The motor was picked from a 1.7cuft GE Mini Fridge. I planned on using the motor for the compressor build, and the actual refrigerator case for paint storage, so no wasted parts. The refrigerator was damaged during a move I was told and the R134a had leaked out of a visible crack in one of the copper lines. This saved me from having to recover the gas to pull the motor, bonus. I cut the lines with a small hobby sized tubing cutter (very small, fits in tight places well for this size copper line) and pulled the compressor motor out of the back of the refrigerator. I drained the oil left in the compressor and did a quick flush with fresh oil and topped it off at about 5.5oz of new oil. A few quick tests of the motor showed it was working well, and very silent.
I needed a small air tank for this build, and I considered a pair of paintball CO2 tanks (40oz total) but settled on a lighter and cheaper alternative, propane tanks. The pair of CO2 tanks would have run me close to $50 for the pair new, where the propane tanks ran me $5.50 for the pair. But now I had to use up the propane to empty the tanks. On go the steaks, and a good self cleaning of the grill and the propane was gone. I removed the small Schrader Valve from the center of the tank in preparation for drilling and tapping. I needed to drill the neck out to 7/16" for the 1/4" NPT tap. I used progressively larger bits and lots of cutting oil while drilling and tapping.
Some people say you cannot use propane tanks for air as you'll never get rid of the oderant that is used in the propane to help detect leaks. This is false. A couple flushes with Dawn dish washing detergent and some Simple Green and the odor is completely gone. After a few good washes to get rid of any oil residue left over from the drilling, the tanks were baked dry out in the summer sun here in Arizona. A thinned mixture of rust prefentative paint was put in each tank and rolled around to coat the entire inside of the tanks. This was set to bake dry again in the sun for a day.
Another trip to Ace Hardware got me everything needed to build the bracket needed to mount the tanks to the compressor and hold the regulator and switches. I built a simple frame by welding some 1/8" x 1.5" flat stock together with a riser for the regulator mount. I then welded the bolts needed to mount the compressor onto the bracket. This is all held together to the tanks with some large hose clamps to make everything more servicable if needed. Between any metal to metal location I used some 1/4" foam rubber to keep all vibrations and noise to a minimum.
I mounted all the air fittings and plumbing as compact and serviceable as possible. I used the same model parts as the previous build, as they have proven to be reliable. Condor MDR-11 4-port Pressure switch with unloader, one-way check valve from Grainger, Water trap and regulator, handful of fittings from Ace Hardware, and a roll of Teflon tape. The tank pressure gauge is mounted to one of the ports on the switch with a 1/4" to 1/8" fitting. Regulator and quick release on another port, all coupled together with 1/4" male/male fittings. The 115psi safety pop-off is mounted to one side of a 1/4" T fitting in the remaining port of the switch. 90 degree fitting goes to the one-way check valve with a 1/4" male/male fitting. The check valve has a 1/8" unloader hook-up on it connected to the unloader on the pressure switch with 140psi nylon tubing, and a compression fitting. The other side of the on-way check valve is connected to the output of the compressor motor with a couple of compression fittings and high pressure hose. The tanks are plumbed with 140psi tubing and compression fittings. Water build up is not a huge issue here in Arizona, but I will be keeping an eye on these tanks, as I do not have a petcock installed for tank drain. I may need to disconnect the tubing and drain that way if needed.
Some pictures of the final compressor:
The pressure switch is has 20-105 psi adjustment range with 20-45 psi adjustable differential and 26 Amp capacity with 4 1/4" ports and unloader valve made by Condor, model MDR-11.
The unloader valve is needed with these smaller compressor motors when trying to start them under high pressure. With the 50psi cut-in setting I'm using, I could get away with not using an unloader as the compressor will restart just fine at that pressure, but the unloader keeps the stress of that initial start up pressure from harming the compressor at all.
All air fitting connections were teflon taped to help with any leaks. Make sure to use nylon compression fitting inserts with nylon tubing. I made the mistake of using the brass inserts on the first compressor build and had leaks everywhere as the brass fittings cut into the nylon tube.
I also had to "tune" the one-way check valve from its original configuration. During the first compressor build mock up when I fired up the compressor for the first time with the check valve inline the check valve was making more noise than the compressor motor itself! There is a nylon block in the valve on a lightweight spring that opens on positive pressure from the input and closes when there is no more front pressure. The problem was the spring was just a little to heavy for this small compressor motor and made a constant slapping noise, almost like a fart, and was 2-3 times louder than the compressor motor. I ended up trimming 2 coils off the spring to lighten it up and its quiet as can be now, but still functions to hold back all the pressure from the tank.
I made a simple and small air filter for the input line much like the filter on previous build. I took a "Yorker" cap from a squeeze bottle and cut the nozzle down until it fit over the input tube on the compressor motor, and drilled a breather hole on the side of the cap. I found a cap from something in my junk drawer that fit over that yorker cap to make it enclosed, and also drilled a matching breather hole in that. I filled the newly made "filter" with stacked filter material from an Air Conditioning filter and pre-filters from an old 3M respirator. Stacked alternating with AC Filter, 3M filter, AC, 3M etc until the cartridge was full.
The new compressor specs are:
- 1/6? - 1/8? HP Motor
- 115psi Max.
