Zeo shutting down: export your data!

Although I don’t know anything about the batteries on these units having never looked, I just wanted to correct your misunderstanding about current. The fact the battery charger says 0.5A says nothing about what the battery can take, and you don’t need a small charger to charge it. As you say it is very unlikely a battery of that size can be 1.5Ah.

Unlike voltage, current is a limit, therefore if you are charging a 3.7V battery, it doesn’t matter if your charger can give 0.5A or 1000A, it will charge at exactly the same rate, and assuming the battery needs less than 0.5A, the same current will go into it.

As to the other matters, if it is soldered to the board, then your options are limited. You can measure what the voltage is when it is on charge and if it gets up to the 3.7v, and how quickly it discharges, but to do much else you would have to either unsolder or cut the leads off.

[I’d originally posted technical details on testing the Zeo battery, etc., but have deleted them, though correct. Lithium batteries are not as harmless as the batteries we were used to before; they can explode and catch fire. If you know what you’re doing, you don’t need me to say anything; if not, don’t do anything more than replace. Don’t try to rejuvenate batteries, or recharge them externally.]

Thanks @tolerant for your information, very useful. I won’t quote in full to keep things short. You report that you’ve opened your non-charging Zeo (mobile phone version) transmitter and are looking to get it working (details on the quantifiedself “Zeo shutting down: export your data!” thread a few posts up).

Batteries are usually soldered into such assemblies, and must be unsoldered for replacement. Be very careful not to heat the battery itself significantly. If you are not accustomed to solder work on miniature equipment you may cause damage to the circuit board.

It appears impossible to get a replacement battery that will fit in the transmitter and supply sufficient energy for a night. Maybe somebody will eventually find one? The external volume of the original battery (20x20x2) is about 80% of that of a rechargeable ML2032 (20mm dia, 3.2mm thick, 65mAh), which suggests ~50mAh for the Zeo battery.

You can replace the battery, either temporarily for testing or permanently, with a larger Li-ion or LiPo one of the same voltage (any single cell will be rated at 3.6-3.7V), being extremely careful not to short-circuit it in the process. If permanent, you’ll have to improvise a casing. Capacity needs to be at least 50mAh, maybe a bit more if I’ve underestimated, but a somewhat larger capacity doesn’t seem to be a problem (it might take longer to indicate full charge, but should last a night even if not full).

I’ll post a summary of information on the battery below, for reference.

HTH


(I wanted this to be a separate message, but it has been appended to my previous contribution.)

This is a summary of information on what is know of the battery used by the transmitter of the Zeo sleep monitor (mobile phone version). The transmitter for the clock version is likely to use the same battery. In case this is copied or posted elsewhere, this was originally posted at
https://forum.quantifiedself.com/showthread.php?tid=654&page=19
and quotes and uses information supplied by quantifiedself member “tolerant”.
There is discussion at quantifiedself of ideas on testing and replacing the battery.

[quote=“tolerant, post:373, topic:561”] (much edited quote): I’ve opened my unit with a knife because the battery doesn’t take charge; the battery is a PP312122AB 3.7V Li-polymer Rechargeable Battery. Square, size 2 x 20 x 20mm, non-standard, not available.

This hasn’t given me much of a clue how it could have been opened more easily. I’m sure it could have been done more carefully, without doing any damage to the case, but at the end of the day, while there are two tiny plastic pins that go into tiny holes there, and while it appears to “clip” to some extent when I tried to reassemble it I think that the case is ultimately held together by glue. The person who does it next (I hope it’s not me again trying to open another faulty unit) should try suction cups as a matter of first resort. The charger is rated 5v 500mA.
[/quote]

The external battery volume (20x20x2mm) is about 80% of that of a rechargeable ML2032 (20mm dia, 3.2mm thick, 65mAh), which suggests ~50mAh for the Zeo battery; this is an estimate only.

From memory, I think that the battery of the clock version runs for about 16 hours on a full charge, and takes about 1 hour’s charging (not 3) until the display indicates full charge (I tested these times). This may vary between units, and my recollection may be wrong!

A comment about lithium technology (Li-ion, LiPo) rechargeable batteries: I believe that lithium cells are very susceptible to damage by either over-charge or under-charge. This is avoided in equipment by circuitry that does not allow these circumstances. Lithium rechargeables self-discharge only very slowly, so that they may be left for long periods; however they may eventually self-discharge below the safe limit and become unusable if left for long enough. It is not safe to try to rejuvenate them in an external charger without protection circuitry. I don’t know if this form of damage actually occurs from personal experience; but I have a spare transmitter, and swap the two occasionally to avoid the possibility.

Here is a copy of the picture:

@woodinblack, how exactly would I measure the voltage of the battery, whether or not it’s on charge? Do you mean with the case of the transmitter open or closed? Can I use the gold-coloured “things” at the bottom of the transmitter to measure the voltage remaining on it? If not, can I use the points on the circuit board to where the battery wires are soldered to measure the voltage? As regards the 0.5A written on the charger, I don’t know what it could be referring to. A LiPo battery should not generally be charged at higher than capacity. Assuming the Zeo battery has a capacity of 50mAh, it could safely only take 50mA of current when charging.

