How can I build a better pulsed pressure generator?

[Joerg]

Folks,

I need to generate pressure pulses of 3-4psi above ambient, frequency
around 1Hz, more or less arbitrary waveform with frequency components up
to 40Hz in there (but those are less than 2psi) and delivered via short
hose. Currently just air pressure.

I got this (sort of) running with a compressor but after a while my ears
fall off from all that noise. Can't really place the compressor outside
here. So then I tried various speakers. They don't move enough air
around. When I pushed one of them a bit harder a few minutes ago its
flexible surround rubber burst, that thin stuff doesn't hold much pressure.


In other words I need something similar to an arbitrary function
generator but where the output is air pressure. What I want to avoid is
hacking a motor, removing a valve and using the piston.

Any other ideas? Can one buy something like that at less than 4-digit
prices?

 

[Lasse Langwadt Christensen]

Den onsdag den 18. september 2013 22.17.29 UTC+2 skrev Joerg:

Folks,



I need to generate pressure pulses of 3-4psi above ambient, frequency

around 1Hz, more or less arbitrary waveform with frequency components up

to 40Hz in there (but those are less than 2psi) and delivered via short

hose. Currently just air pressure.



I got this (sort of) running with a compressor but after a while my ears

fall off from all that noise. Can't really place the compressor outside

here. So then I tried various speakers. They don't move enough air

around. When I pushed one of them a bit harder a few minutes ago its

flexible surround rubber burst, that thin stuff doesn't hold much pressure.





In other words I need something similar to an arbitrary function

generator but where the output is air pressure. What I want to avoid is

hacking a motor, removing a valve and using the piston.



Any other ideas? Can one buy something like that at less than 4-digit

prices?

solenoid or servo motor and a pneumatic cylinder?

-Lasse

 

[Joerg]

One of those old style Wagner paint sprayers with the vibratory motor?

(I think the new ones actually have a turbine

Those are too wimpy. They can't move this much air around. They can
shoot out a quart of paint in 10-15 minutes (unless the motor burns up)
but I have to move 1/50th of a gallon of air around in less than a second.

Speakers with a big enough amp can do this but they suffer seriously
because they are meant to be operated against a fairly rigid air cushion.

 

[Joerg]

Den onsdag den 18. september 2013 22.17.29 UTC+2 skrev Joerg:

solenoid or servo motor and a pneumatic cylinder?

Something like that but I've not seen any solenoid-cylinder combination
that is this fast. They all had too much mass and were designed for much
higher working pressures (Festo et cetera).

 

[Don Y]

Hi Joerg,

I need to generate pressure pulses of 3-4psi above ambient, frequency
around 1Hz, more or less arbitrary waveform with frequency components up
to 40Hz in there (but those are less than 2psi) and delivered via short
hose. Currently just air pressure.

I got this (sort of) running with a compressor but after a while my ears
fall off from all that noise. Can't really place the compressor outside
here. So then I tried various speakers. They don't move enough air
around. When I pushed one of them a bit harder a few minutes ago its
flexible surround rubber burst, that thin stuff doesn't hold much pressure.

In other words I need something similar to an arbitrary function
generator but where the output is air pressure. What I want to avoid is
hacking a motor, removing a valve and using the piston.

Any other ideas? Can one buy something like that at less than 4-digit
prices?

If you didn't care about the "waveform", I'd have just suggested
driving a small "lawnmower" engine with an electric motor
(plumbing your hose to the openings for the plug(s)). Possibly
adding a pressure valve to ensure a greater "leading edge"
if need be...

Should be simple to experiment with (assuming you have the time
and an old lawnmower/snowblower/chainsaw/etc. handy)

 

[Tim Williams]

Try taking apart one of those really small shitty subwoofers... long-throw
speakers on the order of a few inches diameter. Turn the enclosure from
ported (they usually are) into fully sealed and you get rid of the back
wave; arrange suitable means of coupling on the face side.

Assuming AC coupling is fine, of course. If you need DC, you could always
bleed in bias from a servo valve and PID.

It would take some work, but you might be able to hack a hard drive head
motor onto an especially nice ball valve (normally, they're rather sticky
from the o-rings; otherwise, if you can salvage a strong head motor from a
1980s hard drive..). Add encoder for positional feedback, and calibrate
output for the nonlinearity of the valve (or feedback loop with a nearby
pressure sensor).

