Water level alert

[Tatooed Lady]

Interesting...it's something I've briefly thought on, as well, but I'm not a super techie type, so once I think, "gee, wouldn't it be cool if..." everything in my head goes into 4th grade mentality...hoses, loops, wires, Dr. Jekyl's lab, etc...
Realistically, though, you would THINK that, in theory, it would be similar to one of those pet watering setups...you know, the upside down 2 litre bottle deals? Maybe just put a hole in the lower front of a humidifier tank, run a hose to the lower end of a water bottle, another humidifier tank, whatever...sort of similar, also, to the semi trucks with dual fuel tanks having a crossover tube to level the amount of fuel in both tanks...and POOF...so long as all the seals are good, you SHOULD have success. Or need a mop in aisle 2.

[JDS74]

Does anyone know of a commercially available water level alert, or an extra capacity humidity tank for the PR 60 series? I sleep so soundly with my machine, that many nights the tank will run bone dry several hours before I wake up with stuck together mouth parts..

External water capacity must never have a water level above that of the internal level. Thus is true for two reasons. 1. If the level is higher, it will add additional pressure to the system. I, for example, it is 10 cm higher, that will add pressure to the system and that will make the system pressure sensors 'think' that less pressure is needed and therefore airflow will be reduced. 2. If the level is high enough, water will overflow into the blower unit.

Adding an external passover unit in series with the internal humidifier will work but may interfere with humidity and temperature control. For example, suppose you are using a heated hose. Then you must connect between the end of the heated hose and the mask. Now you need to control the temperature of the external water supply otherwise it will cool down the airflow and exacerbate any rainout issues.

Adding an external room humidifier will increase the humidity of the air entering the blower unit and therefore reduce the rate at which the water level goes down during the night.

So, from an engineering point if view, if one were to construct a humidifier tank that is identical to the existing tank where it goes into the humidifier and extends out and is sealed so that air pressure within the system is retained and is always level (see the first requirement) then additional capacity would be available. The best characteristic of such an extension would logically be to eliminate the exterior wall of the existing tank, i.e., imagine a tank just like an OEM one but just a little longer.

[palerider]

External water capacity must never have a water level above that of the internal level. Thus is true for two reasons. 1. If the level is higher, it will add additional pressure to the system. I, for example, it is 10 cm higher, that will add pressure to the system and that will make the system pressure sensors 'think' that less pressure is needed and therefore airflow will be reduced.

I believe you are almost wolly incorrect, think, as was mentioned, the upsidedown 'watering globes', or, for instance, the upsidedown water containers for small pets.

[Guest]

Just as long as we are brainstorming...

Adding an external passover unit in series with the internal humidifier will work but may interfere with humidity and temperature control.

If using the OEM heated hose yes the hose heater will no longer work. As mentioned an aftermarket heated hose can be added if that heat is needed or wanted. Also worth noting here if using the HC 150 it does come with a heat plate to heat the water but it can also just be used in passover (w/o heat). That may help some.

Adding an external room humidifier will increase the humidity of the air entering the blower unit and therefore reduce the rate at which the water level goes down during the night.

Prob the easiest solution. I normally turn on a room humidifier when the furnace is turned on.

I have ordered a fish tank water level sensor.

Looking at this it appears those sensor are rather large and both hi & lo need to fit into the OEM water chamber. But if this is just an alarm feature meaning I would have to wake up to add the water I wouldn't be in favor of this for me <smile> and would opt for the HC 150 addition, tho it does cost a little more you may be able to get your insurance to pick up the cost if the OEM is not adequate for your needs.

JDS74 wrote:External water capacity must never have a water level above that of the internal level. Thus is true for two reasons. 1. If the level is higher, it will add additional pressure to the system. I, for example, it is 10 cm higher, that will add pressure to the system and that will make the system pressure sensors 'think' that less pressure is needed and therefore airflow will be reduced.

This may be the most complicated approach of all. The added bottle will not only need to be air & water tight - don't forget there will be air pressure inside the OEM chamber - my question is will this pressure keep any water in the added bottle from filling the OEM chamber? Or will the use of two holes or hoses (one up higher for air) be needed for this to work? Will the added holes cause whistling or other noises?

Honestly, when I first started I needed the extra heat from a heated hose but as time went on I found that heated me up from the inside out and now feel that passover is enuff for me. If/when I need more moisture I turn up the heat on the OEM humidifier. YMMV

[chunkyfrog]

Since the tank on my Airsense 10 is so easy to remove and replace,
I think I will swap tanks when I run low. Luckily, my nose tells me when the water gets low.
I wish everyone had this gift, though some may be taken aback by some of the unpleasant odors.

[Guest]

Luckily, my nose tells me when the water gets low.

Mine does too but I just remember when I get up (later that same day) I need to do something about it...cuz I do have the external passover backup <smile>, so nothing is urgent.

[AlabamaAl]

Just as long as we are brainstorming...

I have ordered a fish tank water level sensor.

Looking at this it appears those sensor are rather large and both hi & lo need to fit into the OEM water chamber. But if this is just an alarm feature meaning I would have to wake up to add the water I wouldn't be in favor of this for me <smile> and would opt for the HC 150 addition, tho it does cost a little more you may be able to get your insurance to pick up the cost if the OEM is not adequate for your needs.

