diy solar

diy solar

(SOLVED) Solving the climate control crises

I've been diving deep into solving the temp control environment for vanlife. The biggest gripe of van and small space dwellers who want to live off grid is the struggle to keep their space a constant 70 degree temperature in their van or small space.
........
There's other variables at play, what climate are you in? Are you in the desert where evaporative cooling is quite effective but requires a ready supply of water, or are you in hot / humid climates where you need active moisture removal? My personal suggestion is to use an inverter based AC unit, in particular the Midea 8k BTU U-style window unit. I have the same unit in my room that Will has in his. These AC units are awesome. These have a "special" feature in them as they use their condensation to sling water onto the condenser coils getting a little boost in performance that no other AC unit (outside of water slinging window units) can obtain. The best part of this is that you can boost performance on-demand, just dump in 20oz of filtered water.

All AC units have curves showing their capacities. The hotter the inside and outside temps, the less cooling capacity they have. This plagues all units, there's no way around this. However, having a unit that can utilize condensate to hack the curves drastically changes things. Suddenly a 100 degree day isn't a 100 degree day for the AC unit that's able to utilize evaporative cooling on its condenser. It gets a huge bump in efficiency / capacity. You can apply this hack to other AC units by way of a mist system, but you must use filtered water to feed it as mineralization and bleach are bad for units. AC units that are simply on or off (traditional) don't really get a bump in power usage realm, just the frequency at which they turn on and off (saving total usage power, but not running power). For inverter units on the other hand, the computer will see it's doing a great job and can slow down the pump thus using less energy to produce the same amount of cooling.

The window units are gravity drained like all window units and need to have the rear lower than the front. Not drastically, but a condition that must be met at all times during operation. There's no easy inside (as of from the inside portion of the unit) way of adding additional water to them, but a 1/4" line could be run to them where you can periodically turn on water to fill its basin. As mine is a window unit, and the benefit of these as window units allows you to retain opening the window, I just open the window and dump in some water if I want a boost in capacity. I envision these being installed in-place and then closing in the pass through. (google midea U inverter ac if you're not sure what they are, and what I mean by closing it back up, see crude drawing attached)
 

Attachments

  • Screenshot 2021-06-28 202705.png
    Screenshot 2021-06-28 202705.png
    98 KB · Views: 11
A nice benefit of the inverter units is that they run non-stop when it's cooling time (at lower power), have no large in-rush current (inverter benefit), and don't require large / high surge power requirements. Though I haven't tried it, I'd bet it could run off of a cheap 750w modified sin wave inverter. Inverters typically turn the AC into rectified DC and don't really care what the sin wave looks like, let alone if it's even a sin wave. Most appliances (should really be all, but do your own due diligence) that have an inverter will run both AC and DC of the same voltage. Experiment on your own dime ;)

Ice chest coolers / ice in general is just a really poor solution. You can't make your own ice (conservation of energy) and you'd get more cooling capacity out of buying the same $ to $ ratio of gasoline and powering an inverter gen. Trying to get a small cooler / refer unit that runs on 12v to make you that much ice is a fools game. Do the math. Most of the cooler / ice suggestions are debunked with basic thermodynamics and energy cost to benefit analysis. It will always be easier to make cold air than it will be to make ice, that in turn makes cold air. Especially if a heat pump isn't used to make the ice, and regardless if a heat pump is present, if it's dumping its heat load into the occupied space, you also have the additional heat of the system's inefficiency -- all of that on top of the original heat load. No, no no -- besides, a small cooler based refer unit isn't making you that much ice.
 
Last edited:
Midea 8k BTU U-style window unit.
I too use the midea 8000 BTU/hr unit and it’s very efficient and power thrifty. Amazing device. Curse the hecking beeping though… refuse to use wifi to disable the beep. One person posted a tutorial on how to remove the beeper entirely :)
These have a "special" feature in them as they use their condensation to sling water onto the condenser coils getting a little boost in performance that no other AC unit (outside of water slinging window units) can obtain.
I’ve encountered a whynter brand 12,000 BTU/hr unit last week in person that has two tubes and it slings the water onto the hot condenser to evaporate it and blow it out the hot tube. They say condensate drain not needed. Not trying to be contrarian; just sharing because the efficiency gain from driving water condensate to the hot condenser is really neato and well done mentioning it!

