Slow Flow Air Pump With Zero Vibration

How about the exhaust-fans being used computer CPU?

They through lots of air and are not pumps. Good thinking. These fans run fast so airflow may be somewhat smooth. Agreed.

I am looking for veryfast electromechanical air pump with small air flow of few L/m and some way to connect tubes to it.

An aquarium (fishtank) air pump is extremely quiet and because of the cleverly designed feet has very little vibration, it is a diapragm pump which can be adjusted in flow rate between 1 litre/min to over 3 maybe 4 litre/min in free air...

They can also generate 2 to 3 psi of pressure.

John.

Any good source of these pumps? In India we can not get noise free pumps.

Just do an internet search for "aquarium supplies" and find thousands of sources.

If the pulsations from the diaphram pump ar objectionable, feed the air to a tank first (it doesn't need to be very large) and then bleed off your air.

Hi Shyam, some years ago I saw something about generating a flow of air from an electric potential. It worked as it shows in the diagram, when you put a high negative voltage on the pin and positive voltage on the plate you get a considerable flow of air in the direction of the blue arrows. I can't remember the principal behind it but I know they were experimenting with it as a method of cooling microprocessors. I had a look to see if I could find anything on the internet about it but to no avail. Sorry I can't remember any more and I have never tried this myself but it might be worth an experiment or two.

Masu, that principle looks very similar to an 'ion pump'..

Basically an ioniser with a target, one of those negative ion generators has this effect, the ions 'released from the sharp point create a stream or flow of molecules away from the charged source... I played with something similar in the late 70's but the air flow was very low and the pressure barely perceptable... When ever I tried to get a decent flow it led to corona discharge and dangerous voltages!!

John.

I agree that flow is usually small. Such pumps can be used on gases, fluids, and solids that have static charge capability.

Dear Masu

Very intelligent reply indeed. Yes, this is what is used as ion spray in ion mobility spectrometers and also in the powder coating machines. Once ionized, the ions gain flow kinetic energy as well.

Shyam, I had a further look on the internet and found something called a corona_cooler. It works much like the cooling method I remember seeing and thought it may be of help.

Dear Masu

Do not ever try that for the chips. The impulse current in corona discharge generate EMP and inductive field so high that it easily burns thin wires of the chip. I used this technique in 1976 (is part of my Ph.D. thesis) to test the ruggedness of the chips with corona discharge impulses and many EPROMs were damaged. I then used Fuse type PROM which worked. This is a good way to test if your chip can take EMP. You may very well end up with result and no-chip.

Corona also generates large amount of UV and that will kill the brain neurons very easily. I was losing my memory during these experiments and then looked for permitted levels of O3 in air and was shocked to know the ppm level being toxic to brain. I knew the reason then and fortunately everything was not lost and I still knew my name. You can easily find out that welders who use electric discharge method are often dumb people. That is what remains of them.

Air discharge generates about 10eV UV photon, which we also call vacuum UV. It can give cataract easily but will show up 5 years later due to latency period involved.

What pressure? Does it have to be a positive dispalcement type pump? Aquarium pumps are high vibration pumps. Are you pumping air or some other gas? Small twin lobe rotor pump feasable? Small is relative. be more specific please.

What I need is for pumping drift gas into Ion Mobility Chamber with 1L/m to 5L/m at best. 12V battery operated pumps are used and perhaps some mechanism to damp the flow oscillations is required. Flow is atmospheric pressure and perhaps slightly more pressure as some filters are inbetween to remove moisture.

I recently engaged a test stand design that required slow flow, slow pressure, cleaniless (uses almost pure oxygen, so, no dirt or "cabum!"), and in our case it had to have a cycling work much like a human breath.

Al last, we built a custom piston pump made on teflon that runs at very low speed, is clean and is capable of giving us positive and negative pressure as required. Add a pressure relief valve and/or a pressure regulator and your mechanism is set.

If you build the cylinder big enough and uses a motor and a reduction gear small enough, you can adjust the speed and pressure to a quite steady state work, specially if you include some accumulator (not really big) at the outlet.

I do not know if it solves your problem, but it's an approach.

I will try some reservoir and then release linearly. It looks better idea. Drift gas is dry N2. Teflon can not be used as it can generate static charge. Idea is to inject neutral gas and let small ion current drift through it. Ion current is only pA level and any other charge generated will be of great problem.

Ok, if your process is not a stedy state one, and if you are able to do so, there's always the alternative to install an accumulator and release pressure through a pressure regulator to your system. Easy, and clean. The only requirement would be to charge the system from time to time.

IMS gas injector byitself is a reservoir. I am worried about its uniform flow in the chamber. gas is often injected from one corner using small pump. If gas is not flowing uniformily then drift time will differ. Ions flow opposite to drift gas and towards Farady cup.

So, I'd change to the following design. This would give you steady state and almost continuous process. Any method of cylinder alternation (even manual) would fit well. If you do not find a pump... do not use it!

Regards

bhre@brturbo.com.br

My signal frequency is from 100Hz to 100kHz range. If this nozzle arrangement cause some kind of vibrations or standing wave then will have serious problem.

