Dehumidifier controlled with humidity sensor

Hello all and thanks in advance.

I have a fairly large dehumidifier that was used to control the humidity level in a crawl space below my dads house. The control board has apparently gone out. It gives an error code, but the compressor/fan will not come on no matter what is pressed. I have bypassed the circuit board and checked the compressor/fan and they both work... Here is the label on the unit:

This company offered a 5 year warranty on the units when it was purchased, but
A. I am fairly certain it has been just over the 5 year mark sense it was purchased (lucky him) and
B. The company has sense gone out of business.

My questions:
A. I am assuming with no documentation/schematic on the board, it would be very difficult to repair the current board?

B. Would it be practical to use an arduino, some form of a humidity sensor, and a relay to have the unit (fan and compressor) turn on and remove the humidity from the crawl space when the sensor is above a set level? It doesn't have to be fancy just on if its above x and off if its below x. If this sounds reasonable, what parts would you suggest to use to do this (ex. sensors, relays, etc)?


this is a typical control application to could be processed by an Arduino.
There are a lot of humidty sensors and relays on the market available.
You schould make a analyzis for the selection wrt to technical needs and environmental conditions and requierements.
Beside this analyzis you can start with the design of the sketch.
If you have no experience with system design the study of the IPO-model is highly recommended.
Have a nice day and enjoy designing and coding.

The board may be fixable, but you'd need someone with a scope and some expertise to spend time on it. You might find that one of the replacement parts sites carries one instead.

The Arduino solution would be relatively simple I would think, although I would be reluctant to do it as the device is hidden away in a crawl space - your insurer would not be impressed if your home-brew device caused a fire.

I expect too that there are considerations about not short cycling the compressor and there may be requirements to run the fan for a while after the compressor is done. Probably temperature constraints too. Hard to know unless the manual is unusually thorough.

I could do it easily enough, but I wouldn't if it were my house.

I have done quite a bit of searching for a replacement board, and I have been unable to find anything! Here is the board that I am assuming is faulty. I don't have a scope and I definitely don't have the expertise :rofl:!

I don't know that he will want to use this dehumidifier for his crawl space any more, but we don't like to throw things away either. I want to make use of the device in some way or another.... As you know @wildbill, I have a sawmill, and the dehumidifier could be used in the future as part of a drying kiln... Therefore, I want to "fix" it one way or the other.

I would assume having the fan run for a while after the compressor shuts off would not be an issue, and definitely would not hurt. I don't know that I can even find a manual for this, and if I did not sure what I would be looking for...


My first Arduino project was a greenhouse sensor system. My dad wanted to track the temperature out there so he would know that freezing conditions were over. It reported data over wifi which I stored on a server at my house.

It never worked because of all the brick walls between his router and the greenhouse but the mark three is still operational in my dining room.

For your first attempt, I would try a DHT22 and an optoisolated relay board. You already know how to bypass the controller apparently. It looks like there are four relays there though. Can you tell why?

What is that?

I’m sure you already know this, but I have used ubiquiti nanostations (mine are the old versions loco m2) to serve as a wireless bridge for internet access. I have had great success with them. I run one over 400’ through tree branches and get pretty decent speeds…. I’m sure there are others, but you can pick them up the ones I use pretty cheap used on eBay. If that is the problem with your green house project.

Anyway, there are actually five relays. One is covered with a sticker in the picture. They are labeled:

F. Pump
And fan.

My dad had already disconnected some stuff so I’m not sure that it was all connected, I would have to check with him. I don’t think anything was connected to r.g.a.

I am wondering if the two pump connections are to run the condenser in reverse for defrost?
Not sure what e.v. Is? Evaporator?
Or r.g.a?

Edit: one relay is hid under the large transformer

Just as with your bandsaw project, the greenhouse project went through multiple iterations. I used several different x-bee variants after the wifi trying to get it going, but the construction of the house combined with the fact that it's three thousand miles away meant I never solved it. I'll bear the ubiquiti nanostations in mind, but my dad moved recently so there is no need to get it going now.

The current version (that I'm estimating to be the mark (version) three) has a Honeywell temp/humidity sensor and runs on a Pi because my Arduino wifi shield can't handle the newer encryption protocol that my router uses.

I forgot I was talking to a computer coder when I suggested the ubiquiti :rofl:

The compressor will probably be the largest relay.

It probably has a start capacitor and maybe a run capacitor.

Other than that it could be plugged in and rin.

You MUST let it rest for at least 10 minutes before the next start.

When it runs. The fan should run.

When it stops the coil will drain. The condensate pump run any time by a float switch.

If the compressor is good.
A relay of proper size is needed.
Easy to control with an arduino.

It does have (what I would call a start capacitor). This is how I bypassed the circuit board on it. Ran power directly to it and the fan.

I didn’t think about that. But most thermostats on ac units and I guess dehumidifiers too since they have same basic principles as an ac unit have a “lock” out feature. Where nothing will happen for a set amount of time to prevent damage.

