Building a portable AC for my car

[iduhfuse]

Got a datasheet for those fans ? Any notes in it about mounting plane ? Are they brushed or electronic ?
12 V ? Car runs 14, surely they're not that tender.
However - inductance of water pump could be generating a turn-off transient...reverse polarity spike when you flip the switch off
one of those cheap bicolor LED voltage testers would probably show, look for a green flash at turnoff
http://www.walmart.ca/en/ip/auto-voltage-tester/6000119380174try a flywheel diode ?

old jim

I checked system running at 12.36 volts (car off) with multimeter. The fans are Corsair SP 120 High performance. Also can you send me a link for a flywheel diode?...i'm seeing diagrams. I have no idea what that looks like lol. OOPS ****Another thing I forgot to mention. I did remove the pump from the circuit and tested each fan individually. They both did not work.

 

[jim hardy]

Also can you send me a link for a flywheel diode?...i'm seeing diagrams. I have no idea what that looks like lol.

most any diode will do
it ought to be rated for as much current as the pump draws
if it's a 1 amp pump get a handful of in4001's they're dirt cheap and everywhere
1n400X series comes in various voltages , last digit is voltage rating in hundreds 1n4001 is 100 volt, 4004 is 400 volt etc but they're only 1 amp
probably these3 amp would be fine,
https://www.amazon.com/dp/B0087YKERK/?tag=pfamazon01-20

lifetime supply for two bucks ?

it goes right across pump motor terminals
end with the stripe to positive

Are the fans polarity sensitive? Do the instructions warn against reverse voltage?
I saw inductive kick from a Harley Davidson generator kill an expensive "Electric Frank" electronic voltage regulator... it's not something all designers think of.

I hope you find something simpler like a loose crimp connection

old jim

 

[Tom.G]

Good catch, Jim!

Diode: 1N4001. They are about 5 cents. Put one directly across the pump; reverse polarity so it doesn't conduct.
If that's the problem, the fans may be toast. So for good measure, the next time around, put a diode in series with with each fan.
EDIT: and a capacitor across each fan, maybe 0.1u to 0.47u.

I see you beat me to it Jim.

 

[rbelli1]

Don't bother with the 1N4001 diodes. Get 1N4007. They are 1000V and usually the same price as the 50V ones. If you ever need the higher voltage then you already have them. Get a bunch. Almost anywhere something just specifies a diode without any more information they will work.

For series diodes you will get better performance with Schottky diodes. It probably doesn't doesn't matter in this application but the lower forward voltage is useful for higher currents and applications where the voltage drop could be problematic. Don't use them for half wave rectification. The reverse leakage current is bad for polarized capacitors. And a bigger capacitor doesn't help. Or at least so I've been told !

BoB

 

[iduhfuse]

Don't bother with the 1N4001 diodes. Get 1N4007. They are 1000V and usually the same price as the 50V ones. If you ever need the higher voltage then you already have them. Get a bunch. Almost anywhere something just specifies a diode without any more information they will work.

For series diodes you will get better performance with Schottky diodes. It probably doesn't doesn't matter in this application but the lower forward voltage is useful for higher currents and applications where the voltage drop could be problematic. Don't use them for half wave rectification. The reverse leakage current is bad for polarized capacitors. And a bigger capacitor doesn't help. Or at least so I've been told !

BoB

Damn, I understood about a 1/4 of that hahaha. I will wait tomorrow until fans are completely dry. Who knows, maybe water did get into them (I'm hoping not) the rad wasn't even condensating yet but oh well. If the fans are shot, there goes $40...any other alternatives to push air through the rad?

 

[jim hardy]

I see you beat me to it Jim.

sorry - heat of the chase

Another trick would be run whole thing through a ten amp bridge
that way if cord to cigarette lighter plug gets wired backwards - no sweat, plus the bridge does the flywheel.

https://www.amazon.com/dp/B01HDZRYFI/?tag=pfamazon01-20

 

[Tom.G]

I just realized that 50V spikes are common in car electrical systems, and I wouldn't be suprised to see an occassional 100V spike. If those fans are the common 12V units sold for computers, they are brushless DC. The IC in them seems to be a Hall detector and commutator all in a 3-pin TO-92 package. Those would definitely need surge suppression/filtering on the power leads. The question is, what would be the most appropriate? Perhaps back-to-back 16V 5W zeners with a low amp fuse on the incoming power? What say, Jim?