- 2 - 16.5oz tanks (33oz total)
- Weight 21 lbs
- 30db Noise Level
- Operating pressures 50-85 psi
- 35 psi differential, cut-in 50 psi, cut-out 85psi
- 1 min 45 sec empty to 85 psi
- 50 psi to 85 psi fill takes 45 sec
The new mini compressor is very, very quiet and works great. Couple coats of Hammerite Slate Blue and Black to give it some color like the previous build. This motor works a bit harder than the larger motor to keep up with the air demands, and I think I'll be adding a 45db 120cfm 120mm fan to the unit to keep it cooled off for extended use.
Here is a pic for size reference with a painted bowling pin.
Total cost if built from scratch with new parts would be about $110-$120
Here is a video of all the silent compressors running side by side for a noise level comparison:
Adding a Fan:
I've used this compressor now on a few jobs, and it works great. The only possible issue I can foresee is heat with this little motor. Although I've never brought it even close to a thermal shut down, I don't want to cause un-needed stress and damage to the compressor if it can be avoided.
I've been running it with a small desk fan pointed at it and it has kept nice and cool, without it, it does warm up a bit after 10 cycles or so. So I ventured out to the local Fry's Electronics and picked up some goodies to help keep it cool.
Fry's had a 120mm 115v Fan (out of stock though) that puts out about 65-70cfm @ 40db. That would be enough cooling, but 10db louder than the damn compressor itself, that wont do...
So I kept looking, branching out to the 12v Case fans and found a perfect fan. SilenX 120mm 12v .2A Fan. 6-14db!! 24-72CFM with thermistor. The thermistor will be a perfect bonus on this setup as it will speed up the fan to full power as the temp rises, and slow it back down when cool.
Ahh, but you say.. its 12v, and the compressor is a 115v system... This was solved with a small 115v to 12v wall wart adapter for charging stuff. I found the smallest one there that had enough Oompa Loompas to handle the current draw of the fan at full power (200mah). The plan was to pull the adapter out of the case and wire it into the compressor directly so I have 12V on-board. No extra plugs etc. I was able to take the components out, rewire to fit my situation, add an on-off switch and fit the whole bundle inside the current pressure switch! I added a micro toggle into the pressure switch case so the 12v power can be cut when shutting down without having to unplug the compressor to shut off the fan.
Couple holes drilled in my riser that holds the regulator and pressure switch and the fan was mounted on the side of the compressor. This pushes the air past the entire height of the compressor motor, especially the top (where the most heat is generated from the hot oil slinging on the case).
I did some testing and there was a significant difference in the temp from the fan cooled side and the opposite "Dark Side"... After running solid for 30min without shutting off, the compressor was still cool enough to put your hand on.
Here are some pics, as always, click to enlarge.
Temperature Testing:
So I setup a test to see what the temps were over a time frame. I used a digital thermometer (no contact type) measured every 5 minutes of continuous run. The temps were measured at the top of the motor enclosure as it was a good spot that didn't get any direct fan blowing on it. I setup the compressor with a constant leak that kept the compressor running without switching off by reaching cut-out pressure. The leak was set to keep a constant 40psi in the tanks and continuous 25psi drain. This ensured the motor had a load on it (fighting 40psi).
The room temp was 78f. There was no airflow in the room from other fans etc to keep a controlled test environment. Start temp of the compressor at measurement site was 80f for each test. The compressor was allowed 2hours to cool back down to 80f before the 2nd (fan on) test was done.
The first set of results here is the temps without the fan running.
| TIME | TEMP | INCREASE |
|---|---|---|
| 5min | 90.8f | 10.8f |
| 10min | 103.8f | 13.0f |
| 15min | 115.4f | 11.6f |
| 20min | 123.7f | 8.3f |
| 25min | 132.3f | 8.6f |
| 30min | 139.2f | 6.9f |
So 30min continuous run-time = 140f or thereabouts.
I cooled the compressor back down to the baseline of 80f and started the same test with the new fan on.
| TIME | TEMP | INCREASE | TIME | TEMP | INCREASE | |
|---|---|---|---|---|---|---|
| 5min | 89.3f | 9.2f | 35min | 115.7f | 1.5f | |
| 10min | 99.1f | 9.8f | 40min | 117.3f | 1.6f | |
| 15min | 105.3f | 6.2f | 45min | 118.9f | 1.6f | |
| 20min | 108.5f | 3.2f | 50min | 120.2f | 1.3f | |
| 25min | 112.0f | 3.5f | 55min | 121.5f | 1.3f | |
| 30min | 114.2f | 2.2f | 60min | 122.6f | 1.1f |
After 60min I was still at a lower temp than the 20min mark without the fan, and the 5min temp increase was getting down to almost 1f for every 5min!
So thats 3x the duty cycle as far as the heat test goes.
Another impressive number that I noticed was the post run temp decrease with the fan on of almost 15f in 5min. Without the fan the decrease was much slower at 3-4f in 5min. So the fan will help to keep the motor cool when its running, and even cooler between cycles.
If you have any questions or comments, please feel free to contact me or comment below.
I get a lot of requests for part numbers, I can only give you the part numbers I used from Grainger, I can't help you with some other company in your country, or alternatives etc.
*Grainger Part Numbers
- Check Valve 5X780
- Pressure Switch 3EYP2