Longevity of refrigeration compressor as a vacuum pump?

[DTJ]

Do we have any refrigeration techs here?

I've got a compressor out of a small domestic refrigerator that I've re-purposed as a vacuum pump for my manual pick and place pen.

See pics attached.


I ran it for a while until most of the oil spluttered out the exhaust line. I've run it for a few hours and it still runs whisper quiet and pulls ~27"Hg vacuum.

Any ideas on how long this will last running with just the residual oil inside the pump?

[noidea]

Hi DTJ
Unfortunately its a "how long is a piece of string question"

Hermetic refrigeration compressors are designed with two things in mind, one they will always pump a small amount of oil out of their discharge as a refrigeration system is sealed and it circulates through the system and makes its way back to the compressor eventually, how much is still in the sump of yours is a mystery.

Secondly they are also designed to be cooled by the flow of the refrigerant gas back over the windings of the motor inside the can, in most cases this is going to be below about 10 degrees C or if from a freezer below 0 deg C so running them with ambient air being drawn in puts them behind the 8 ball. Yours with the four pipes coming out of it indicates that it probably came out of a freezer as the extra ports are an oil cooler loop where the refrigerant gas passed through to further cool it further.

The compressor label indicates it was designed for R134A refrigerant which would have probably been charged with POE Polyolester oil which is very hygroscopic and sucks up moisture which it then binds with and starts to form acid.

All in all it is probably best treated as a disposable unit, some sort of separator / coalescer on discharge would help trap the oil that will continue to pump out of it which you could then pour back into the intake port periodically. Some way of filtering the intake would help and not running it too long will also help so it doesn't overheat

[Mechatrommer]

http://www.randombytes.org/vacuum_pump.html
http://www.belljar.net/refrig.htm
http://www.belljar.net/fbleeconversion.pdf

[Gyro]

Maybe fitting a sealed screw type oil filler port in the top of the canister would be the answer (leaving the port open would probably seriously reduce the vacuum it could pull and increase the splutter rate). You'd have to fit a permanent oil collector on the outlet but you could then recycle the oil back into the compressor when it starts to run low. The act of pulling a vacuum ought to provide some degree of cooling.

Hopefully you already have a non-return valve on the inlet pipe!


Edit: Mechatrommer's links seem to have it covered.

[BradC]

Tip it up and pour the oil out of the suction line (that ties straight into the can). Measure how much came out.

Grab the data sheet for the compressor and find out how much oil it holds. Fill it back up through the suction line with ISO100 air compressor oil.

Do this periodically and it'll practically last forever, with the lifetime pretty much defined by how hot it gets. If you're not running it until its too hot to touch, and you're not sucking up noxious chemicals it'll pretty much last longer than you want it to. These compressors rely on refrigerant flow to move heat, and pulling a vacuum gives you very little flow, so in this application it's all about thermal inertia and convection around the can.

Get the POE oil out of it ASAP.

[Zero999]

Tip it up and pour the oil out of the suction line (that ties straight into the can). Measure how much came out.

Grab the data sheet for the compressor and find out how much oil it holds. Fill it back up through the suction line with ISO100 air compressor oil.

Do this periodically and it'll practically last forever, with the lifetime pretty much defined by how hot it gets. If you're not running it until its too hot to touch, and you're not sucking up noxious chemicals it'll pretty much last longer than you want it to. These compressors rely on refrigerant flow to move heat, and pulling a vacuum gives you very little flow, so in this application it's all about thermal inertia and convection around the can.

Get the POE oil out of it ASAP.

That seems like sensible advice. You may also have to flush it with several lots of compressor oil to get rid of the POE.

[BradC]

That seems like sensible advice. You may also have to flush it with several lots of compressor oil to get rid of the POE.

I meant to say also as you are in Aus, grab 1L of cheap compressor oil from Supercheap or Repco. That'll pretty much last you forever.

Yeah, a couple of flushes with mineral won't hurt to get the POE out, but to be honest just getting most of it out the first time will probably do the job. Tip it up so the suction port is at the bottom, and leave it to drain overnight. Do it when it's good and hot will make it go faster too.

The process tube (the crimped and soldered one) is a second line into the can, so you can cut the end off that to let air back into the can to help you when you do oil changes. You'll just want to plug it up again for vacuum duty.

A cheap and nasty oil separator for the discharge would be something like an aluminium can stuffed full of steel wool. Put the discharge pipe down the hole in the top almost to the bottom of the can, poke a couple of *small* holes in the very bottom to let the oil drain out and most of the vapor will get trapped on the wool on the way out, taking most of the oil out of the discharge . If you put the can over a container, you can measure the oil as it comes out and you'll know when and how much to put back in periodcially. Don't re-cycle it. Compressor oil is too cheap to stuff around with.

[HighVoltage]

One of the best oils for vacuum pumps is "Anderol-Oil type 555" and this oil works extremely well for this type of vacuum pump as well.

I have done the procedure described above by BradC for changing the oil to the 555 and the pump is running for years already without any problems.

[MagicSmoker]

Personally, I'd ditch the refrigeration compressor and use an aquarium air bubbler pump instead.