@iduhfuse If the fans are dead, don't trash them. A post mortem would be helpful. Specifically, If there is an IC or transistor in them, let us know the part number and other markings on it. That way we can look for a data sheet and find out what their voltage rating is.

 

[jim hardy]

What say, Jim?

Don't forget that inductive kick reverses polarity.
When the switch opens
instead of current being pushed through the pump motor by the battery
the pump motor now tries to suck current through the switch, which of course it can't, and the only place it can get it from is backward out of the fans.
A motor's inductance can make a lot of voltage trying to do that.A flywheel diode let's the motor suck current , harmlessly circulating it around through the diode, with only one diode drop (-0.6V) which won't hurt anything.

https://www.westfloridacomponents.com/blog/what-is-back-emf-and-what-does-it-do/

old jim

 

[Tom.G]

I was trying to say that 50V to 100V spikes, of either polarity, are part of the automotive environment and was looking for some refinement to the idea of back-to-back zeners, perhaps with a bit of source impedance added. The flywheel, or freewheeling, diode at the pump would still be advantageous and is what joggled my memory of the noisy power environment.

 

[iduhfuse]

Not sure if this helps but I may need to mention that I have a 500watt amplifier hooked directly up to the battery for my speakers...

 

[iduhfuse]

I just realized that 50V spikes are common in car electrical systems, and I wouldn't be suprised to see an occassional 100V spike. If those fans are the common 12V units sold for computers, they are brushless DC. The IC in them seems to be a Hall detector and commutator all in a 3-pin TO-92 package. Those would definitely need surge suppression/filtering on the power leads. The question is, what would be the most appropriate? Perhaps back-to-back 16V 5W zeners with a low amp fuse on the incoming power? What say, Jim?

@iduhfuse If the fans are dead, don't trash them. A post mortem would be helpful. Specifically, If there is an IC or transistor in them, let us know the part number and other markings on it. That way we can look for a data sheet and find out what their voltage rating is.

If there were 50V or 100V spikes etc... would that not blow my 8A fuse I wired into the circuit.

 

[iduhfuse]

Don't forget that inductive kick reverses polarity.
When the switch opens
instead of current being pushed through the pump motor by the battery
the pump motor now tries to suck current through the switch, which of course it can't, and the only place it can get it from is backward out of the fans.
A motor's inductance can make a lot of voltage trying to do that.A flywheel diode let's the motor suck current , harmlessly circulating it around through the diode, with only one diode drop (-0.6V) which won't hurt anything.
View attachment 103197
https://www.westfloridacomponents.com/blog/what-is-back-emf-and-what-does-it-do/

old jim

I like the idea of this, but if I'm understanding this correctly...the idea is my fans would not be getting power? However, when the pump is out of the circuit they don't function. So, in addition to this possibility I think I have another problem elsewhere...Only data I got on the fans are that they are 12V at 0.25 Amps.

 

[iduhfuse]

I just realized that 50V spikes are common in car electrical systems, and I wouldn't be suprised to see an occassional 100V spike. If those fans are the common 12V units sold for computers, they are brushless DC. The IC in them seems to be a Hall detector and commutator all in a 3-pin TO-92 package. Those would definitely need surge suppression/filtering on the power leads. The question is, what would be the most appropriate? Perhaps back-to-back 16V 5W zeners with a low amp fuse on the incoming power? What say, Jim?

@iduhfuse If the fans are dead, don't trash them. A post mortem would be helpful. Specifically, If there is an IC or transistor in them, let us know the part number and other markings on it. That way we can look for a data sheet and find out what their voltage rating is.