While in a perfect world, The best solution would be to enlarge the tank capacity, but since I no longer have access to the tools and equipment necessary to prototype this, I felt the sensor was my best option. The overfill sensor would not be necessary to install in the tanks, since it just senses an overflow presence, this will be taped in a dry location, it does not matter where. The low tank sensor will be the only one in the tank. I don't mind waking up to refill the tank, I just need an alert that it needs refilling. After all pre cpap I was awakened five or six times a night for bathroom breaks. Once or twice to refill a tank is nothing. I have also swapped tanks during the night, but refilling is just as easy.

[archangle]

Even if you had the temperature cranked up on your humidifier, your humidifier will use more water but should last a few nights. Your tank should not rune bone dry and the fact that you sleep soundly has nothing to do with the level in the humidifier. When I first started cpap therapy, I used a humidifier and it ran dry. I called my DME and they gave me replacement humidifier while mine was being repaired. Maybe you should do the same.

Sheriff

My PRS1 will run dry some nights even if I start with a full tank and the humidifier on 4. It depends on leak rate, pressure, and room air temp and pressure. Down there in the swamp where you are, there's more humidity in the air to start with.

BTW, I do have "system one humidification" turned on.

[Guest]

Certainly do what you feel is best. From what I can see it will certainly be a challenge. I am not saying it can't be done but the way I see it the sensor will have to go in to the front part of the chamber; the water line is about 1.25 in high and that sensor looks fairly large in the picture. Even if it fits in that space every time you remove the chamber you will have the wires to deal with. For me that is another deal breaker.

I still think the HC 150 and an aftermarket heated hose will improve the humidity enuff for most people.

For those who like to tinker search this forum for the Repti-Heat cable. Just be sure to get the lowest watt (I think it's 15W) and to just go in a straight line - Do NOT wrap the wire around the hose or it will melt it. Also using a timer on the cable is prudent as you will surely forget to turn it off.

[aytikvjo]

I also have the problem of the humidifier running dry if I sleep more than 8 hours. I'd love to have it cranked at tube temp 5 and humidity 3 all night, but it just doesn't last in this winter air.

Disclaimer: Rambling

I've thought about this problem a little bit off and on. My solutions are tailored to the Respironics system one humidifiers though because of the exposed face. Don't know if an S9 would be possible, but S10 could be.

So far what I've got based on the following design criterion:
1.) Fault tolerant and graceful failure
2.) Minimal or non-contact with as few entries into water chamber as possible
3.) Non-interference with temperature control loop of humidifier plate.

Since the sides of the chambers are clear, an optical sensor is probably the best bet as they are accurate, non-contact, and maintenance free. There exist ultrasonic sensors that are pretty interesting, but they are pricey and tend to be for large tanks.

So in order to minimize effect of addition of cold water into the chamber on the output humidity as well as the machines sensory of delivered pressure and flow, the system control needs to have either a very small hysteresis band or be continuous. We can't accidentally make the heater control loop go unstable (unlinkely, considering this is a medical device and the engineers on these tend to do their homework) or introduce too much fluid at once to measurably affect the pressure of the circuit. For this reason my tendency is to use a slightly more expensive linear image array sensor (like you see in flatbed scanners) than multiple discrete sensors and a non-trivial control loop.

I'm thinking something like the TSL1402R for the sensor. Cheap-ish at $14 in one off quantities. Responds to light in the near IR bands, so the LED needed to generate some light would be pretty hard to see and won't be distrubing. Nice easy digital output and a physically large enough array. The major question is whether there is enough contrast between the water level and air level at this wavelength to reliably detect the liquid level. Things like condensation on the side walls could potentially throw this off. Water is pretty reflective in IR, but pretty transmissive in visible: 800nm is cutting it kinda close. The transmissivity of the polycarbonate(?) water chamber depends on forumulation a lot (50 series are clear, 60 series are tinted blue, S10 is frosted). So experimentation in this area is required.

I'd need at least another sensor to act as a failsafe against overflows (these could potentially destroy the motor). So this would probably act based on hardwired logic and kill pump power and/or trigger a fault in the MCU.

A couple issues come to mind using an optical sensor though, for instance if the sides of the chamber are occluded (mineral build up from non-distilled water), the sensor could report false levels. For this reason it'd would be advantages to have the failsafe sensor use a different sensing technology than optical.

As far as pumps go, a peristaltic pump would be the best. They offer absolutely minimal contact with the fluid and are relatively maintenance free (tubing might need to be replaced from time to time, silicone would last longest). Noise would be an issue, so we want to encase it in sound proofing and run it as slow as possible without introducing nasty time delays into the control loop (these things usually have only 2 or 3 lobes making the output highly non-linear).

To interface with the chamber, all you'd need is one small hole to accept 1/8" tubing or so and a non-return valve. That way the refill system could be taken off without adverse operation.

The logic could be implemented on pretty much any microcontroller because the time constant of the system is so huge, a little 8 bit PIC would work great. Component count is relatively small for the board: only need some voltage regulation, a couple power FETS, some passives etc.

The real work would be in the control loop. At best i'd think it would be 3rd order time variant. I'd have to do some math later. The output of the peristaltic pump is a challenge to linearize without additional sensors. I'll have to think about that some more. Operationally, we'd aim to deliver as little fluid as possible in as many discrete steps as possible. This way the temperature of the chamber should be relatively unaffected and the introduction of such a small volume into the circuit should be below the error level. More math to determine what those volumes and times would be needed.

The cool thing is that you could theoretically set up a gallon jug of distilled water, prime the pump tubing, and press go. Wouldn't have to worry about refills for like a month.

Oh there are other ways of going about this that are simpler and more cost effective: completely passive systems that could maybe double or triple the water capacity, simpler systems that work purely on hysterisis control, humidifier in the central air system, etc. but they are simply not as much fun as the above.