

Just a quick update :

I have changed the power supply out for one with no fan and the power usage went from 20 w down to a current draw of 0.09 amps when idle @ 12 v or 1.08 w . It appears that it used approximately 0.2 KWh over a 48hr period or around 1.5 cents a day .

Enough of the power thing for now - it is time to move on to efficiency and maximising that so now i need to seal it off properly .

All Topics | Latest Posts

Please login to post a reply.

Magazine Locator

Want the latest issue of Model Engineer or Model Engineers' Workshop? Use our magazine locator links to find your nearest stockist!

Find Model Engineer & Model Engineers' Workshop

Latest Forum Posts
Support Our Partners
Allendale Electronics
Eccentric July 5 2018
Tee London LMES 6th Dec
Eccentric Engineering
Subscription Offer

Latest "For Sale" Ads
Latest "Wanted" Ads
Get In Touch!

Do you want to contact the Model Engineer and Model Engineers' Workshop team?

You can contact us by phone, mail or email about the magazines including becoming a contributor, submitting reader's letters or making queries about articles. You can also get in touch about this website, advertising or other general issues.

Click THIS LINK for full contact details.

For subscription issues please see THIS LINK.

Digital Back Issues

Social Media online

'Like' us on Facebook
Follow us on Facebook

Follow us on Twitter
 Twitter Logo

Pin us on Pinterest