All I could find on the fans here: http://www.corsair.com/en/air-series-sp120-high-performance-edition-high-static-pressure-120mm-fan

 

[jim hardy]

I like the idea of this, but if I'm understanding this correctly...the idea is my fans would not be getting power?

no, all you do is add a diode to the pump
so that it doesn't throw an inductive fit when power is switched off

 

[jim hardy]

Looking at post #46
it is not clear to me how air will circulate through your radiator... Both fans point same way ? They can't pull air out that didn't get in someplace...
What did i miss ?

old jim

 

[iduhfuse]

Looking at post #46
it is not clear to me how air will circulate through your radiator... Both fans point same way ? They can't pull air out that didn't get in someplace...
What did i miss ?

old jim

Hi Jim, yes you're correct. I still need to make a 3 inch hole saw cut to vent in air otherwise negative pressure builds in the cooler.

 

[OmCheeto]

Hi Jim, yes you're correct. I still need to make a 3 inch hole saw cut to vent in air otherwise negative pressure builds in the cooler.

Or you could physically reverse the direction of one of the fans.

 

[iduhfuse]

Could really use that AC today *cries*

 

[OmCheeto]

View attachment 103216 Could really use that AC today *cries*

sissy...

 

[iduhfuse]

sissy...

View attachment 103220

Jesus Christ, at that point I'd buy a new car with AC lol

 

[iduhfuse]

Update: I got one fan working again! Somehow the blade part of the fan came loose so I popped it back onto the motor. Weird...I hoped that would be the same issue for the other fan but it is attached just fine. It does not work and I can't figure out why.

 

[iduhfuse]

Update***: swapped out bad fan for a different one. Both are now working on the circuit. Have yet to include any diode etc. Will see if it kills another fan, if it does then I definitely know something is wrong. Also, water pump is rated 13.6v at 3 amps DC. So far so good. Will let you know results tomorrow or next day. I'm tired haha

 

[Tom.G]

@iduhfuse In your post #71, image three, (https://www.physicsforums.com/attachments/img_20160714_175550-jpg.103235/) there is an IC shown mounted at the right side of the circuit board. What are the markings on the IC? Also markings on any other ICs on the board?

EDIT: For testing purposes, can you connect to a 12Volt supply other than your vehicle? This will separate vehicle system issues from any device issues.

 

[iduhfuse]

@iduhfuse In your post #71, image three, (https://www.physicsforums.com/attachments/img_20160714_175550-jpg.103235/) there is an IC shown mounted at the right side of the circuit board. What are the markings on the IC? Also markings on any other ICs on the board?

EDIT: For testing purposes, can you connect to a 12Volt supply other than your vehicle? This will separate vehicle system issues from any device issues.

Only one IC. Will look at it tonight. I have a 12V adapter that outputs 1.4 amps. Not enough for the system but enough to test each fan. So far everything is working good. Except for the one fan that died :( I doubt they're going to give me a replacement because I cut the 3 pin connector off

 

[iduhfuse]

VERDICT*** Okay, I apologize for the long wait. Things got extremely busy for me so I had to put off time on this project.

So far, did some testing today with 8lbs of ice in the cooler.(cubes). Started at 105f and dropped to about 95f in the course of 15-20mins. However, it would not go any lower for the next hour :((( I have a feeling the fans are not pushing enough air past the radiator to extract the negative heat from the ice water. What I will do next is get a powerful single slim fan that will be placed on the opposite end of the cooler top and blow air inside and hopefully air will blast out from the radiator. The two fans on the rad will be eliminated. Can anyone see this as a bad idea/upgrade?

 

[Tom.G]

Will probably work better that way, if you can keep the cover on the cooler when it is pressurized. The important fan characteristics are CFM vs Back Pressure, usually shown as a graph on their data sheet. You can measure the pressure difference of your present setup with a homemade water differential pressure gauge like this one. Made of clear plastic tubing with water in it. Stick one end in your Cooler box and the other end exposed to free air. It's convenient beacuse fan back pressure is measured in inches of water.

