Attention Manufacturers! Changes we'd like in our equipment

[jedimark]

Not sure we will see WiFi on next generation of machines. Manufacturers do not allow access to their software and lock us out of clinician settings. Seems to me they do not want us accessing our data.

PR offers BlueTooth module for communicating with SleepMapper... I wonder if it would talk to SleepyHead too?

I have one of these very modules.. Unfortunately I have no idea how to work it yet, to make it work with SleepyHead.

I can't even get SleepMapper in Australia yet to get a feel for what this module is capable of.

[jedimark]

I've got a few ResMed firmware bugfixes I'd like to see....

ResMed really needs to do something about their (mis)use of the EDF+ file format. Kudos to them for using an open format, but their execution is rather crummy.

I'd like to see ResMed machines record settings and summary information *per session*, not per day.. Whoever thought of trying to stuff summary data into the inflexible STR.EDF file is quite a few volts short of lighting a bulb.

And I'd like to see old files safely *archived* instead of deleted every 7, 30 or 365 days. SD cards are comparatively massive compared to what they are using them for. On a related note, I seriously hope they aren't overwriting the same spot on the SD card every time the STR.edf summary file is updated.

Plus, I'd like to see them keep their channel codes consistent between foreign language machines, because it's currently an awful mess.

And, then there is their disturbing habit of using the same channel name for two different channels in STR.EDF

And the ugly bug that puts a delay between session filenames and start times in the same session group needs to be fixed.

Well, That's that.. if only they'd listen

[archangle]

ResMed needs to stop subtracting and discarding "intentional" leak data in the CPAP machine itself. Either record total leak and let the PC software figure it out, or at least stop clipping the recorded data at zero.

The machine can't actually calculate unintentional leak directly, so it's sometimes throwing away useful data.

[palerider]

The machine can't actually calculate unintentional leak directly, so it's sometimes throwing away useful data.

why can't it?

[Droopy]

Can't the cat sleep in another room?

[archangle]

The machine can't actually calculate unintentional leak directly, so it's sometimes throwing away useful data.

why can't it?

It doesn't have enough data. All the machine can measure is the airflow going down the hose. Average that over a short period of time and you have the total leak rate. There is no way for the machine to measure how much goes out through the exhaust vent (intentional) vs. leaking around the mask our out of your mouth.

In theory, if you tell it the exact mask, it can calculate the vent rate of the mask based on the pressure. However, the machine probably doesn't have a listing of all mask types. Even if it knows the exact mask, the leak rate probably varies from one mask to another of the same model, and changes as moisture, dust, dirt, or minerals build up on the exhale vent.

Also as another example of why it's so stupid to discard the total leak rate, think about what happens if the mask vent gets blocked by the user's arm, a pillow, or clogs up. On a PRS1 machine, you will see that the total leak rate goes to zero. On an S9 machine, you'd get no indication that the user may now be rebreathing the same air over and over. The user would (hopefully) wake up with a feeling of suffocation and wouldn't know why without total leak rate data. The S9 machines measure the total leak rate, but throw it away if the leak rate is below the expected intentional leak rate.

[palerider]

The machine can't actually calculate unintentional leak directly, so it's sometimes throwing away useful data.

why can't it?

It doesn't have enough data. All the machine can measure is the airflow going down the hose. Average that over a short period of time and you have the total leak rate. There is no way for the machine to measure how much goes out through the exhaust vent (intentional) vs. leaking around the mask our out of your mouth.

In theory, if you tell it the exact mask, it can calculate the vent rate of the mask based on the pressure. However, the machine probably doesn't have a listing of all mask types. Even if it knows the exact mask, the leak rate probably varies from one mask to another of the same model, and changes as moisture, dust, dirt, or minerals build up on the exhale vent.

have you ever watched your s9 calcuate the leak rate? it's pretty nifty.

it actually should have all the data it needs, which is the lowest leak rate that it sees when you crank things up. presumably, most of the time, you'll have a zero additional leak early on, and that's where it sets its initial baseline. watch the leak stats on the machine when you first turn it on, they jump up and down for a few breaths until the machine can see "well, it's not going lower than x..." and it settles to zero and stays there while you breath.

Also as another example of why it's so stupid to discard the total leak rate, think about what happens if the mask vent gets blocked by the user's arm, a pillow, or clogs up. On a PRS1 machine, you will see that the total leak rate goes to zero. On an S9 machine, you'd get no indication that the user may now be rebreathing the same air over and over. The user would (hopefully) wake up with a feeling of suffocation and wouldn't know why without total leak rate data. The S9 machines measure the total leak rate, but throw it away if the leak rate is below the expected intentional leak rate.

now, that's a wholly different point, and a good one of course, if you've been in that situation long enough, you won't be lookin at your graphs.

