Snoredog

See if you can find a spec sheet somewhere for the AWM90000 series MAS. That's what I think we may be looking at. And yup. These are pneumotach flow sensors.

Snores

What I am saying in summary is

1) A pressure sensor as defined in the links I gave above, can measure variations in pressure & can be used to measure vibratory pressure changes.

2) A microphone cannot be seriously used to measure pressure especially in a xPAP machine.

The single nozzel devices shown in the pics of the Remstar & Resmed, I believe we all agree, measure the output pressure from the xPAP & can double up to be used to measure the vibration (pressure fluctuations) caused by snoring.

Calling them microphones is just not valid. Using them to achieve the effect of a microphone (detecting snores) is what is confusing everyone.



Jeeze this lil ole discussions is rolling along and along - I've enjoyed it - it proves we can have different perspectives and discuss them amicably. That is some achievement that we have made

DSM

#2 UNLESS, there is some other device there (that I can't see) that functions as a sound detector.

_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): resmed

SWS: here is another type of MAF sensor:

Membrane sensor

An emerging technology utilizes a very thin electronic membrane placed in the air stream. The membrane has a thin film temperature sensor printed on the upstream side, and one on the downstream side. A heater is integrated in the center of the membrane which maintains a constant temperature similar to the hot-wire approach. Without any airflow, the temperature profile across the membrane is uniform. When air flows across the membrane, the upstream side cools differently than the downstream side. The difference between the upstream and downstream temperature indicates the mass airflow. The thermal membrane sensor is also capable of measuring flow in both directions, which sometimes occur in pulsating situations. Technological progress allows this kind of sensor to be manufactured on the microscopic scale as microsensors using MEMS technology. Such a microsensor reaches a significantly higher speed and sensitivity compared with macroscopic approaches. See also MEMS sensor generations.

I wonder if sandwiched between those grates is the above.

Here is a MAF that has a thermistor like found in your fuel injected automobile (themistor found in lower right about 5'Oclock), Andy Granatelli the STP guy makes a performance one for about $300 bucks:

I'm familiar with the PITA turning those backwards can be when you relocate them onto a cold air intake (CAI), Snoredog. But these APAPs all use pneumotachs, and not thermistors. Here are the Mas Airflow Sensors (pneumotachs) that I think we see pictured:
http://www.google.com/search?q=awm90000 ... f8&oe=utf8

Snoredog, Doug, isn't that pretty darn cool when we interleave information like that. I have learned bunches from you two and bunches from SAG (and so many others on this forum).

What a cool, cooperative think tank!

That is a carburetor right ?

Seems like an interesting technology & certainly if it can scale down & not cost a mint in its minature form, it could apply to medical devices and xPAPs.

Do they provide any indication of its sensitivty to very light volumes of air (say 10 lpm).

DSM

-SWS wrote:I'm familiar with the PITA turning those backwards can be when you relocate them onto a cold air intake (CAI), Snoredog. But these APAPs all use pneumotachs, and not thermistors. Here are the Mas Airflow Sensors (pneumotachs) that I think we see pictured:
http://www.google.com/search?q=awm90000 ... f8&oe=utf8

Snoredog, Doug, isn't that pretty darn cool when we interleave information like that. I have learned bunches from you two and bunches from SAG (and so many others on this forum).

What a cool, cooperative think tank!

Well looks like you found the one on Respironics PCB, Honeywell makes it, hey save that link for parts hehehe.

http://components.argussoft.ru/honey/de ... M90000.pdf

So what are they then calling the pickup tube if the MAF is on the PCB?

So would it contain just the tube and grates for separation of tubes at a precise distance?

dsm wrote:That is a carburetor right ?

Seems like an interesting technology & certainly if it can scale down & not cost a mint in its minature form, it could apply to medical devices and xPAPs.

Do they provide any indication of its sensitivty to very light volumes of air (say 10 lpm).

DSM

no, vehicles this is found on don't have carberators, fuel injection replaces that. The throttle valve is just a air valve (controls amount of air entering the system) the above is a MAF or MAS Mass AirFlow Sensor found on today's fuel injected automobiles. The ECM computer needs that data to know how much air flow is entering the engine (based upon your accelerator position), then the ECM opens the injectors to inject the proper amount of fuel to obtain a 14.54:1 mixture.

Actually, see SWS's link above for the AWS9000, the Remstar MAF is found there.

Snoredog wrote:

blarg wrote:I guess I'm just confused as to why they'd need to measure outside sound frequencies when detecting snores, because we hear them....

that is because the airflow is looked at (by the machine algorithm) as a sine wave like you would see on an oscilloscope. The breath or flow is monitored by assigning numeric values to those waveforms (like shown in the center blue line in the various forms of SDB events below).

Ok, obviously you're not running a machine with a mic. That was never confusing for me, but specifically for detecting snores you could use a mic, but they don't, but the device they do use is functioning as one specifically when it's used for detecting snores. Am I interpreting correctly?

snoredog wrote:So would it contain just the tube and grates for separation of tubes at a precise distance?

I think the grates are used here to straighten airflow. I believe the large grates you see are to get the airflow entering the narrow tube perpendicular to cross-section. Then into the pneumotach housing you'll probably see smaller air grates once again used to straighten out even finer air turbulence vectors. Again, the air flow vectors need to enter the bundled cross-sectional tube openings "straight on" so to speak. Small vector path disturbances can skew measurements. Hence even "micro-turbulence" needs to be straighted out. Again, they're bundled because we're measuring differential flow-based pressure across multiple heads of each pneumotach conduit or sub-tube.

blarg wrote:

Snoredog wrote:

blarg wrote:I guess I'm just confused as to why they'd need to measure outside sound frequencies when detecting snores, because we hear them....

that is because the airflow is looked at (by the machine algorithm) as a sine wave like you would see on an oscilloscope. The breath or flow is monitored by assigning numeric values to those waveforms (like shown in the center blue line in the various forms of SDB events below).

