Work In Progress: A new way to let "magic smoke" escape from Resmed CPAP machines.

[rickst29]

The Resmed "Airsense 10" uses 24VDC power, but it also requires a 3rd "pin" on the Power Supply to feed it 3.3V. I had read that this is "tested" at startup time, to verify that a genuine Resmed PSU is attached and perhaps to also verify whether the PSU can supply 90 Watts of power, or some lesser amount.
- - - -
Picture hints: Power comes into red twist connectors at the left. From the Booster, "Yellow" is +24V, and "Black" is Ground (international color coding), each going into 5-way "push-down" connectors. A 2700 uF Electrolytic "Filter Capacitor" has one wire into each connector, with proper Polarity observed. (My thanks to Bruce Perens for correcting my initial mis-wiring of this capacitor). Another pair of connectors leads to the "Buck Converter" mini-board inputs ("+ input", "- input)". The last wires from each Connector connect to the Resmed output cord leads "Ground" (outer cylinder) and "+24V" (inner cylinder), using yellow twist connectors.

Input to the "Buck Converter mini-board" (24V) is tuned to create just 3.3V output. I didn't bother to connect the "Output -" terminal, it is wired in common with the "input -", and Resmed doesn't want multiple "Ground" connectors. Buck Board "+ Output", at 3.3V feed directly to the Resistor, which then connects to the "middle pin" wire on the Resmed output Cord.
- - - -
Using a jumper cable, I opened the hood on a turned-off SUV and connected the input power cord to the SUV battery. I verified the "+24V Bus" to be running at 24V, and the "Ground Bus" to have continuity with SUV frame (that will become TM Frame "Ground" during Trips). Then, I moved my Voltmeter over the the Buck Board output terminals, and dialed down the output Voltage until it showed 3.3V. Finally, I verified the same values (Ground, +24V, and +3.3V on the correct parts of the tri-axial connector.

I stepped inside, unplugged the Resmed CPAP, and re-verified the exact same Voltages to be present on the Resmed AC-> Power supply connector. Then I took the Resmed outside and plugged it in: The power-up sequence was the same, the power switch ran correctly, and it also auto-adjusted for an "apnea event" - because I was not yet wearing the mask. After presenting it with a few seconds of "good breathing", it returned to normal mode. I shut it down, waited for "SD Card Activity" to be finished, and then pulled the plug. It worked great!

Subsequent to taking the picture, I changed the filter Capacitor from 2700uF (35V) to 1000uF (50V).

[BrucePerens]

You don't want their $330 battery, huh?

Given that there are aftermarket R9 power supplies, and they're pretty cheap, I guess 3.3V is all there is on that pin, not some sort of more sophisticated digital rights management.

I am more worried about releasing the magic smoke from your face.

The problem you could run into is that you are going from a power supply that had a negative wire that was isolated from ground to one that connects the negative wire to the chassis ground of the camper and the protective ground for 120V power. If the CPAP is double-insulated, this will probably not be a problem. If it is not double-insulated and has metal components connected to the negative line that touch your body:

1. It might connect your face to ground, which would not be nice if you happen to touch something 120V hot.

2. If metal parts are connected to the positive wire rather than the negative one, letting those parts hit something grounded can create a short.

3. A 12V circuit across your wet insides really does hurt. Try licking the top of a 9V battery.

These are probably unlikely. Most human-contact equipment is carefully isolated. But keep in mind that you are replacing one of the isolated components with one that is not isolated.

[rickst29]

Thanks, Bruce! I know you to be expert in these things, and I'm definitely not.

I don't think that I have that problem (but I might have a different problem). The ground wire comes in from CPAP, goes into the 5-way push-down Connector. The shared Connector's source of Ground Voltage is the big Converter's 24V "Output -" lead wire. They gave me different Ground wires for ~ 12V "input" and 24V "output", the output Ground may very well be a "floating" termination for the 24V side of things.

But if it is, I have (perhaps) a small mistake on my 3.3V output "Ground": I've tied it directly to the big Push-Down, because it's CPAP return current comes in on the shared CPAP "Ground". Q1: Do I maybe need to make a change there?
- - - -
My "Different Problem" is more basic and bigger: I have wired a "filter capacitor" with +24V on the "+" side, and +24V to the CPAP coming out the "-" side. Presenting just a small in-line resistor, with no energy storage and release? Q2: Do I need to move that Capacitor "-" over to the 24V Ground Connector, or is it effective in the current position?

(My idea had been: Resmed shares the computer stuff AND the DC motor on the same 24V circuit. As the rotor spins (with, typically, 16-24 contact areas), the gaps present short moments of very high resistance - creating an asymmetric square wave of current, running at 16-24x the motor's rotational speed. I SWAG a frequency of 600-3000 Hz for this waveform. The AC-DC Resmed supply which I destroyed, in order to to cannibalize it's output cord, has a 560uF electrolytic there. But I can't tell how the connection was made, and I now suspect that their Capacitor "-" was a Ground connection, rather than inline "+". )

[Padgett]

Something like this maybe ? Spec says can handle 8a 12vdc in 12vdc out.

