Engine Dehumidifier DIY

Huckster79

Pattern Altitude
Joined
Nov 22, 2018
Messages
2,322
Display Name

Display name:
Huckster79
I recently built an Engine Dehumidifier, very similar to commercially available unit called the Engine Saver. I believe the Engine Saver retails around $325, I have $40 into mine. The idea is our engines are humid regions from combustion gasses making their way into the crank case, and moisture is a necessary element for our arch enemy: corrosion. I have heard it debated if these are effective but my thought is for $40 I'll take my chances its a wasted $40... Im not the inventor of the DIY version, I believe its been floated around online from some EAA groups, and like any home built rig there are variations.

Basic concept is to pull air over a desiccant, usually silica gel to dry the air. There are open looped units that just pull ambient air over desiccant and pump it in the oil fill tube and let it vent out the vent tube. There's a closed loop systems that pump it in, but also captures it on its way out recirculating the same, or hopefully most of the same air back in. I went with a closed loop unit but am going to use it in both fashions. At shutdown I plan on hooking up the "In" and leaving it open loop as I pack up for the day and closing the loop before going. Thought being the crank is filled with nasty corrosive gasses at shutdown I'd rather just push those out and have them gone. But the closed loop system I believe has an advantage in that desiccant dehumidifying isn't mechanical dehumidifying and takes time, always drawing ambient air I don't believe gives the desiccant long enough to absorb the humidity.

Basic design is two rubbermaid tubs for "left overs" one has the largest aquarium pump I could buy, a 60 gallon unit. It was double outlet so I put a "T" in to combine it from the two tiny aquarium hoses into one old air compressor hose I had. I sealed the pump in the tub, siliconing the cord and hoses coming out. I placed a hygrometer that reads current RH and also historic High and Low RH in the tub. For the air draw into the pump tub I used another chunk of the air compressor hose and ran that to the same type of tub. In that tub I drilled a hole near the bottom and one near top. I put fibrous aquarium filter material over the air hose end going out the upper hold, and then same material on the bottom of container over top of the inlet hose so the silica wasn't tight around the inlet hose end. The cheapest way I have found to buy Silica is "Crystal Kitty Litter" its a pure silica gel kitty litter they sell at Wally World. Then filled that container with the silica.

On the outgoing air from the pump I put an oil/water/particulate filter for air compressor tools in the line to catch any silica dust, though undisturbed and low flow of air I doubt there will be much, but don't want to pump it into the Continental if there is... I ran the hose from the pump to a rubber drain pipe adapter I bought at hardware that was 1 1/2" to 2", bought a plug for the 1 1/2" side and drilled it out for the hose. So to install I remove the dipstick and put this over the oil fill and a couple quicks twists on the attached hose clamp clamps it down to the oil fill tube. On the hose for the air inlet to the silica I run that to the vent tube and by chance it fits snuggly right in it.

One of the EAA designs built in a light bulb to "dry off" the Silica periodically. I'm not convinced thats enough heat to really dry it, and the $9 bag of Silica I bought will fill up this tub multiple times over. A rise in the RH on the hygrometer should tell me when the Silica is saturated. At that point I can swap it out for fresh and once I have most of my bag of fresh used up I will cook off the used stuff in the oven (One of the many advantages of keeping my home-front a bachelor pad)

I plugged it in yesterday and I stopped this morning to check on it, it was 42% and had been up to a high of 57% after originally plugging it in. So it seems to be doing something. Again I can't say with any guaranteed certainty this will help my engine last longer, but I am hard pressed to find a reason it would hurt. I understand frequent operation is best but my thought is having torn apart car engines that once the oil drips off a part it can get a light surface rust on it in a nights time... So I understand there are plenty that will say it will help nothing, but I'm not sure either side of that argument can truly prove definitively either way. I'm willing to gamble my $40 it will do something to help and if I'm wrong I'm out 8 gallons of AvGas money...
 
Last edited:
The rubber unit and plug I used for oil filler attachment point is center towards topA6085719-FE6F-4DFC-9FD7-E5FAFAA53FE2.jpeg
 
Cool. How are you hooking it up to the engine?