When I start to try and write code with this, I assume it would be better to use millis for this lock out feature instead of delay?

So, if humidity level > 50% and 10 minutes has passed since the unit was last on. Then turn on compressor (relay) and fan (relay).

I don’t think this unit has a condensate pump. I’m Not looking at it at the moment, but I think the water produced drains via gravity and a hose. But I will need to check.

In regards to sizing of relay, the specs in the image at top of thread show 4.2 amps at 120v. So, I need a relay with the high side higher than this rating?Obviously, the relays logic side needs to be 5v to be controlled by arduino.

Solid stat relay? Or does that matter?


I'd go with SSR - it should last longer than anything mechanical.

Ok, you suggested earlier a optoisolated relay board. What does the optoisolated mean? The low side (5v) is separate or isolated from the high side 120v?

The relay isolates the 5V logic from the 120V side - that's its purpose.

However, the relay coils themselves can cause a voltage spike, particularly when they are turned off. In consequence many boards add an additional layer of protection so that the 5V side is isolated from the relay coils too. That's where optoisolation comes in - there's a chip that contains an led and a phototransistor. The net effect is that you can use it to pass a signal across without the two sides being electrically connected al all.

in the case of relay modules, 'opto-isolated' is like 'new and improved' soap.
it means nothing for what the words mean.

the optoisolator is only there to allow a very low current signal switch the relays. it absolutely means nothing about isolation.

using an opto-isolator allows for the use of low voltage signals from any device. however it is connected to the power supply and it eliminates the isolation application.

the existing board you posted a photo and you list 2 pumps.
I can only guess that there is an optional pump that can pump the condensate out to a drain. it is silly to gravity-drain into the ground or even a bucket around where your machine is.

if the pump relay is not used, that is fine.

the fundamental principal is heat transfer from air to refrigerant and from refrigerant to air. the fan moves air both into the coil and away from the coil.

so, compressor and fan are the core working part of the system.

get them right and all the rest are support systems.

the requirement for the delay is simple and goes to how DX or Direct Expansion regrigeration works.

the compressor increases pressure and the refrigerat has a PHASE CHANGE. from vapor to liquid
the liquid is moved past the condensor coil and heat is pushed out.
the liquid moves to the expansin valve or capilary tubing. as it passes the valve or thru the tuing, it expands, evaporates and the cooling occurs. the liquid turns back to vapor.

the compressor does it's job, compresses the vapor to liquid.
when you shut off the system, it is in a state of pressure extremes and one of the fundamental laws of physics is equilibrium.
the high pressure moves into the low pressure and the system at rest equalizes.
when the compressor starts it compresses vapor.
if you short-cycle, the compressor can have liquid. since you cannot compress liquid, the compressor cannot start and the compressor will have a locked rotor and either trip a breaker or burn up it's coil.
for such small devices, there is no protection and the compressor typically burns up. it only takes a few seconds before the system is damaged.

put a hefty delay in Startup(), 3 minutes maybe.
a loss of power and instant re-start has to be allowed for.

Most of the relay boards you'll see from the major suppliers actually are opto-isolated. The difficulty is that for reasons of convenience, they allow you to use a single power supply to run both the relays and the driver electronics. If you do, the onto protection is defeated.

Thinking of getting these:
HiLetgo 2pcs SSR-25DA 25A 250V Solid State Relay Module

I assume this will work for my application.

So, if you use the convenience of the powering both the relay and driver with same power supply and something goes wrong, it burns up the microcontroller (arduino in this case)?

It could, perhaps. The more usual symptom is that the board reboots. Or crashes.

The reason you need to make a practice of completely avoiding "delay()" is that you generally wish to attend to more than one matter at a time and even if you do not to start with, you are likely to wish to add additional functionality later. If you code using "delay()" to start with, you generally lock yourself out of improving the code.

Unless it was real cheap on Aliexpress or whatever! :grin:

Fair explanation.

Thorough nonsense! Or at least, badly expressed. :roll_eyes:

Slight improvement in the description. :face_with_raised_eyebrow:

Some of the available modules are designed such as to make the opto-isolators useless as both the input and output are connected to the same things. Most have separate connections for input and output circuits but confusingly have multiple supply connections and a a removable link which connects the two sides so that you can use a single supply and defeat any advantage of isolation.

Not at all. You will only damage the microcontroller if you make some major blunder in the circuit, nothing to do with the presence of an opto-isolator.

The more critical point is that you need to understand the requirement of keeping the wiring to the input side of the opto-isolator neatly separate and with control and return (whether "VCC" as on that diagram, or ground) travelling from one part to another as a bundle, and the supply to the relay similarly as a pair to the actual output connections of the power supply whether or not that supply is also powering the Arduino.

The problem arises from the impulse of switching the relay being capacitively or inductively coupled from the relay circuit back into the microcontroller logic circuits by loops in the wiring, or from the power supply if there is shared wiring from that supply to both relay circuit and microcontroller.

You might want to take a refresher course on transitors FETs and industrial signaling.

You completely missed the point.