HackPAP

[AfibApnea]

I may have to change from CPAP to AutoCPAP, to BiPAP, or to some other more complicated air delivery system.

I have been told that when one “upgrades” from a CPAP machine to a more ‘sophisticated’ air delivery method (e.g., AutoCPAP, BiPAP, etc.) that essentially the machine stays the same and that what’s different is that another software algorithm (firmware) is programmed into the machine. I believe that if I have to change that I will be charged an outrageous amount for what is essentially the same machine with its different firmware.

Has anyone else heard this?

Is this true?

[Guest]

I may have to change from CPAP to AutoCPAP, to BiPAP, or to some other more complicated air delivery system.

I have been told that when one “upgrades” from a CPAP machine to a more ‘sophisticated’ air delivery method (e g , AutoCPAP, BiPAP, etc.) that essentially the machine stays the same and that what’s different is that another software algorithm (firmware) is programmed into the machine. I believe that if I have to change that I will be charged an outrageous amount for what is essentially the same machine with its different firmware.

Has anyone else heard this?

Is this true?

No. For all intents and purposes, they can't make an APAP out of a straight CPAP, nor can you make either one into a BiPAP with programming. BiPAPs can (in most cases) be made to run as an APAP or straight CPAP, but not the other way around.

Usually, a person can be switched to APAP or BiPAP if they "fail" at straight CPAP therapy.......by being provided with a different machine.

[JDS74]

I don't know if it is literally true but it does make sense.
What an xPAP device does is to deliver air at a given pressure and measure flow and/or pressure changes and then modify the outgoing air pressure to accommodate.

If I haven't over simplified too much:

A CPAP delivers air at a fixed pressure and maintains that pressure.
An Auto Pap delivers air at a pressure but can increase or decrease that pressure.
An Auto BiPap delivers air at a pressure for inhaling and a different pressure for exhaling and can increase or decrease those pressures.
An Auto BiPap ASV delivers air just like an Auto BiPap but can change pressures much more rapidly and can switch to a large difference between inhale pressure and exhale pressure.

They all require a way to deliver air, to measure pressure and possible flow rate, and to change air pressure.
For full data machines there is some added cost for the recording device and for the display. But even for bricks, they still record.

One can imagine that a single hardware platform could accommodate all of these modalities with the difference being the firmware alone. So the unit marginal hardware manufacturing cost should be the same for all.

What is likely to be different is the regulatory cost for each set of firmware.
Assume that the regulatory cost for approval is $10 million (its probably much more).
And assume that the market for these devices is 100 thousand per year for one manufacturer.
First, assume that 90% of patients can be treated successfully with a simple CPAP.
Second, assume that of the remaining 10%, 8 % can be treated successfully by an Auto Pap or Auto BiPap.
And third, assume the last 2% require an Auto BiPap ASV for successful treatment.
We neglect those who fail in usage because they fall into one of the categories and are counted as if they succeeded.

The regulatory cost for the simple CPAP is $111 per unit. (90,000 units)
The regulatory cost for the Auto machines is $1,250 per unit. (8,000 units)
And the regulatory cost for the Auto ASV is $5,000 per unit. (2,000 units)

That's why the big jump between a straight CPAP and an auto unit.
It also is a reminder that the Auto ASV is probably a loss leader and costs the company more to build than they can charge.

Your guess is likely better than mine on the total regulatory costs and on the market size in total and by device. This in not meant to be a definitive analysis but an illustration of the cost problems facing the manufacturer. If the market is 500 thousand, then the regulatory costs per unit go down dramatically, etc.

[palerider]

I may have to change from CPAP to AutoCPAP, to BiPAP, or to some other more complicated air delivery system.

I have been told that when one “upgrades” from a CPAP machine to a more ‘sophisticated’ air delivery method (e.g., AutoCPAP, BiPAP, etc.) that essentially the machine stays the same and that what’s different is that another software algorithm (firmware) is programmed into the machine. I believe that if I have to change that I will be charged an outrageous amount for what is essentially the same machine with its different firmware.

Has anyone else heard this?

Is this true?

depends, the hardware inside the resmed s9 series from the cheapest to the most expensive is the exact same physical hardware, the only difference is the programming.

the respironics system one's do have different circuit boards in them, because some of the fancier ones have an extra button.

you don't upgrade them, they don't do that sort of thing, you'll get another one that has a different name on the outside.