I had looked at building one, but the biggest problem I have is that my breather tube hooks right to the top of my oil filler, so there wouldn't be any circulation through the engine. I suppose what I really need is to build one that has a cap on the oil filler tube, and that cap would have one short tube to catch the return air and one long tube that would drop all the way down. As it is right now, I just leave the filler cap open between flights and hope that diffusion does the trick well enough.
 
Interesting contraption. In most places where we want to dry something out, whether carpet, drywall, etc., we apply moving air. I figure an air pump that pushes regular old air into the oil filler and exhausts through the crankcase breather will do as much as anything. Just keep the air moving to ventilate the terrarium, so to speak. I guess I’ll need to test that theory.
 
Interesting contraption. In most places where we want to dry something out, whether carpet, drywall, etc., we apply moving air. I figure an air pump that pushes regular old air into the oil filler and exhausts through the crankcase breather will do as much as anything. Just keep the air moving to ventilate the terrarium, so to speak. I guess I’ll need to test that theory.
In Alaska where the air is dry, maybe. I wouldn’t advise that approach in Florida.
 
If the air temp and engine temp are in equilibrium, why not?
 
so....how are you pulling/pushing this dry air thru the engine? And....remember there is an intake and exhaust valve that will be open and expose an air passage thru the engine....so you either need to seal those or deal with that somehow.
 
so....how are you pulling/pushing this dry air thru the engine? And....remember there is an intake and exhaust valve that will be open and expose an air passage thru the engine....so you either need to seal those or deal with that somehow.
It wouldn’t be airtight, but I would think the rings would keep most of the air going out the breather. With the closed loop system, the low pressure would make that even more likely, and with the open loop system, who cares if some of the air goes out the valves?
 
so....how are you pulling/pushing this dry air thru the engine? And....remember there is an intake and exhaust valve that will be open and expose an air passage thru the engine....so you either need to seal those or deal with that somehow.

Your valves have an open pathway to the crankcase?
 
I’d enjoy watching you try to pour fuel into the intake ports of a horizontally opposed airplane engine!
 
How will you determine the device is effective at reducing humidity in your engine?
 
If that question is to me? I'll watch condensation in the rocker box drains (Lycoming.) Humidity readings at the exhaust end (crankcase breather) should decline with time, too. It's all an experiment.
 
I’d enjoy watching you try to pour fuel into the intake ports of a horizontally opposed airplane engine!
Not a problem ...
38496356-img-small-plane-flips-over-due-to-cross-wind.jpg
 
You could heat the air and it would double as a engine heater. Since it would be left on even a mildly warm air would keep the engine ready to go.

For that to have a significant effect, you'd probably have to have insulated tubes as well as either much hotter air, or moving a higher volume of air. Putting a 100-watt bulb in the box and not insulating anything probably wouldn't do a whole heck of a lot.

But, now that you mention it, the thermal effects should be considered. I'm thinking that if I had an un-insulated setup like this, that my engine (with its Reiff preheater) would actually get *colder* with this running.

Does anyone know of a device that could measure temp and RH inside the engine and record that over time, or would I need to build that to establish a baseline?
 
Last edited:
Does anyone know of a device that could measure temp and RH inside the engine and record that over time, or would I need to build that to establish a baseline?

One of these down the dipstick tube would work, but it’s a bit pricey.
Draminski Hay & Straw High-Moisture Meter with Probe Plus Temperature Display https://www.amazon.com/dp/B00EX9L0EO/ref=cm_sw_r_cp_tai_cTbhCbEA7FVSF

If you work with an Arduino you can use a DHT11 or DHT22, but they are too big to get into the engine easily. A SHT11/15 would work if you are handy enough to construct a probe.
 
Last edited:
:frown3::frown3::frown3::nonod::nonod:

RIP dear Skywagon.

My Skywagon was the fourth one after this one. It still has an active registration, by the way. Somebody'll fix it, or already has.