[CM800]

Could buy one of these maybe?

http://www.ebay.co.uk/itm/220V-Vacuum-Pen-Suction-Pen-Smt-Smd-Repair-Air-Vacuum-Pump-Pick-And-Place-B-/381999483511?hash=item58f0f18a77:g:xw4AAOSwJ7RYThsl

[Mechatrommer]

Personally, I'd ditch the refrigeration compressor and use an aquarium air bubbler pump instead.

or this...
http://www.ebay.com.my/itm/DC12V-Mini-Vacuum-Pump-Negative-Pressure-Suction-Pump120kpa-5L-min-With-Holder-/222259792486?hash=item33bfb74a66:g:mncAAOSwo4pYEEju
i bought 2 of these in preparation for my next PnP project. easier to embed in 12-15V system, easier to control through mosfet switch, space saving etc. from my experience, the suction power of this (by putting finger on the suction tap) is much more than aquarium air bubbler (which i also have one laying about for years already). its just a bit noisier than the refrigerant compressor. and refrigerant compresor should last longer than this small oiless brushed motor.

[NiHaoMike]

Personally, I'd ditch the refrigeration compressor and use an aquarium air bubbler pump instead.

Not even close in terms of how much of a vacuum can be pulled.

Since the compressor is running at way less than its design load, reducing the voltage by 20% or so using an autotransformer can let it run cooler.

[Mechatrommer]

if just there is an easier way to run mains line at 80%

[MagicSmoker]

Personally, I'd ditch the refrigeration compressor and use an aquarium air bubbler pump instead.

Not even close in terms of how much of a vacuum can be pulled.

Okay, sure, but how much pressure differential and flow rate do you need to pick up a damn SMT part, anyway?

I have a manual P-n-P system at my home lab for assembling prototypes (paid for itself many times over already) and it uses an aquarium pump for the vacuum pickup. It works perfectly fine, even picking up relatively heavy parts like 10mm aluminum electrolytic capacitors (need the rubber suction tip for that, of course).

[Zero999]

Since the compressor is running at way less than its design load, reducing the voltage by 20% or so using an autotransformer can let it run cooler.

It will have an induction motor which will attempt to turn at just below the mains frequency. Reducing the voltage and not the frequency will increase the current draw and it may struggle to start, causing it to overheat. It may work if you reduce the both the voltage and frequency to 20% but there's no guarantee it will work.

[CJay]

Yuo can buy parts designed for use with R134A aircon systems, particulate separators, filter driers, oil traps etc. Not terribly expensive but by the time you've bought them and plumbed them all in you're in the region of a decent second hand vacuum pump (a robinair, again an aircon service pump) so it's a value judgement as to which route to choose.

Personally, if I had easy access to scrap/free refrigeration units and needed a vacuum pump that could pull that sort of level, I would most definitely give it a go.

[james_s]

There is actually a line of commercially built super quiet air compressors that use refrigeration compressors. They are modified with a removable top on the hermetic dome and have a threaded dipstick to check the oil level. So long as you keep oil in the sump, a refrigeration compressor should last a good long time. They have a spiral groove in the shaft that pumps oil up from the sump and centrifugal force sprays it over the top of the compressor. One issue remaining though is that the compressors are normally designed to be cooled by a small amount of liquid refrigerant returning from the evaporator but with sensible duty cycles you can usually get away with running one open loop. YMMV but I used a refrigeration compressor as a small air compressor for years.

[calexanian]

First thing. What are you doing with it? 27 inches of vacuum is not that much. Additionally without oil it will not last very long.

[james_s]

Who said anything about running it without oil? Keep oil in the sump and they will stay lubricated. Might need to add a little now and then but it's pretty easy to rig up a primitive oil separator on the discharge line so most of the oil will condense and drip back down into the sump.

[noidea]

One issue remaining though is that the compressors are normally designed to be cooled by a small amount of liquid refrigerant returning from the evaporator but with sensible duty cycles you can usually get away with running one open loop. YMMV but I used a refrigeration compressor as a small air compressor for years.

It's not worth arguing on the internet but they are designed to pump vapour only. Liquid return to the compressor has its own set of issues.

[DTJ]

Wow, a flood of useful info guys.

As per Noidea & Bradc, I drained the balance of the oil last night. It's clear and has very low viscosity. I only got ~1cc out of it.
I've looked online at some compressor oil from local auto spares. I wonder if it will be too thick and I should just put some light machine oil into it?
I'll try do do this slowly to avoid potential "hydraulic locking" that noidea picked up on.


I had neglected to put a non-return check valve on between the compressor vac side and the rest of the gear - Gyro's comment prompted me to run out and check to see if it had back flowed a pile of oil into the solenoids / reservoir tank. It hadn't. I had ordered a check valve with with some solenoids a few weeks back and fitted it.


I'm going to put a catch can on the outlet with one of those stainless pot scourers in there to trap any oil mist.



We have 6 monthly hard waste collections in our area so there is no shortage of fridge compressors - I'll see how long it lasts and if it does fail I can grab another one easily enough.