Tim

 

[Joerg]

Speaker driving into a chamber with only outlet your tube? Remember
piston area "gain" from Freshman mechanical engineering electives?

That's exactly how I blew up two of them. No smoke, the rubber surround
gave up.

In the end the tube doesn't matter much, the total volume of air does.
It has to be compressed by xx percent no matter what. With liquids this
would be a ton easier to do.

 

[Joerg]

Are the "cone" and surround durable enough to stand back pressure?

Cone yes, surround no :-(

I'm curious, because I've thought of looking at such a scheme for a
positioning project.

And: How low a frequency will they work at? I have in mind 1-2 Hz.

They work down to DC. You can actually position the membrane and then
let it sit at that point of travel. That's also ok for the surround but
it sure does not like more or less static pressure.

 

[Joerg]

Try taking apart one of those really small shitty subwoofers... long-throw
speakers on the order of a few inches diameter. Turn the enclosure from
ported (they usually are) into fully sealed and you get rid of the back
wave; arrange suitable means of coupling on the face side.

That's exactly what I did for the first attempt. And promptly killed the
surround :-(

Assuming AC coupling is fine, of course. If you need DC, you could always
bleed in bias from a servo valve and PID.

I used DC.

It would take some work, but you might be able to hack a hard drive head
motor onto an especially nice ball valve (normally, they're rather sticky
from the o-rings; otherwise, if you can salvage a strong head motor from a
1980s hard drive..). Add encoder for positional feedback, and calibrate
output for the nonlinearity of the valve (or feedback loop with a nearby
pressure sensor).

I've got a valve thing right now and it works but then I need that loud
compressor to generate air pressure. Because the valve works like a
waste gate.

 

[Joerg]

Some compact woofers use linear motors.

Then there is this:

http://soundimage.dk/Different-col/LinearMotor.htm

But it costs as much as a car so it's out.

 

[Lasse Langwadt Christensen]

Den onsdag den 18. september 2013 23.04.37 UTC+2 skrev Joerg:

Which ones? Cost? ... <pant, pant, wanna have>







Cone yes, surround no :-(










They work down to DC. You can actually position the membrane and then

let it sit at that point of travel. That's also ok for the surround but

it sure does not like more or less static pressure.

you need a beefy woofer, 3psi is something like 180dB spl


-Lasse

 

[mike]

Folks,

I need to generate pressure pulses of 3-4psi above ambient, frequency
around 1Hz, more or less arbitrary waveform with frequency components up
to 40Hz in there (but those are less than 2psi) and delivered via short
hose. Currently just air pressure.

I got this (sort of) running with a compressor but after a while my ears
fall off from all that noise. Can't really place the compressor outside
here. So then I tried various speakers. They don't move enough air
around. When I pushed one of them a bit harder a few minutes ago its
flexible surround rubber burst, that thin stuff doesn't hold much pressure.


In other words I need something similar to an arbitrary function
generator but where the output is air pressure. What I want to avoid is
hacking a motor, removing a valve and using the piston.

Any other ideas? Can one buy something like that at less than 4-digit
prices?

Might be helpful to know what you're trying to accomplish.
A solution for the general case is likely to be big/expensive/complex.
How much volume you need to move is critical.
The hose can have a major effect.
As will the volume at each end.

How about using two independent pistons operating at higher
frequency and varying the phase to
create the pressure waveform?

 

[Adrian Jansen]

Folks,

I need to generate pressure pulses of 3-4psi above ambient, frequency
around 1Hz, more or less arbitrary waveform with frequency components up
to 40Hz in there (but those are less than 2psi) and delivered via short
hose. Currently just air pressure.

I got this (sort of) running with a compressor but after a while my ears
fall off from all that noise. Can't really place the compressor outside
here. So then I tried various speakers. They don't move enough air
around. When I pushed one of them a bit harder a few minutes ago its
flexible surround rubber burst, that thin stuff doesn't hold much pressure.


In other words I need something similar to an arbitrary function
generator but where the output is air pressure. What I want to avoid is
hacking a motor, removing a valve and using the piston.

Any other ideas? Can one buy something like that at less than 4-digit
prices?

How much volume ?

Would one of those el-cheapo air compressors you see in auto parts
places for inflating tyres do ? 12 volt motor, so easy to DC control.
At least you have a motor already hooked to a cylinder and piston.
Replace the motor with a stepper ?