The water condensing onto the evaporator side is cold!! Put it back on the refrigerant condenser for a big win!
 
I’ve encountered a whynter brand 12,000 BTU/hr unit last week in person that has two tubes and it slings the water onto the hot condenser to evaporate it and blow it out the hot tube. They say condensate drain not needed.
I've got a single vent system that does the same -- well it did, but something happened and now it just runs 3hrs before needing manually emptied. Duct vented portable ACs are a cancer to the user. Avoid if at all possible. I thought everyone knew to avoid those :cool:

The amount of wasted heat those things create with the ducts is nuts. There's really no way to make those portable and efficient. If someone can fit a 6" duct, they can fit a 3" hole for a mini split :)
 
that sounds delightfully frustrating! i’ve heard many sad stories of single tube type and the way they pull hot air from outside to move air.

if a dual hose type is properly configured, i would assume it should be better behaved than the single tube type.

but i only have experience with Dual Tube and U shaped (midea) so far.

definitely would want to place the hot condenser outside to begin with

i’ve also read about corrosion of the condenser coil when it’s used to evaporate condensate, in models that do not have some anti corrosion coating such as epoxy

i’m still seriously considering building my own device from parts using water and corn glycol as heat exchange work fluid.


easier for me to engineer a pleasant mount for that thin radiator than a mini split

of course this isn’t for everyone, don’t overcomplicate! but i sure as heck like DIY so it’s important for me to mention for other DIY minded HVACkers
 
Corrosion is always a possibility, especially when not properly mitigated. Aluminum fins on copper coils, the copper wins. It's especially brutal in saltwater environments -- the air contains salty moisture and then the AC pulls it out and now has saltwater getting slung on aluminum. Chlorine will absolutely destroy an aluminum rad in just a couple years. Properly designed systems have a resin on the copper to help mitigate it, some will also nickel plate the copper to mitigate it. Getting at least 6 years out of a $400 AC unit is okay in my book -- I typically get 10 at least.

Not quite sure what you're suggesting on the AC side of things, going from air-to-water and then then water-to-air creates more loss. There is a way to get slightly better efficiency, but you'd have to use a much larger sized rad to find it (assuming that the original solution was somehow undersized). Each time you change mediums, there's an efficiency loss. It's why air to air intercoolers are preferred over air-to-water-to-air. There's only 2 reasons to use the latter and that's due to space constraints --- it's at times difficult to duct air to a place where an air-to-air intercooler works, and the 2nd use case is for 'doping' the water loop with some other means of cooling like adding a lot of ice into the tank so that the intercooler is using substantially colder water than could be achieved otherwise. The latter is only sustainable for drag racing for the most part as there's no easy way to continually supply such a tank with ice while also pulling out the water. Ford had a lightning concept at one point that used the AC system to chill the water tank to much colder temps allowing for more power, but it was never seen on a production vehicle. Manifold type super chargers typically use air-to-water-to-air because the SC is the intake plenum, and it's better than nothing.
 
Eventually I will like to do an experiment with a solid state heat pump for cooling a small volume of air, for example equivalent to a small van interior.

Radiator on the outside circulates water/glycol to dissipate heat to outdoor ambient (40-50C max)

Water block on radiator loop interfaces with solid state heat pump. Heat pump heats up external radiator, cools internal water reservoir. I would probably want to cool the water reservoir to just above dew point for the indoor air. Then run it through a radiator+fan to cool the indoor air. And also have another tap to run the water through silicone tubing on chair or blanket for direct cooling. Very non-common type use. That partially motivates the water chiller approach.

It's a lot easier to engineer a compressor and just pump refrigerant to get a certain amount of BTU/hr pumped. Definitely acknowledge that.