Ok, obviously you're not running a machine with a mic. That was never confusing for me, but specifically for detecting snores you could use a mic, but they don't, but the device they do use is functioning as one specifically when it's used for detecting snores. Am I interpreting correctly?

I believe you have it right

But the tech specs of the device used to measure the output pressure & the snore pressure vibrations define it as a pressure transducer (not even a hint of the word mic). Calling it a mic is a very loose use of the definition of a mic & in pure tech circles would not be accepted.

I believe this confusion (as I see it) may have come about because all the early papers written on detecting snoring all did the lab tests using conventional microphones. Then some early Autos were produced where they used 'acoustic sensors' to measure snores, then the manufacturers fairly quickly switched to 'pressure sensors' which allowed accurate measurement of exit air pressures and doubled up as accurate for measuring snore vibrations. It became easy to refer to them as a mic even thought they aren't.

I believe we have all established that there are 2 types of pressure sensors used in xPAPs, MAF sensors (rigged in pairs) and Pressure sensors (used individually to measure output pressure and snore vibrations).

In the Resmed S7 they show a pair of MAF chips where one of the pair are used for pressure differential measurement & of the other pair, only one half is used and it is primarily used for measuring the output pressure and secondarily for measuring air pressure vibrations that are caused by snores. The pressure & frequency can be measured & used to determine the intensity of the snores.

It does seem to me that the snores could be detected by either set of pressure sensors. In fact it may make more sense if the flow detection pair get used for snore detection. (I can't recall if we established conclusively the the output pressure sensor is the only one used for both exit pressure & snore vibrations).

Now have we reached consensus, maybe, maybe not. But this techhead is in no doubt what these pressure gizmo gadgets are

DSM

blarg wrote:

Snoredog wrote:

blarg wrote:I guess I'm just confused as to why they'd need to measure outside sound frequencies when detecting snores, because we hear them....

that is because the airflow is looked at (by the machine algorithm) as a sine wave like you would see on an oscilloscope. The breath or flow is monitored by assigning numeric values to those waveforms (like shown in the center blue line in the various forms of SDB events below).

Ok, obviously you're not running a machine with a mic. That was never confusing for me, but specifically for detecting snores you could use a mic, but they don't, but the device they do use is functioning as one specifically when it's used for detecting snores. Am I interpreting correctly?

yes it is doing basically the same thing as a Mic would convert sound pressure waves to electrical signals.

But even with OUT the audio sound from a snore, it would "detect" the snore flow signal as shown on the chart above.

This is why like on a Remstar EncorePro report it may record snores yet if you sleep with a partner they may never hear you "audibly" snore, but the vibrations may still be present for snore where the device can still detect it.

The airflow "signal" coming back down the cpap hose to the machine is what gets picked up, similar to a hydraulic circuit. Apparently it is very sensitive as some have reported that they only dragged the hose across the bed and it registered as a snore.

This is the two-way feedback the machine needs to see to function properly. It probably uses both flow and pressure sensors to determine when you exhale for like the Cflex feature. If you have ever used Cflex you know it is pretty darn good at detection of your inhale/exhale.

dsm wrote:Now have we reached consensus, maybe, maybe not. But this techhead is in no doubt what these pressure gizmo gadgets are

Same here. Now what do we use them for? Mine doesn't seem to be cutting the grass properly.

As another comment. Irrespective of what labels we choose to stick to (pressure device or mic) I believe that we all have contributed to a pretty good understanding and mind picture of the devices and their roles in the machines in this area.



DSM

-SWS wrote:

snoredog wrote:So would it contain just the tube and grates for separation of tubes at a precise distance?

I think the grates are used here to straighten airflow. I believe the large grates you see are to get the airflow entering the narrow tube perpendicular to cross-section. Then into the pneumotach housing you'll probably see smaller air grates once again used to straighten out even finer air turbulence vectors. Again, the air flow vectors need to enter the bundled cross-sectional tube openings "straight on" so to speak. Small vector path disturbances can skew measurements. Hence even "micro-turbulence" needs to be straighted out. Again, they're bundled because we're measuring differential flow-based pressure across multiple heads of each pneumotach conduit or sub-tube.

I just think there are about 3 or more of those wafer grates (have no idea why they are seperated, but the Honeywell spec says:

NOTICE
LAMINAR FLOW
Due to the fast response time of the sensor, these specifications were generated using laminar flow. Airflow instability or “turbulence” present in the airstream will result in an increase in measurement uncertainty.
The turbulent flow problem can be corrected by either straightening the airflow using flow laminarizing or by slowing the response of the sensor using a simple RC time constant on the output of the sensor. This, of course, slows down the sensor response time. The values needed depend on the amount of turbulence present in the application.
Several techniques for laminarizing the flow include adding hex shaped honeycombs, foam, screen materials or adding constrictors (frits) to the flow stream. There are various commercial laminar flow elements that can be purchased. Unfortunately the greater the efficiency of the laminarizer, the greater the increase in pressure drop in order to establish a given flow rate. Plastic honeycomb material probably gives the most improvement for the least pressure drop. In any test fixture, the avoidance of sharp radii is an absolute requirement.

So I guess those are laminarizing the flow to smooth it out.

dsm wrote:As another comment. Irrespective of what labels we choose to stick to (pressure device or mic) I believe that we all have contributed to a pretty good understanding and mind picture of the devices and their roles in the machines in this area.

Especially the part about cutting the grass. Where does the four-cycle gas engine strap on these APAP thingies? I've already got the wheels and handles mounted on mine...

Wha? Class is dismissed already?