Those things are almost always born again. My mechanic once told me when it comes to Skywagons you can tell how badly it was last wrecked based on how nice the paint, interior and avionics are. ;)
 
Those things are almost always born again. My mechanic once told me when it comes to Skywagons you can tell how badly it was last wrecked based on how nice the paint, interior and avionics are. ;)

Mine was on it’s back very early in its life. Hit a peat bank while taking off on a short lake and went onto it’s back. Before I knew anything about that a mechanic friend went with me to look at it he hadn’t gotten out of the truck yet and told me it had damage history. He saw flush rivets in the leading edges. Cessna bonded the leading edges. He did a thorough inspection and told me to buy it. Said it was put back together better than Cessna built them. :)
 
One of these down the dipstick tube would work, but it’s a bit pricey.
Draminski Hay & Straw High-Moisture Meter with Probe Plus Temperature Display https://www.amazon.com/dp/B00EX9L0EO/ref=cm_sw_r_cp_tai_cTbhCbEA7FVSF

Yeow. When it costs $350 to see if you have a problem and $100 to fix the problem... Just fix the problem! :eek:

If you work with an Arduino you can use a DHT11 or DHT22, but they are too big to get into the engine easily. A SHT11/15 would work if you are handy enough to construct a probe.

The DHT22 is questionable, but the DHT11 should be able to fit into my engine just fine. I would want to figure out a way of ensuring it couldn't actually fall off the end of the cable and into the crankcase, though! :eek:

A few years ago, I was designing a system like this in my head. Unfortunately, my head works a lot faster than my hands (or my wallet). :rofl: Eventually I had figured on a combined automatic closed/open loop system that would keep it open until the humidity was down to ambient so as to make the desiccant last as long as possible, and maybe even try to sense some of those corrosive gasses so as to not keep them in the system when the loop was closed. And, of course, there would be an automatic system to bake the desiccant when necessary, and and and... then I didn't build it because it got way too complex in my head before I even started. :rofl:

Oddly enough - On my latest BFR, my CFI gave me crap for leaving the oil cap open ("something could crawl/fall in there!") but then the IA praised me for doing it when the plane went in for annual! I suppose if I actually built this thing, I would make them both happy.

I have both an Arduino and a RasPi, but I haven't experimented with them nearly enough. It seems like the Arduino would be better suited for the test/measurement/control aspects, but I'm also thinking about logging data onto an SD card. How hard would that be to do with an Arduino?
 
I have both an Arduino and a RasPi, but I haven't experimented with them nearly enough. It seems like the Arduino would be better suited for the test/measurement/control aspects, but I'm also thinking about logging data onto an SD card. How hard would that be to do with an Arduino?

Ah, Google. What did we ever do without you?

https://www.arduino.cc/en/Tutorial/Datalogger

Of course, Google still can't help me find what the inner diameter of the oil filler tube is on an IO-550-G. :dunno:
 
Cool. How are you hooking it up to the engine?

I’m using that 2” rubber fitting over the oil fill to pump air in and my vent tube comes out bottom of cowl and literally my air compressor line slides right in- snug but not to where it’s gotta be forced. I would think you would need a fancy cap to do it.

For yours I’m picturing you could use that rubber fitting like I did- there’s a pic of it... and simply have two lines through the top of it- one that terminates to pump air in and one that’s just long enough with some sort of rubber cork or whatnot that second hose is ran through to stick in the vent as you put the fitting on... if that doesn’t make sense I could scratch you what I’m thinking
 
I have both an Arduino and a RasPi, but I haven't experimented with them nearly enough. It seems like the Arduino would be better suited for the test/measurement/control aspects, but I'm also thinking about logging data onto an SD card. How hard would that be to do with an Arduino?
Easy, just get the daughter card with sdcard.
 
Interesting contraption. In most places where we want to dry something out, whether carpet, drywall, etc., we apply moving air. I figure an air pump that pushes regular old air into the oil filler and exhausts through the crankcase breather will do as much as anything. Just keep the air moving to ventilate the terrarium, so to speak. I guess I’ll need to test that theory.

I think in dry enviornment that would work, and even in damp enviornment would at least push the corrosive gasses out. But this system was so quick and easy to build the added reduction of RH I can't think won't help...

How will you determine the device is effective at reducing humidity in your engine?

I would say the historic RH measurements on the hygrometer would be indicative... As it first starts to run the hygrometer should show a relatively accurate RH of the crank as that silica won't absorb much RH on first pass, so the High indicated on the hyrometer should show close RH to the engine and even in one day I seen a 15% reduction in RH... Technically if the engine is kept warm the RH in the engine would be less than whats in the pump tub as the pump tub right now is pretty cold and engine warm, so though absolute humidity would be the same, relative humidity would be less in the warm engine vs the cold pump tub given the same absolute humidity.