Compressor heat pump can achieve Coefficient of Performance 3-5 commonly from my limited reading

However, thermoelectric cooling can achieve Coefficient of Performance 2-4 from my experience. If driven at high voltage they fall to CoP <1 though.

By stacking the plates, each plate contributes a fraction of the overall heat pumping potential, allowing more efficient operation. Some people attempt to achieve a delta T of 30-50 degrees C with only one layer of peltier element and that is fundamentally inefficient from my limited reading. When each module contributes delta 3-10 degrees C then more efficient operation is possible. Each layer can be considered analogous to a phase in a multiple phase compression type heat pump. This implies a need for 15-20 layers or phases of thermoelectric coolers. This is a very large complexity to regulate!

However, I've already tested 3 phase and 7 phase solid state heat pump, and the initial results are satisfying enough for me to continue testing.

A human only needs 100-150 W of cooling by themselves. The environment needs many more watts of cooling. By recognizing the two overlapping ways to cool myself, I hope to find a pleasant solution for myself. Can anyone tell me what it's like to sit in a chair with 200W of direct contact cooling? How much would this reduce the need for me to cool the rest of the air all the way to e.g. 20-24C?
 
Last edited:
Eventually I will like to do an experiment with a solid state heat pump for cooling a small volume of air, for example equivalent to a small van interior.

Radiator on the outside circulates water/glycol to dissipate heat to outdoor ambient (40-50C max)

Water block on radiator loop interfaces with solid state heat pump. Heat pump heats up external radiator, cools internal water reservoir. I would probably want to cool the water reservoir to just above dew point for the indoor air. Then run it through a radiator+fan to cool the indoor air. And also have another tap to run the water through silicone tubing on chair or blanket for direct cooling. Very non-common type use. That partially motivates the water chiller approach.

It's a lot easier to engineer a compressor and just pump refrigerant to get a certain amount of BTU/hr pumped. Definitely acknowledge that.

Compressor heat pump can achieve Coefficient of Performance 3-5 commonly from my limited reading

However, thermoelectric cooling can achieve Coefficient of Performance 2-4 from my experience. If driven at high voltage they fall to CoP <1 though.

By stacking the plates, each plate contributes a fraction of the overall heat pumping potential, allowing more efficient operation. Some people attempt to achieve a delta T of 30-50 degrees C with only one layer of peltier element and that is fundamentally inefficient from my limited reading. When each module contributes delta 3-10 degrees C then more efficient operation is possible. Each layer can be considered analogous to a phase in a multiple phase compression type heat pump. This implies a need for 15-20 layers or phases of thermoelectric coolers. This is a very large complexity to regulate!

However, I've already tested 3 phase and 7 phase solid state heat pump, and the initial results are satisfying enough for me to continue testing.
Honestly the "thought energy" and money would be better spent acquiring the parts and knowledge for building your own DC heat pump system. Rumor has it US taxpayer money (DARPA) paid to have these things developed. https://www.aspencompressor.com/vertical-mini-compressors (convert watt x 3.4 = btu)

Peltier just isn't there IMO. Check out this build for some inspiration ;)
 
Last edited:
I've built a chiller that used bags of ice cubes.

After we engineers specified a thermal chamber the company bought it, but set it up in their Bangalore office.
So when I needed to characterize a DUT (electronic Device Under Test) I assembled a temperature cycling machine out of a computer controlled turkey roaster and a Styrofoam chest of ice. While monitoring a thermocouple, controller switched 24V on/off to the heating element, regulating high temperature setting. Then it turned on a fan and sucked air from the bottom of the ice chest, so room air was drawn over the cubes.

It got the job done, but bag of ice was melted within a couple hours. You may be comfortable for a little while with your "Frankencooler", but don't expect it to last long.


1624944733064.png
 
Thank you for sharing that interesting DC refrigerant pump ? Eventually I would like to learn how to braise copper.
1624944579447.jpeg

DC small scale refrigerant compressor does certainly seem like the most realistic option if I don’t want to be in R&D forever.