 

[jim hardy]

What part of the world are you in?
Does weather lady tell you the dewpoint?
Air cools easily down to the dewpoint but to get below that you have to remove the water vapor, which is steam.
Condensing steam takes roughly 1000 BTU's per pound compared to air's 0.24 BTU's per pound per degree.
Melting ice takes 144 BTU per pound.
So to condense a pint of water vapor out of humid air melts $\frac{1000}{144}$ = ~7 pints of ice.
That's why car airconditioners are so big. Where i live they leave a puddle of water in the parking lot.

Do you have any leaks in your car that soak the carpet ?

 

[iduhfuse]

What part of the world are you in?
Does weather lady tell you the dewpoint?
Air cools easily down to the dewpoint but to get below that you have to remove the water vapor, which is steam.
Condensing steam takes roughly 1000 BTU's per pound compared to air's 0.24 BTU's per pound per degree.
Melting ice takes 144 BTU per pound.
So to condense a pint of water vapor out of humid air melts $\frac{1000}{144}$ = ~7 pints of ice.
That's why car airconditioners are so big. Where i live they leave a puddle of water in the parking lot.

Do you have any leaks in your car that soak the carpet ?

I live in Toronto, Canada. So yeah weather probably does state dewpoint most likely. Humidity RH was 36% in the car. And no zero leaks

 

[jim hardy]

Humidity RH was 36% in the car.

Toronto ! wow...

I typed 95 F and 36% into this
http://andrew.rsmas.miami.edu/bmcnoldy/Humidity.html
and got dewpoint 64 degF
The air coming out of your radiator probably will only creep below that as the car dehydrates . Fortunately you're not having to dry outside air.

Fun project, keep us posted ?

 

[anorlunda]

I have one of these things, and I'm pleased with the cooling I get on very hot days. I point it right at my face, or the back of my neck.

$8.99 retail. You should be able to convert it to the car's 12 VDC easily, both fan and squirter. One per passenger.

 

[iduhfuse]

I have one of these things, and I'm pleased with the cooling I get on very hot days. I point it right at my face, or the back of my neck.

$8.99 retail. You should be able to convert it to the car's 12 VDC easily, both fan and squirter. One per passenger.

I tried this. My car fan on full blast and a sprayer squirter right in front of the vents with cold ice water. It def works but my shirt and face get soaked lol

 

[sophiecentaur]

I have a feeling the fans are not pushing enough air past the radiator to extract the negative heat from the ice water.

There will be a hard limit to the rate at which you can get heat into the surface of the iced water if you are just interfacing at the top. You need a good heat exchanger with a big surface area (like the real thing has) and good circulation of water around the melting ice. Have you measured the temperature on the surface of the (stationary?) water in the cooler? If it is not higher than 0°C then you can conclude that not much heat is being transferred in the way you wanted.

I have one of these things, and I'm pleased with the cooling I get on very hot days. I point it right at my face, or the back of my neck.

$8.99 retail. You should be able to convert it to the car's 12 VDC easily, both fan and squirter. One per passenger.

Looks like the same principle as a Swamp Cooler but very localised.

 

[iduhfuse]

There will be a hard limit to the rate at which you can get heat into the surface of the iced water if you are just interfacing at the top. You need a good heat exchanger with a big surface area (like the real thing has) and good circulation of water around the melting ice. Have you measured the temperature on the surface of the (stationary?) water in the cooler? If it is not higher than 0°C then you can conclude that not much heat is being transferred in the way you wanted.Looks like the same principle as a Swamp Cooler but very localised.

Measured the temp of the water after its been running right? I haven't yet.

Update*** did another test today with 12lbs of ice...little change :( surface of water was roughly 6C

 

[OmCheeto]

Measured the temp of the water after its been running right? I haven't yet.

Update*** did another test today with 12lbs of ice...little change :( surface of water was roughly 6C

?