For that to have a significant effect, you'd probably have to have insulated tubes as well as either much hotter air, or moving a higher volume of air. Putting a 100-watt bulb in the box and not insulating anything probably wouldn't do a whole heck of a lot.

But, now that you mention it, the thermal effects should be considered. I'm thinking that if I had an un-insulated setup like this, that my engine (with its Reiff preheater) would actually get *colder* with this running.

Does anyone know of a device that could measure temp and RH inside the engine and record that over time, or would I need to build that to establish a baseline?

I agree preheating this way could be tough, as the air flow is so low. I would also assume the same of its cooling effect. The whisp of cold air isn't going to do much cooling when that whole engine and all that oil is warm... The CFM of this is pathetically low, but the idea is its a constant trickle of dried air, and its effectiveness is based on the "slow and steady wins the race" theory.

On engine RH, it has to be somewhat reflected on the hygrometer in the pump tub though maybe not the percent accurate factoring in temp differences and the fact its a closed loop not a sealed loop..


so....how are you pulling/pushing this dry air thru the engine? And....remember there is an intake and exhaust valve that will be open and expose an air passage thru the engine....so you either need to seal those or deal with that somehow.

An aquarium pump pumps the air in and on opposite end sucks it out... It is not a sealed system by any means I understand that its not 100% sealed, but limiting sucking in constant new air I figure will lower the RH the most possible.
 
Buy a big dehumidifier and run in the hangar.

You would need a hanger that was sealed about as good as a house or it would be like peeing in the wind, not to mention very pricey to run if you are paying electric, and you would still need an air pump to push that dried air into the engine for it to have any effect on the RH inside the crankcase in any amount of reasonable time I would think. Big dehuey units cost a lot more than $40 to purchase even if you get electric included with your hanger...
 
44% RH today in the pump tub and ambient was 85+%... so the continental is drier inside than it is outside...
 
Last edited:
Does anyone know of a device that could measure temp and RH inside the engine and record that over time, or would I need to build that to establish a baseline?

I came across an interesting article from The Twin Cessna Flyer where they did test the temp and RH of the engines, as well as how long the oil protects them. Good stuff!

http://nebula.wsimg.com/1fd3c69cbb2...3181A996A1861998A&disposition=0&alloworigin=1

And, considering he's in Chicago and conditions are fairly similar to here, I guess now I need to buy/build a dehydrator. They did find some temp/RH data loggers tho:

https://www.lascarelectronics.com/data-loggers/temperature-humidity/
 
Last edited:
Build.. super cheap n easy...

Hahahahahahaha.... Yeah, in theory.

I tend to over-do things though. The dehydrator in my mind isn't just a dehydrator. It's got an automatic open/closed loop switching system based on sensed internal and ambient temperature and humidity, logs all four of those plus time and local weather, and heats the air it pumps into the engine to avoid working against the engine heater in the winter. ;)

Oh, and it makes me enough free time to build it. :(
 
So, a question:

Let's say you could build a heater and install it under your cowl. That heater would use radiant heat to raise the temperature under your cowling to (say) 60F. You leave the heater plugged in all winter, and your engine experiences low RH all winter because of the warmer than ambient environment under cowl.

If you left the oil cap loose, wouldn't your engine dry out in a day or two after use?

Wouldn't this be easier than a dehumidifier?

Is there a reason it wouldn't work?
 
So, a question:

Let's say you could build a heater and install it under your cowl. That heater would use radiant heat to raise the temperature under your cowling to (say) 60F. You leave the heater plugged in all winter, and your engine experiences low RH all winter because of the warmer than ambient environment under cowl.

If you left the oil cap loose, wouldn't your engine dry out in a day or two after use?

Wouldn't this be easier than a dehumidifier?

Is there a reason it wouldn't work?

That's essentially what I already do. I have a Reiff heater on the engine, a blanket over the cowl, cowl plugs, and I crack the oil cap.

That's all well and good in the winter - His data for a similar situation showed 8% relative humidity in the engine with no dehydrator. But, in the humid summer months, when the engine heater isn't doing anything, it gets interesting. After shutdown, the engine gets hotter inside for about 20 minutes before reversing course and cooling down, and after about 8 hours the humidity inside the engines is over 80% and stays there. :eek:

The other factor to consider in my case is that on my IO-550, the breather tube attaches to the oil filler tube just below the cap, not directly to the engine. So, there's no airflow *through* the engine, and I'm relying on diffusion to dry it out. Not particularly effective.