(I disassemble and reassemble xpaps as a hobby... (and to de-cigarette smoke use ones))

[aytikvjo]

For the most part they try to homogenize the bill of materials between products because it reduces the overall production cost of the device in a couple ways:
1.) Development costs (Engineering time and Q&A time is expensive)
2.) Tooling costs (Molds are expensive, spinning pcb's is expensive)
3.) Materials costs

For the most part, components between machines of the same series will be the same with the main differences being in software and facing.
Bi-Level devices sometimes have an extra component in the air path is basically a voice-coil connected valve for the purpose of dumping pressure overboard.

For Respironics series 60 NIV devices (S/T, AVAPS, ASV) there's an extra button so at least the face plate is different.

In all likelihood a Respironics / Resmed engineer with the appropriate tools could flash a regular CPAP to the APAP firmware with no difference in hardware. The Bi-level could probably be flashed to an auto-bi level too. It's also probably likely that a normal bi-level could use the ASV/AVAPS firmware with minor modifications to the code (if they don't use a blower with a more aggressive performance profile).

But for the consumer to do this is simply not realistic, nor is it really worth trying for the sole purpose of saving money. These are medical devices and a lot of effort has gone into their design and assurance that they operate correctly and fail safely. If you have a junk one and are handy with a logic analyzer and oscilloscope then you are welcome to probe a bus or two, but I would never open my main device for which I use for therapy.

*Note, if they are using a simple SoC with built in program flash, they've probably made use of the flash read/write protection features of most higher end chips anyway. Though if they are loading program from an external eeprom.... Skies the limit. Not to mention the attack vector of hijacking the normal firmware update process.

You may think this unfair, but this is standard practice in a wide variety of industries. It helps them keep costs down. The high price of developing specialized devices helps keeps prices of the low-end ones down and it's all the same to the consumer.

Example: Modern digital oscilloscopes, which can range from very cheap at $350 to all the way over $10k for some very high bandwidth models. Usually the bandwidth is limited in software, so a $800 70MHz device has the same hardware as the $2500 300Mhz device. Only difference is a software key (which people have subsequently found out how to generate and unlock the hardware for some brands / models). Agilent, Tektronix, all the big names do this.

Intel and AMD do it with their processors. Modern car stereo / nav are like this sometimes. Commercial software does it all the time in various ways. Even video games do it in their own way.

TLDR: Thinking this practice is unfair comes off as kind of entitled. Time to develop these higher-end features is very expensive and homogeneous platform design keeps costs down.

[purple]

I am not an expert, but I have been told that the primary diff between models of machines is firmware, not hardware. One can NOT upgrade the firmware to ones' sleep apnea machine to have the higher end features. (At least it would take an expertise that most of us do not have, and the manufacturer would be opposed to us doing that) In theory, companies are being reimbursed for the extra costs of developing the firmware (and the risks-costs entailed if a mistake is made). While I would like to advocate that nothing less than an auto bi level be sold, if Resmed, and PR used an auto Bi Level (all right a few changes to emulate all the machines have up to that point) Then some company is going to come along and manufacture a cpap brick, and offer it for sale for less money, and insurance companies will require we get that instead of the auto bi level machine I advocate. I say that the cost of the bi level could be distributed among all the machines, and we could have the full abilities of auto bi level for not much more than what a brick costs.

I once spoke with a tech and mentioned that the documentation on a mask could be easily improved, and the tech said that there was, an enormous dollar cost in changing the documentation because of the legal consequences. All kinds of level of review by both the company and lots of different countries -government agencies, all for a number of different languages as well.

[chunkyfrog]

Considering that so many people drop out of therapy, mostly due to comfort issues;
forcing everyone to jump through a gauntlet of hoops to get optimal treatment, is foolish and cruel.
--IMHO

[SleepyCPAP]

... I believe that if I have to change that I will be charged an outrageous amount for what is essentially the same machine with its different firmware.

Hi AfibApnea,

I'm going to read a different question than the above have answered, focusing on your statement about being charged more as you transition to new modes of therapy. If the equipment you listed for yourself is correct, you already have APAP. So if you are moving from CPAP to APAP there is no change, other than selecting the new mode in the clinical menu. No extra cost there for the machine to do it (as it is built-in, and a menu option) -- I have no clue what a DME charges for doing the change of setting, but of course there are manuals and online advice here for those who get the Dr.'s new settings and then DIY.