[BrucePerens]

Originally Posted by rickst29

They gave me different Ground wires for ~ 12V "input" and 24V "output", the output Ground may very well be a "floating" termination for the 24V side of things.

You can measure continuity between them. Usually these are boost converters and do not isolate their inputs from their outputs.

3.3V output "Ground": I've tied it directly to the big Push-Down, because it's CPAP return current comes in on the shared CPAP "Ground".

Why don't you list out all of the connections on the CPAP and your power supplies, so that I'm sure what we're talking about.

I have wired a "filter capacitor" with +24V on the "+" side, and +24V to the CPAP coming out the "-" side. Presenting just a small in-line resistor, with no energy storage and release?

Filter chokes go in series, filter capacitors go in parallel. The capacitor goes between + and - of your power supply, and make sure it's rated at least 1.5 and preferrably 2 times the operating voltage, and you have the plus wire connected to the plus of the power supply. Electrolytic capacitors break if they are reverse polarized. Tantalum ones that are reverse polarized explode like a gun shot.

I am seeing pre-built converters with 12V plugs or car battery clamps for Airsense 10 for $85 online. It might be better to just buy one.

[rickst29]

Originally Posted by Padgett

Something like this maybe ? Spec says can handle 8a 12vdc in 12vdc out.

Similar, but creating regulated 24V instead of 12V. (From there, I use the mini-board to create a sort of fraudulent 3.3V 'Resmed Authenticity Certification'.) I used this one: http://www.ebay.com/itm/AweKing-Wate...R/232401317134

[rickst29]

Originally Posted by BrucePerens

You can measure continuity between them. Usually these are boost converters and do not isolate their inputs from their outputs.

Zero Ohms -> yes, they''re connected.

The CPAP connector is identical to that used in older Dell Notebook AC Adapter PSUs: The 3 wires which lead into the cable, from a "refurb" AC/DC Resmed PSU which destroyed to acquire the cable were "mapped" as follows:

• The RED wire at the cable end connected the "outer" cylinder as "Ground".
• The WHITE wire connected the "inner" cylinder, as +24VDC.
• The BLACK wire connected the "center pin", as +3.3VDC.

I have read, elsewhere, that the 90W PSUs present a series Resistor of 2.7k Ohms along the 3.3V wire. I could test that, in a basic way, by running "center pin" output from the original AC-DC Power supply through an identical resistor, and see if the Voltage drop through the load/test resistor ~ 1.65V. If the "Genyoowin Resmed Test" is more complex (RC timing, or data-related) then I won't be able to pass it.

Originally Posted by BrucePerens

Filter chokes go in series, filter capacitors go in parallel. The capacitor goes between + and - of your power supply, and make sure it's rated at least 1.5 and preferably 2 times the operating voltage, and you have the plus wire connected to the plus of the power supply. Electrolytic capacitors break if they are reverse polarized.

Thanks! I'll move the "-" wire over to the "Ground" Connector, and bring the Resmed cable "White" wire over to the "24V" Connector. If my gadget doesn't work really well, then I'll buy the genuine article, and simply upgrade its 12V source (plug or clamps) into my already purchased twist-lock set.

[BrucePerens]

Originally Posted by rickst29

I have read, elsewhere, that the 90W PSUs present a series Resistor of 2.7k Ohms along the 3.3V wire.

OK. This is current-limited logic "high" going directly to a GPIO pin of a microprocessor. You need to protect it from static.

About the only timing I'd expect is that 3.3V should not go high until 24V is stable.

Thanks

Bruce

[rickst29]

Part #1: A sacrificial AC-DC power supply for Resmed A10. (Refurbs run about $18 on Ebay.)

Part #2: A DC-DC "Boost" Converter, creating regulated 24V output from 11-15V input. I used and recommend this one, which costs $18 as I write this:http://www.ebay.com/itm/Waterproof-D...-/272574605057.

Part #3: A DC-DC Variable "Buck" Converter, to create 3.3V output from 24V input. I used and recommend one like this, which costs barely $2: http://www.ebay.com/itm/Step-Down-LM...t/232405283762.

Part #4: A Resistor (2.7k Ohms).

Part #5: An electrolytic Capacitor (50V 1000uF).

Part #6: A project box, like this: http://www.ebay.com/itm/158mm-x-90mm...W/182278707309

Miscellaneous Parts: some wire connectors An input power cord with strain relief (and some kind of 12V Connector); some #16 stranded wire to hook it all up, and some electrical tape.