12V 10A 120W power 360W pumped CoP=3
24V 8.5A 204W power 460W pumped CoP=2.3
48V 4.2A 202W power 550W pumped CoP 2.7

550*3.412= 1876.6 BTU/hr!



with a stack of peltier i’m expecting to pump 10-30W with each element depending on system conditions.

1624945336623.jpeg

1624945357572.jpeg

i don’t see anyone trying the efficient peltier thing, so i’m “wasting” my time checking ☀️waiting on some thermal pads.. thermal paste is so unpleasant to work with

edit: for example, in recent vehicles with more complex cooling systems, it's becoming common to use heat exchangers and different working fluids to achieve the desired temperature ranges for the different areas of the machine. for example R-134a, water/glycol, and oil

 
Last edited:
Thank you for sharing that interesting DC refrigerant pump ? Eventually I would like to learn how to braise copper.
Braising is basically soldering with higher temps. You're not welding, just soldering with more heat. These are good practice.....

Technically those are soldering, but Al melts at such a low temp, it's on the border of braising and some in the industry still consider it braising. Braising needs a torch, propane is fine, doesn't need anything more exotic.
 
Braising is basically soldering with higher temps. You're not welding, just soldering with more heat. These are good practice.....

Technically those are soldering, but Al melts at such a low temp, it's on the border of braising and some in the industry still consider it braising. Braising needs a torch, propane is fine, doesn't need anything more exotic.
Thank you for your guidance ? one day I'll charge a loop safely ?at least it doesn't require argon or some inert gas flow (y)

I wish to use CO2 as working fluid for heat pump one day, but accept that due to the pressures involved the first system would probably be R-134a.

Cheers!


edit: an interesting aspect of thermoelectric elements is that the seeback effect is acting just as it does in a fluke meter with a thermocouple. to ensure efficiency, it’s important to stay below 10 C difference across a single element. so then regulation of it becomes a relatively simple task of alternating between A) applying 1-2V to a given peltier element, and B) disconnecting power momentarily and reading the voltage potential across a given element to assess delta temperature.

if the seeback voltage indicates a delta temperature of close to 10 degree C, then that module would be disconnected from power for an extra portion of upcoming time cycles.

this provides time for the copper thermal reservoir to normalize the module’s temperature and ensure efficient operation. perhaps reminiscent of mosfet drive modes and efficiency.

the entire goal is that each layer would be an equal number of degrees (0-10C in this case) closer to the target temperature. copper plates serve to maintain stable intermediate temperature regions to ensure high efficiency operation.

in this way, the peltier elements act as very small low capacitance thermal battery cells that need to be balanced. once “charged” to a delta T of 10C it will then create a voltage potential itself until the temperature difference across the two faces equalizes.

1624957006734.jpeg

just like a pack with even one lithium cell overvolted, that over voltage condition in a single module can dramatically impact system efficiency (not to mention flammability) in this metaphor the copper plate is a very fast active cell balancer

for example, this test device in the above picture has two aluminum water blocks. 240x80x17mm

the decidedly inedible thermal conductivity sandwich format is:
Code:
water inflow              ( 0.6W/(m•K))
aluminum water block      (~200W/(m•K))
thermal pad               (  ~8W/(m•K))
copper sheet 1mm          (~400W/(m•K))
thermal pad               (  ~8W/(m•K))
peltier alumina pad       ( ~30W/(m•K))
peltier bismuth telluride (~1.5W/(m•K))
peltier alumina pad       ( ~30W/(m•K))
thermal pad               (  ~8W/(m•K))
copper sheet 1mm          (~400W/(m•K))
thermal pad               (  ~8W/(m•K))
aluminum water block      (~200W/(m•K)
water outflow             ( 0.6W/(m•K))

with the copper and peltier segment copy and pasted for multiple layer design “multiple phase” cooling to achieve improved efficiency when compared with a single layer at 12V (very inefficient low CoP)

two phase would look like this:

Code:
water inflow              ( 0.6W/(m•K))
aluminum water block      (~200W/(m•K))
thermal pad               (  ~8W/(m•K))
copper sheet 1mm          (~400W/(m•K))
thermal pad               (  ~8W/(m•K))
peltier alumina pad       ( ~30W/(m•K))
peltier bismuth telluride (~1.5W/(m•K))
peltier alumina pad       ( ~30W/(m•K))
thermal pad               (  ~8W/(m•K))
copper sheet 1mm          (~400W/(m•K))
thermal pad               (  ~8W/(m•K))
peltier alumina pad       ( ~30W/(m•K))
peltier bismuth telluride (~1.5W/(m•K))
peltier alumina pad       ( ~30W/(m•K))
thermal pad               (  ~8W/(m•K))
copper sheet 1mm          (~400W/(m•K))
thermal pad               (  ~8W/(m•K))
aluminum water block      (~200W/(m•K)
water outflow             ( 0.6W/(m•K))

it may seem like a bunch of random components thoughtlessly tossed together, and maybe that’s what i’m doing ??but i do think there is a good reason for me to investigate more for now.

the copper sheets are to ensure uniformity of temperature across adjacent faces, which is a direct driver of efficiency. also, to facilitate transmission of heat to the aluminum water block.
 
Last edited:
There's other variables at play, what climate are you in? Are you in the desert where evaporative cooling is quite effective but requires a ready supply of water, or are you in hot / humid climates where you need active moisture removal? My personal suggestion is to use an inverter based AC unit, in particular the Midea 8k BTU U-style window unit. I have the same unit in my room that Will has in his. These AC units are awesome. These have a "special" feature in them as they use their condensation to sling water onto the condenser coils getting a little boost in performance that no other AC unit (outside of water slinging window units) can obtain. The best part of this is that you can boost performance on-demand, just dump in 20oz of filtered water.

All AC units have curves showing their capacities. The hotter the inside and outside temps, the less cooling capacity they have. This plagues all units, there's no way around this. However, having a unit that can utilize condensate to hack the curves drastically changes things. Suddenly a 100 degree day isn't a 100 degree day for the AC unit that's able to utilize evaporative cooling on its condenser. It gets a huge bump in efficiency / capacity. You can apply this hack to other AC units by way of a mist system, but you must use filtered water to feed it as mineralization and bleach are bad for units. AC units that are simply on or off (traditional) don't really get a bump in power usage realm, just the frequency at which they turn on and off (saving total usage power, but not running power). For inverter units on the other hand, the computer will see it's doing a great job and can slow down the pump thus using less energy to produce the same amount of cooling.

The window units are gravity drained like all window units and need to have the rear lower than the front. Not drastically, but a condition that must be met at all times during operation. There's no easy inside (as of from the inside portion of the unit) way of adding additional water to them, but a 1/4" line could be run to them where you can periodically turn on water to fill its basin. As mine is a window unit, and the benefit of these as window units allows you to retain opening the window, I just open the window and dump in some water if I want a boost in capacity. I envision these being installed in-place and then closing in the pass through. (google midea U inverter ac if you're not sure what they are, and what I mean by closing it back up, see crude drawing attached)
VERY helpful man. Looking into this now. I don't know how stealth my Honda Odyssey will be after installing one of these though but this definitely seems to be the best window AC unit for vanlife as I've looked at many.
 
Apologize for the triple post but I finally got a chance to read through everything and I'm still excited to try my build when my ICECO 60QT 3 in 1 comes in November. I missed the $499 deal and had to pay $560 and I'm getting it a month later but I expect the price to jump to $800 when it comes for regular sale.

I will not build a Frankencooler type build as it melts the ice much faster than the copper/aluminum tube builds and requires a bilge pump in the bottom of the cooler. My plan is to run 3" aluminum pipe or waterproof duct through the bottom of the ICECO. I want the freezer to completely freeze over and don't want to have to worry about the bilge pump freezing over. I fully expect the ice to melt inside the freezer within 3 days but it will save me having to move ice from my freezer into another cooler and 60QT should last a long time.