But yes, the way things are set up now you'd have a charge for getting a BiPap. Maybe your insurance picks up that cost?

--SleepyCPAP

[archangle]

In general, the difference between the cheapest data capable CPAP up to the "normal" auto bilevel machine is just the firmware. The dataless bricks lack some airflow sensor hardware. The higher level machines such as ASV might be different, I'm not sure.

The manufacturers have a large financial incentive to make sure customers can't reprogram the machines to do the higher function. I haven't heard of anyone hacking the software yet. One would think that, given the large financial incentive, the machines would be "secure" against hacking, but so many modern high tech firms have really poor security practices, you can't be sure.

[palerider]

In general, the difference between the cheapest data capable CPAP up to the "normal" auto bilevel machine is just the firmware. The dataless bricks lack some airflow sensor hardware. The higher level machines such as ASV might be different, I'm not sure.

I'd love to get my hands on a s9 escape for about 10 minutes. (longer if I were to take pictures) and see if it's the same hardware as the elite/adapt units.

[Nick Danger]

Hmmm... I might look over your disassembly post and take apart my old Escape. Of course, I have no idea what to look for other than comparing to your photos.

[SleepWrangler]

I have been told that when one “upgrades” from a CPAP machine to a more ‘sophisticated’ air delivery method (e.g., AutoCPAP, BiPAP, etc.) that essentially the machine stays the same and that what’s different is that another software algorithm (firmware) is programmed into the machine.

Has anyone else heard this?

Is this true?

It could be true. Hardware is only part of the cost. Where software is involved there may be licensing issues where the manufacturer owes royalties to other stakeholders based on the model because of features and capabilities included in each model.

*Note, if they are using a simple SoC with built in program flash, they've probably made use of the flash read/write protection features of most higher en
d chips anyway. Though if they are loading program from an external eeprom.... Skies the limit. Not to mention the attack vector of hijacking the normal firmware update process.

In 2010 EET Asia announced that ResMed is using an STM32. I've seen linked-in profiles of ResMed engineers saying they have experience with ARM, Cortex M4, STM32.

The STM32 has flash, eeprom, ram built-in and the device can be locked from reading its contents via the programming / debug interface. PAP equipment some people in this forum are using does allow the firmware to be upgraded. Unfortunately ResMed is not one of them. No proof, and although security can be a concern, I would suspect licensing is the reason each model is locked to a particular set of features, capabilities, and version.

[archangle]

I'd love to get my hands on a s9 escape for about 10 minutes. (longer if I were to take pictures) and see if it's the same hardware as the elite/adapt units.

Somewhere in my addled brain, I have a fuzzy recollection of convincing myself that the S9 Escape did really lack the separate airflow and pressure sensor. However, as I said, fuzzy recollection.

[Guest]

I may have to change from CPAP to AutoCPAP, to BiPAP, or to some other more complicated air delivery system.

I have been told that when one “upgrades” from a CPAP machine to a more ‘sophisticated’ air delivery method (e g, AutoCPAP, BiPAP, etc.) that essentially the machine stays the same and that what’s different is that another software algorithm (firmware) is programmed into the machine. I believe that if I have to change that I will be charged an outrageous amount for what is essentially the same machine with its different firmware.

Has anyone else heard this?

Is this true?

I believe the discussion has drifted away from the original poster's questions/concerns. Just because the "innards" of the machines can be similar or even identical (except for the programming), doesn't mean that someone outside the factory assembly line is capable of changing the firmware. From my own experience with the Respironics Legacy models (Auto and Pro 2), the PCA board in the Pro 2 actually says "Auto C-Flex Domestic" on a label on one of the chips. However, there's no way to reprogram it to be an Auto outside of the factory. Even if it were possible, the machines would have to be re-calibrated with the appropriate special equipment. These boards are created with automated manufacturing equipment for a minimal amount of money. So, it would be more feasible to swap the actual PCA (motherboard) for a different one. And, even then it would have to be checked with special equipment for calibration.

To answer the OP's question/concern........it ain't gonna happen the way he described.

[palerider]

Hmmm... I might look over your disassembly post and take apart my old Escape. Of course, I have no idea what to look for other than comparing to your photos.

the mechanicals are easy to spot, and I can point you to high res pics of the front and back of the circuit board, if you don't see missing parts, then it's the same hardware

all you need is a thin flat screwdriver to fit into the slots on the bottom, and a long shaft torx 10 driver.