What I'm envisioning for a dehydrator is to have a second oil cap, with a hole drilled in it so that I can run a tube through it. That tube will run down into the engine (just to the top of the oil), the other will be connected to the other end of the breather. That way I'll at least get some fresh air down into the inside of the engine. It won't fully flow through the engine, but it'll improve the diffusion process dramatically, probably by a couple orders of magnitude since the air will no longer be stagnant and it'll be happening down inside the engine rather than up at the top of the filler tube. (Yes, I know, it happens everywhere, but I'm bringing much drier air much farther into the engine.)

Part of what I'm interested in is the data logging portion. That will tell me if there's any advantage to hooking up the dehydrator in the winter, and how bad the situation is during the summer. I'll build the logger first and incorporate it into the rest of the design if I take it further.

It's also occurring to me now that I might be able to get away with just adding a fan to the aforementioned oil cap apparatus.
 
don't forget.....the valves. one or two will be open, but not to the breather.
 
That's essentially what I already do. I have a Reiff heater on the engine, a blanket over the cowl, cowl plugs, and I crack the oil cap.

That's all well and good in the winter - His data for a similar situation showed 8% relative humidity in the engine with no dehydrator. But, in the humid summer months, when the engine heater isn't doing anything, it gets interesting. After shutdown, the engine gets hotter inside for about 20 minutes before reversing course and cooling down, and after about 8 hours the humidity inside the engines is over 80% and stays there. :eek:

The other factor to consider in my case is that on my IO-550, the breather tube attaches to the oil filler tube just below the cap, not directly to the engine. So, there's no airflow *through* the engine, and I'm relying on diffusion to dry it out. Not particularly effective.

What I'm envisioning for a dehydrator is to have a second oil cap, with a hole drilled in it so that I can run a tube through it. That tube will run down into the engine (just to the top of the oil), the other will be connected to the other end of the breather. That way I'll at least get some fresh air down into the inside of the engine. It won't fully flow through the engine, but it'll improve the diffusion process dramatically, probably by a couple orders of magnitude since the air will no longer be stagnant and it'll be happening down inside the engine rather than up at the top of the filler tube. (Yes, I know, it happens everywhere, but I'm bringing much drier air much farther into the engine.)

Part of what I'm interested in is the data logging portion. That will tell me if there's any advantage to hooking up the dehydrator in the winter, and how bad the situation is during the summer. I'll build the logger first and incorporate it into the rest of the design if I take it further.

It's also occurring to me now that I might be able to get away with just adding a fan to the aforementioned oil cap apparatus.

If you use a fan make sure you choose the right one... axial (like a fan used in house in summer) moves most cfm per watt but lacks the ability to push air through resistance- curves, tubes, dects etc... that’s why our furnaces have a Squirell cage style- it’s not as efficient cfm per watt used but the moving air has is “tougher” to keep moving.

I think even a venting system with no dehydration would work better than nothing especially in cold dry months..

I chose the fish pump, very low flow but it’s at work 24/7 so it’s a trickle but let’s face it not a huge volume of air inside our crank cases... even a big engine.

I’m considering a rebuild of my dehuey for summer, once I finish sewing my custom cowl cover- I’m close just gotta sew in Velcro and a cinch strap... thinking of a small air compressor if I can get a valve on timer and do it in “bursts”. Not high flow, throttle the flow with a valve but to let the air drawn into the silica tub sit a bit so silica can do its thing then move the air... silica isn’t fast like an electric dehuey so I thought the sitting time in the silica tub then a “blast” might do more good. Though with this model I would think warming the air may be of benefit more...

Idk my exact internals but air inside cowl is in mid 80s w my preheater and a blanket even when it’s 15 degrees out, and I have oil temp showing a bit over 100 at start up.

Great article btw for us preheat nerds :) confirmed what I was planning from other reading - I will preheat even at temps I will be laughed at for doing it... I’ve even pondered doing it in summer for a couple hours if doable... no reason to not have her as close to operating temp as possible at start up... I think it’s accepted that all engines endure greatest wear at start up, and temp isn’t only influence on that- but why not mitigate one of the start up wear factors if possible?
 
Back
Top