[archangle]

have you ever watched your s9 calcuate the leak rate? it's pretty nifty.

it actually should have all the data it needs, which is the lowest leak rate that it sees when you crank things up. presumably, most of the time, you'll have a zero additional leak early on, and that's where it sets its initial baseline. watch the leak stats on the machine when you first turn it on, they jump up and down for a few breaths until the machine can see "well, it's not going lower than x..." and it settles to zero and stays there while you breath.

I've looked and I disagree with your assessment of how it works.

Also as another example of why it's so stupid to discard the total leak rate, think about what happens if the mask vent gets blocked by the user's arm, a pillow, or clogs up. On a PRS1 machine, you will see that the total leak rate goes to zero. On an S9 machine, you'd get no indication that the user may now be rebreathing the same air over and over. The user would (hopefully) wake up with a feeling of suffocation and wouldn't know why without total leak rate data. The S9 machines measure the total leak rate, but throw it away if the leak rate is below the expected intentional leak rate.

now, that's a wholly different point, and a good one of course, if you've been in that situation long enough, you won't be lookin at your graphs.

You might wake up partially and move enough to unblock the vent if it's your arm or the pillow, or cause a leak around the mask. The vent might not be fully blocked. Even with a fully blocked vent, when you exhale, air will go out the back of the blower unit when you exhale and some fresh air will come in. There will be some mixing of fresh and stale air in the hose and in the machine, so you might get enough O2 to stay alive.

[palerider]

have you ever watched your s9 calcuate the leak rate? it's pretty nifty.

it actually should have all the data it needs, which is the lowest leak rate that it sees when you crank things up. presumably, most of the time, you'll have a zero additional leak early on, and that's where it sets its initial baseline. watch the leak stats on the machine when you first turn it on, they jump up and down for a few breaths until the machine can see "well, it's not going lower than x..." and it settles to zero and stays there while you breath.

I've looked and I disagree with your assessment of how it works.

while I'm not surprised, I'm curious as to what your theory is then.

[archangle]

while I'm not surprised, I'm curious as to what your theory is then.

I think they have a leak vs. pressure curve for each type of mask. It probably has a little "headroom" to allow for mask variations. i.e. it is designed to overestimate the leak.

I've looked and don't see any sign of it starting out with one leak rate and then adjusting, but maybe there's something that I'm not seeing for some reason. Or maybe they do the leak calculation differently with different firmware versions.

I'll have to think about how to do some experiments to see what really happens. I should be able to see something by changing mask types and/or figuring out how to add a calibrated extra leak to the system.

It would be neat if actually does adjust leak rates that way. You often learn new things if you're proven wrong.

[palerider]

I've looked and don't see any sign of it starting out with one leak rate and then adjusting, but maybe there's something that I'm not seeing for some reason. Or maybe they do the leak calculation differently with different firmware versions.

I'll have to think about how to do some experiments to see what really happens. I should be able to see something by changing mask types and/or figuring out how to add a calibrated extra leak to the system.

It would be neat if actually does adjust leak rates that way. You often learn new things if you're proven wrong.

to be clear, what I meant is that if you mask up while watching the info screen (press the knob to toggle between them) you'll see a second by second leak value that jumps around for the first few breaths and then settles down to a baseline. create a leak after that, and it shows it as leaking above whatever baseline it settled on.

I've seen it both on my vpap auto, and on a vpap adapt. I haven't cranked up the old elite as a test case, though I could.

[archangle]

to be clear, what I meant is that if you mask up while watching the info screen (press the knob to toggle between them) you'll see a second by second leak value that jumps around for the first few breaths and then settles down to a baseline. create a leak after that, and it shows it as leaking above whatever baseline it settled on.

I've seen it both on my vpap auto, and on a vpap adapt. I haven't cranked up the old elite as a test case, though I could.

I'm looking at SD card data. I'll look at the info screen to see what you're talking about.

Edit - are you talking a display with numbers like "3 l/m" or some such? I don't see any such thing on my S9 AutoSet.

Once I start up, if I press the "i Info" button, nothing happens. If I press the "check check" setup button button, I get a settings screen, but no leak info. If I press the push dial, it goes back to the humidity level display. If I hold the push dial, it shows me smiley or frowny mask fit.

If I start in either info screen, as soon as the blower starts, it goes to the humidity level screen.

Maybe you have a screen I don't.

[archangle]

to be clear, what I meant is that if you mask up while watching the info screen (press the knob to toggle between them) you'll see a second by second leak value that jumps around for the first few breaths and then settles down to a baseline. create a leak after that, and it shows it as leaking above whatever baseline it settled on.

I've seen it both on my vpap auto, and on a vpap adapt. I haven't cranked up the old elite as a test case, though I could.

Rather than continuing to hijack this thread, let's start a new thread:

viewtopic.php?f=1&t=98205&p=916968#p916968