How did you measure that?
From what I can tell, our systems are nearly identical. (by a factor of 2, for each variable. Which, in the context of this experiment, makes them nearly identical. )

From my July 6 experiment, I was pumping 3 gallons per minute, from a reservoir of 3 gallons. I seriously doubt there was a measurable temperature differential, anywhere in the cooler. (Though, I dumped my ice out of it's bag, into the cooler, and you seem to have complicated things)
I also measured a differential air temperature of 6.5°F, into and out of the radiator.
Given the mass flow rates:

water: 0.194 kg/sec
air: 0.0185 kg/sec​

I would recommend figuring out the change in temperature of the water through the radiator.

ps. I would share more of my data, thoughts, and discoveries, but I'm already 3 hours late for an appointment.

 

[iduhfuse]

?

How did you measure that?
From what I can tell, our systems are nearly identical. (by a factor of 2, for each variable. Which, in the context of this experiment, makes them nearly identical. )

From my July 6 experiment, I was pumping 3 gallons per minute, from a reservoir of 3 gallons. I seriously doubt there was a measurable temperature differential, anywhere in the cooler. (Though, I dumped my ice out of it's bag, into the cooler, and you seem to have complicated things)
I also measured a differential air temperature of 6.5°F, into and out of the radiator.
Given the mass flow rates:

water: 0.194 kg/sec
air: 0.0185 kg/sec​

I would recommend figuring out the change in temperature of the water through the radiator.

ps. I would share more of my data, thoughts, and discoveries, but I'm already 3 hours late for an appointment.

I measured temp using an infrared heat gun. Shoots a laser...I've basically concluded this little experiment is a bust. It was fun, $120 sort of wasted (I can use parts for other things) but otherwise this AC cooler is not effective at all. Guess just have to deal with the heat or either pump that water through a vest...which is tempting

 

[anorlunda]

I tried this. My car fan on full blast and a sprayer squirter right in front of the vents with cold ice water. It def works but my shirt and face get soaked lol

Using the hand-held one, I squirt only when feeling hot and dry. Don't squirt again until the previous water has evaporated. In other words, squirting water must be manually controlled by the person depending on what he feels on his skin.

Here's similar idea that's very easy to manage and that has become popular in recent years because the towels are marketed in garden stores. Wet it, wrap around your neck, then don't wet again until dry.

I tend to favor KISS principle solutions. Most of the things discussed on this thread are far from KISS.

 

[OmCheeto]

I measured temp using an infrared heat gun. Shoots a laser...I've basically concluded this little experiment is a bust. It was fun, $120 sort of wasted (I can use parts for other things) but otherwise this AC cooler is not effective at all. Guess just have to deal with the heat or either pump that water through a vest...which is tempting

I disagree, that the experiment is "a bust".
It only becomes a bust, when the theory is disproven.
You've shared WAY too little data to prove the theory is wrong.

You may not be back, to continue the experiment, but I will...

Lukedafuse; "I'm not afraid to do this experiment!"
Yomda; "You will be..."
Lukedafuse; "The experiment failed! Ahhhhhh!"
Yomda; "That's because, you do no understand the power of the f=ma equation, and have not even delved into the m.dot Q delta.t equations yet..."​

 

[iduhfuse]

I disagree, that the experiment is "a bust".
It only becomes a bust, when the theory is disproven.
You've shared WAY too little data to prove the theory is wrong.

You may not be back, to continue the experiment, but I will...

Lukedafuse; "I'm not afraid to do this experiment!"
Yomda; "You will be..."
Lukedafuse; "The experiment failed! Ahhhhhh!"
Yomda; "That's because, you do no understand the power of the f=ma equation, and have not even delved into the m.dot Q delta.t equations yet..."​

I don't know what other data to provide. Even without doing quantitative data analysis it's pretty simple to tell "omg. It's ******* hot!" ...and then proceed to roll all windows down and get a blast of fresh cool air lol. All jokes aside...I already know my bottleneck. My heat exchanger is way to small and fans are not powerful enough (also need more of them). If I could get a heat exchanger to cover the entire lid of the cooler with maybe 9 fans running then yes it would probably make 20-30f difference instead of only 5-10f difference. However, heat exchangers are expensive and I don't plan on spending any more on the project other than maybe using an old life jacket and running water lines through it to cool me