Idea is good one so perhaps I will use some diffuser design but holes may not be good choice. They may generate sound waves. Perhaps some kind of sponge like damper may be good, but i am afraid to put anything like that inside.

The flow is so low, it doesn't seem to cause any appreciable vibration effect. But I am not used to the process you are trying to build. If this kind of device inside the chamber is not desired, you can still just connect more N2 tubes around the wall behind the electrode instead. This will not give you the same flow homogenity than a "flow straightener" or diffuser would give, but would be better than just one hole breathing.

Just a quick not and I are sure you are aware of this but, if you use an accumulator or storage tank with a pressure regulator, you will need to connect the sensor inputs of the regulator as close as possible to either side of the outlet restrictor. The flow through the system will be the same no mater where you are after the regulator and will be proportional to the pressure drop across the outlet restrictor. Since its this pressure drop you need to regulate the closer you can connect the inputs the better you will be able to regulate the flow. If you try and regulate the pressure in the test chamber or at the outlet of the regulator you may not get the accurate response and control of flow that you require.

I agree. However, the pressure regulator here is placed with the unique purpose of allow a leakage type flow from the high pressure bottles. As far as I see, our friend here is not concerned about the chamber pressure or the outlet regulated pressure (he describes the pressure as ambient and the flow order of magnitude is 5 l/m). I don't think he's gonna need a special installation - an off-the-shelf regulator would fit. I seems that the supply connection to the chamber is more important, to avoid disturbance of the process, and the pressure regulator will work as a flow regulator, keeping the supply pressure constant while the bottle pressure drops over time. As you wrote about the pressure drop and the flow regulation, if the supply pressure from the source was constant, you'd need only a needle valve to keep its value.

In IMS design, the drift gas flows one way and the ions flow the other way. There are also pockets of cavities near side walls due to electrodes inserted in the tube. Unless flow is smooth, the ions will not move smoothly. Ions move in just 40ms the entire tube, and during this time the pressure, and disturbance will affect their travel time which is information.

By the way, what's the lenght and diameter of the chamber?

It seems by your requisites that you really need a flow straightener, or some geometric feature that guarantees your flow completely developed and laminar.

What about extending the chamber behind the faraday cup?

Length about 8" and effective dia. about 1". Drift tube often has projected electrodes in the shape of circular disk to keep the electric field uniform in the drift volume. This also may work as damping source. There is Faraday cup sloe to the gas inlet to collect charge from ions. Drift gas removes molecules that are not having charge or were neutralized.

Ok, I see we are talking about the same solution in different words. Well, I wouldn't have further suggestions except to distribute the gas inlet in 6 or 8 inlet ports around the cup and maybe distribute the gas exit as well. This geometry would make the gas flow more uniform along the tube cross sectional area, and there would be no need to insert any other part inside.

Hi Shyam,

Turbulance and speed of the drift gas are not a conceern, unless they are very, very high.

Tad

Dear Tad

I agree. If drift time is too small, which I am keeping at about 40ms maximum, the turbulence or flow rate change very unlikely to cause serious changes in the ion motion in this period. However, some improvement in peak shape may be possible if smooth flow is achieved. With electrodes project in the chamber, the turbulence in flow is unavoidable. Only smooth drift tube with continuous coating will work have least turbulence in the drift column.

That dual gate FFT technology sounds very interesting. I have read few research papers on that. I am afraid, it may some time give very falls information as well due to processing program incapabilities. Have you come across any such things?

Remember in this instance you aren't really building a pressure regulator you are building a flow regulator, it just happens that the pressure regulator is working like a flow regulator. By the way you can get pressure regulator/controllers like this off the shelf.

Dear Masu

Your points are valid.

Both flow and pressure play a role. Remember that electrophoresis which uses gel for drift media. Pressure also limits the mean free path and one gets the same result but we are talking of flow as well here to remove the recombined ions.

Shyam,

All portable IMS devices use standard diaphragm pumps to provide carrier and drift flow, in a recirculating loop. If you are using a molecular sieve drier, this will usually provide sufficient pulsation dampning. If pulsations are still a problem, staggering data aquisitition time (random variable delay time between aquiring spectra) and averaging will eliminate the standing wave problem.

If you are not recirculating the flows, a compressor and accumulating tank will eliminate all pulsations

Teflon is not a problem. All tubing leading to my IMS is Teflon, and the drift tube itself is coated with Teflon.

Tad

Dear Tad

Very interesting ideas. How do you get a tube coated by Teflon from inside? Is there a standard supplier for it?

The averaging also broaden the peak. The molecular sieve damping idea looks better. Do you manufacture these or use for research? Where are you located? Can I meet you? I have a plan to travel to USA so can make a trip to other places as well. I will like to see some of such designs. Perhaps a joint research work may also be possible.

I am also thinking to use Burle 1mm hole drift tubes. These will hardly require any great air flow volume and do not know if they will work. Right now the drift column dia is 1" and dead area another 1" in 2" diameter tube.

I happened to see this link recently, might help:

www.knf.com/oempiston.htm?source=globalspec0305&vid=92741&sid=10317