 

[Joerg]

Linear electro-hydraulic servo valves have multihorsepower control
capability and hundreds of Hz response. One of them should work with
high-pressure air.

There is a whole generation of Hispanic hydraulic controls experts
created by the low-rider movement.

I think the pistons in there rely on the lubricating effects of the
hydraulic fluid. They could die quickly in pure air, like a two-stroke
engine driven without oil in the gasoline.

 

[Joerg]

Might be helpful to know what you're trying to accomplish.

I am testing a very tiny pressure sensing device for accuracy and
bandwidth, against a known-good large regular silicon pressure sensor.

A solution for the general case is likely to be big/expensive/complex.
How much volume you need to move is critical.

About 1/50th of a gallon to me moved. Total air is a bit more, but not much.

The hose can have a major effect.
As will the volume at each end.

What I have right now is the test vessel which needs 1/50th of a gallon
of air pushed in and out. In order to accomplish that I have about 2" or
1/4" hose connecting from there to an intermediate tank of roughly 1/3rd
gallon capacity. The source for that is a long compressor hose
connecting to a 3 gallon tank at 80psi max, but with a regulator in
between. Works well but is noisy as heck. All the less noisy electrical
constant speed air pumps I've tried aren't powerful enough to keep up.

How about using two independent pistons operating at higher
frequency and varying the phase to
create the pressure waveform?

Yes, absolutely. But I'd need this sort of "ready to roll", without it
turning into a major science project.

 

[Joerg]

---
Recalling how a steam engine's piston and escapement are used to drive
a wheel, connecting a motor to a wheel's axle and adjusting the stroke
of the piston via the escapement should allow you to get the travel
required to get the volume of air you need to generate the pulse.

That's how my compressor works

But it makes a huge racket and so I'd like a simple, less noisy solution.

 

[Joerg]

How much volume ?

1/50th of a gallon each pulse/

Would one of those el-cheapo air compressors you see in auto parts
places for inflating tyres do ? 12 volt motor, so easy to DC control.
At least you have a motor already hooked to a cylinder and piston.
Replace the motor with a stepper ?

Tried that, too weak. Even smaller medical styl;e compressors are too
wimpy for this, it needs a half-horse motor or something like that.

If there would be a speaker that can move this much plus stand the
pressure that would be great. My big one couldn't quite get the
amplitude but made decent pressure pulses. Then the surround rubber tore :-(

 

[Joerg]

I never had much luck with craft projects like that. I'd first have to
find the right material and then the surround would become big and
probably pop because of the pressure.

Or just trap the acrylic party balloon between the cone of a (repaired)
speaker and something flat or suitably curved (probably concave as seen
by the speaker). That relieves the speaker cone support from having to
stand up to any air pressure, and leaves the -- hopefully strong -- cone
working against the party balloon.

That could indeed work. Not sure how many cycles such a balloon would
survive though. I'd have about 60 per minute.

If you use one of those servoed things, expect to mess up the loop tuning.

It can be open loop, I can pre-distort the arb gen input accordingly.
Mine came with a sort of computer-based Etch-a-Sketch.

 

[axolotl]

Speakers with a big enough amp can do this but they suffer seriously
because they are meant to be operated against a fairly rigid air cushion.

You may want to look at Tom Danley's Servodrive speakers for inspiration.

Kevin Gallimore

 

[George Herold]

mike wrote:

I am testing a very tiny pressure sensing device for accuracy and
bandwidth, against a known-good large regular silicon pressure sensor.

OK, then you could hit them both with a 'step' and that should be do the time response. (Like popping a pressure vessel with a balloon on the end.)

About 1/50th of a gallon to me moved. Total air is a bit more, but not much.
What I have right now is the test vessel which needs 1/50th of a gallon
of air pushed in and out. In order to accomplish that I have about 2" or
1/4" hose connecting from there to an intermediate tank of roughly 1/3rd
gallon capacity. The source for that is a long compressor hose
connecting to a 3 gallon tank at 80psi max, but with a regulator in
between. Works well but is noisy as heck. All the less noisy electrical
constant speed air pumps I've tried aren't powerful enough to keep up.

Well if you can keep the volume down then you might be able to get away with a 'fish air pump' type thing. We sell this floating pool ball apparatus that needs a fair bit of pressure...it took a while to find the right pump.
I'll have to look up the name and specs tomorrow.

George H.