My plan is to basically refreeze the water at night and then use 1/2 of the ice during the day when it's hot outside and repeat the process and will probably end up running the ICEO on max all day every day lol.

I finally figured out what panel I'm going to use on my Odyssey


I also did not decide to go with the Pecron as it's the same price as the Bluetti EB150 even though it's lighter. I just didn't want to mess with the fact that it did not have an MPPT charge controller like the Bluetti. It adds probably 20 pounds to get the Bluetti but meh...I've wanted to get a Bluetti for awhile now I figured I might as well get one


Thanks for everyone's input guys keep it up and we will solve the climate control crises together!
 
Nice going!

I really appreciate the focus of this thread on clean slate approaches to human comfort in a mobile setting.

As has been said by others, H2O is cheaper than LiFePO4 and other chemicals.

A solution that requires the purchasing of a few bags of ice is absolutely going to have wider adoption potential than a similar solution that requires purchasing an equivalent weight of LiFePO4 cells and appropriate supporting components.

There are a decent number of industrial users of ice as energy storage. Freeze the water into ice at night buying cheap electricity then melt it during the day with the tiny amount of power (relatively) to just circulate heat into the mass in a controlled way.

If condensation can be managed, then small scale ice as living space thermal buffering could become adopted more.


On another note, a trick that some air conditioners use is to transfer the heat of the refrigerant condenser into the drained condensate that formed on the refrigerant evaporator side. Every watt of thermal energy transferred away into the draining outflow contributes more to the efficiency of the overall system.

Just be careful about corroding the condenser fins due to lack of anti corrosion coating such as a certain type of epoxy.
 
HOBOTECH finally reviewed the one I bought!


SECOP (Danfoss) compressor included now! Can't wait until november! Until then I will use my Prius's climate control setting when traveling lol...
 
Considering this solar generator now over the Bluetti


What do you guys think of this one?
 
Like that it uses LFP. Goal Zero uses NMC and the one I use works fine but LFP chemistry preferable to me. Even with density penalty.

Their images indicate they are using cylindrical LFP cells inside.

1625173713884.png

Sungzu Battery Eh?
“If you know the battery cell and know yourself, you need not fear the result of a hundred charge cycles. If you know yourself but not the battery, for every victory gained you will also suffer a defeat. If you know neither the battery cell nor yourself, you will succumb in every charge cycle.”
― Sun Tzu, The Art of Charging

edit: Looked at more seriously.. nice thick Aluminum case for structural and thermal benefits. If the air is cool then the Aluminum case should keep the cells and BMS happy. -10C claimed so maybe a heater? They mention that the inverter generates heat. One intake and one exhaust fan, good stuff.

I've participated in a number of kickstarters.. this one seems... hopeful... english material could use a little finessing.. technical claims seem believable and in sane range. Would want to learn more before acting personally. 1000 usd for 2500 Wh all in one with inverter and case is a pretty darn fine bargain compared to Goal Zero if it works out. my 2 cents, please mix with grain of salt
 
Last edited:
Like that it uses LFP. Goal Zero uses NMC and the one I use works fine but LFP chemistry preferable to me. Even with density penalty.

Their images indicate they are using cylindrical LFP cells inside.

View attachment 54727

Sungzu Battery Eh?


edit: Looked at more seriously.. nice thick Aluminum case for structural and thermal benefits. If the air is cool then the Aluminum case should keep the cells and BMS happy. -10C claimed so maybe a heater? They mention that the inverter generates heat. One intake and one exhaust fan, good stuff.

I've participated in a number of kickstarters.. this one seems... hopeful... english material could use a little finessing.. technical claims seem believable and in sane range. Would want to learn more before acting personally. 1000 usd for 2500 Wh all in one with inverter and case is a pretty darn fine bargain if it works out. my 2 cents, please mix with grain of salt
This may have just sold me on buying one if these over the Bluetti lol!

One thing I went to clarify before I buy is if it’s compatible with the 340 watt LG solar panel I listed above. I see that their page says it takes up to 500 watts but there is also a circle that says it’s 110 watts solar. I’ll message them.
 
Back
Top