The up side of these pumps is they are very very quiet (and free).


As mentioned previously its for a manual pick and place set up. I'll post up some pictures of the final set up when it's done.

Thanks for all the ideas guys.

[BradC]

Wow, a flood of useful info guys.

As per Noidea & Bradc, I drained the balance of the oil last night. It's clear and has very low viscosity. I only got ~1cc out of it.
I've looked online at some compressor oil from local auto spares. I wonder if it will be too thick and I should just put some light machine oil into it?
I'll try do do this slowly to avoid potential "hydraulic locking" that noidea picked up on.

I'll guarantee if you punch the model number into google and dig up the data sheet it'll take a *lot* more than 1cc of oil. More like 100cc.
If you are worried about oil viscosity, use air-tool oil instead. It's ISO32 and much lighter than pretty much most refrigeration oils. You'll spew a lot more out though as the lighter oil will get whipped up into a thicker mist which will get ingested and ejected. Again, the manufacturers data sheet will tell you what oil it's supposed to take if you are really concerned. All of the reputable compressor manufacturers have comprehensive data on their devices.

Don't worry about hydraulic locking. The suction tube empties into the shell, the actual intake for the compressor is right up the top of the shell. You could half fill the shell and not worry about a hydraulic lock. What you'd to do the rest of the mechanism with drag is another story altogether.

noidea is right. These compressors are designed for vapor return. Having said that, a small amount of liquid return is generally survivable as it will evaporate inside the shell. It will generally result in oil foaming and cause lubrication issues though, so it's a big no-no in a functioning system.

They are designed to cope with liquid, and if you read the compressor data sheet it'll give a maximum system charge. This is the volume of liquid refrigerant that it'll tolerate in the compressor shell at startup (designed for those one in a million instances when the compressor is the coldest part of the system and all the liquid migrates there). Usually in excess of a hundred g or so which for r134a works out somewhere around 80ml. That is less than the oil volume as the liquid refrigerant dissolves in the oil. When the compressor starts up the refrigerant starts to boil out vigourously and the oil pretty much fills the shell with foam. More liquid than specified can lead to damage through hydro-lock of the thick foam or bearing damage due to foam not lubricating properly. Always good to read the data sheet when building a refrigeration system

[DTJ]

I'll guarantee if you punch the model number into google and dig up the data sheet it'll take a *lot* more than 1cc of oil. More like 100cc.
If you are worried about oil viscosity, use air-tool oil instead. It's ISO32 and much lighter than pretty much most refrigeration oils. You'll spew a lot more out though as the lighter oil will get whipped up into a thicker mist which will get ingested and ejected. Again, the manufacturers data sheet will tell you what oil it's supposed to take if you are really concerned. All of the reputable compressor manufacturers have comprehensive data on their devices.

Cheers Brad. I found a data sheet but it does not mention oil capacity. There's no doubt a more detailed data sheet somewhere.
I've got some air tool oil. I'll try 100ml of that in it.

http://www.transtech-co.jp/electronics/PWVK_ES_1.pdf


If I get 40 or more hours out of it I'll be happy.

[BradC]

Cheers Brad. I found a data sheet but it does not mention oil capacity. There's no doubt a more detailed data sheet somewhere.
I've got some air tool oil. I'll try 100ml of that in it.

http://www.transtech-co.jp/electronics/PWVK_ES_1.pdf

If I get 40 or more hours out of it I'll be happy.

Nice job even finding that data sheet to be honest. Not a 1st tier compressor manufacturer. Check out the top level domain for a laugh : www.transtech-co.jp. I did a quick comparo and something like a Secop TL5G is vaguely equivalent +/- enough not to worry about.

http://www.resluk.com/images/pdf/compressors/danfoss/tl5g.pdf

Capacity, displacement and dimensions are vaguely close. Certainly for this kind of thing. That takes 280ml of oil, so 100 will do the job, but keep an eye on the level. As long as you don't cook it, you should get many more than 40 hours out of it.

[james_s]

It's not worth arguing on the internet but they are designed to pump vapour only. Liquid return to the compressor has its own set of issues.

They are designed to pump vapor only yes, but the compressor is cooled by a small amount of liquid refrigerant that returns to it, pours over the motor and vaporizes well before entering the intake of the compressor head. Too much liquid refrigerant returning to the compressor is bad because liquid does not compress. Too little liquid returning to the compressor is bad because the compressor overheats. This is one reason the quantity of refrigerant in the system is somewhat critical, particularly in fixed orifice systems that lack a TXV to regulate the flow to the evaporator. The discharge line comes straight off the compressor outlet but the intake is inside the hermetic dome and to suck liquid into the compressor would require enough liquid returning to fill up the dome up to the level of the intake.

While not my day job, I am licensed as a refrigeration technician and installed heat pumps and central AC systems on the side for a few years. There was significant discussion of this in the class I took to get certified.

I probably should have been more clear though, I realize I've referred to "the compressor" both in cases where I'm referring to the whole assembly and in cases where I'm referring to the actual compressor inside the dome. Obviously you don't want to suck liquid into the compressor itself.