@pol098, I know that when trying to restore LiPo batteries you need to watch them closely and have a LiPo fire/explosion-proof bag. You’re right that the battery is soldered onto the circuit board, and unsoldering and resoldering it could easily damage the circuit board. Maybe an easier way would be, if I was seriously thinking of restoring the battery, to cut the wires at about half-way point and after rejuvenation is complete, to reconnect them by soldering.

I finally got around to updating my zeo to version 2.6.3O. I sucessfully erased the card collected some data. It displays in ZEOVIEWER 0.2.6 and the data displayes there. I am aslo able to export to a scv file… but now the problmes:

1: the zeo machine says that all of teh data is 0. zq, total z (which is fine I was awake) will not display any data or graph, it says no data to display. I was awake the whole time just tryingto gather data.

2: I have sleep apnea and upload my data into a program named sleepyhead. It is expecting a csv file. When I try to upload the file from the viewer it kills sleepyhead.

I will try to gather data tonight and see what happens.

I have read through all of this 19 pages of post, and there are several ways of viewing the data.

So the question:
what is the best way to get a csv file that matches the file from zeo?

is there any suggestions on why there is no graph, it should show the awake graph?

I did see graph data while it was being collected.

you need to measure the voltage at a point where it comes from the battery, either the terminals, or some point on the board where the power exists, so the transmitter would have to be open, unless you can run some leads outside of it.

[quote=“tolerant, post:376, topic:561”]
If not, can I use the points on the circuit board to where the battery wires are soldered to measure the voltage? As regards the 0.5A written on the charger, I don’t know what it could be referring to. A LiPo battery should not generally be charged at higher than capacity. Assuming the Zeo battery has a capacity of 50mAh, it could safely only take 50mA of current when charging.[/quote]

Yes, the points on the circuit board.

The 0.5A written on the charger is the maximum current output of the charger, before the voltage dips. So the charger can give any power up to that amount at the specified voltage.

The 50mAh on the battery doesn’t indicate how much current the battery can give or take, it indicates the actual capacity, so it could provide 50mA for an hour, or 100mA for 2 hours or 50mA for 10 hours. This is only a rough measurement as other factors are in play, but that is the basic idea.


From my initial working on the viewer (and I guess you are using 0.2.6b?) it displays data from the epoch preceding the onset of sleep. So if you don’t sleep it won’t show you anything. Try going to sleep and seeing it. This program is very flakey though, so don’t hold your breath too much!

There are a few sleephead users on this forum, I believe that sleepyhead used to load csv files from zeoviewer but now doesn’t, so maybe an older version would work but that is just second hand info.
Obviously whatever you gave it was not what it was expecting if it crashed.

the zeo doesn’t produce any csv files.

My application exports CSVs, in a very similar format as the original Java viewer, if that is what you need, but again, if it crashes sleephead in that format it won’t make much difference.

Did you backup your data before you updated? My application can deal with encrypted or unencrypted, or both in the same file.

[quote=“tolerant, post:376, topic:561”]
how exactly would I measure the voltage of the battery, whether or not it’s on charge? … Assuming the Zeo battery has a capacity of 50mAh, it could safely only take 50mA of current when charging.[/quote]

Anywhere where the two terminals of the battery are connected, wherever it’s most convenient. If there’s a wire soldered to the circuit board going to the battery, the solder point is fine. If the battery terminal is exposed rather than insulated, also fine. Any solder pad on the circuit board connected to a terminal, also. Use a multimeter set to the appropriate voltage range. It’s possible that you get a measurable voltage across the charging contacts on the transmitter related to the battery voltage; if you hadn’t already opened it I’d have suggested probing the contacts immediately after the 30-second-lasting charge to see if there’s a voltage that decreases rapidly. Avoid bridging or short-circuiting any connections when applying the probes.

[quote]
Maybe an easier way would be, if I was seriously thinking of restoring the battery, to cut the wires at about half-way point and after rejuvenation is complete, to reconnect them by soldering.[/quote]
Whether you try to rejuvenate or test, or simply replace the battery, you’ll have to disconnect it. For a battery connected by wires rather than pins through the board, cutting them is often the best way. It also makes replacing easier (if you don’t cut the wires too close to the board), unless the assembly is so tiny that’s there’s no room to pack in soldered insulated joints.

Regarding the charger current rating: it just has to be capable of delivering the required charging current, higher rating is fine but ignored. I’ve found the Zeo battery seems to indicate full charge in about 1 hour (if memory serves, I did check it long ago), so the actual charging current is probably about 50mA (it would be, say, 5mA for a 10-hour charge, and so on). Any charger that can supply AT LEAST that current is fine. The manufacturer would buy a standard charger; 0.5A at 5V is simply probably the cheapest these days. If you want to fill tiny medicine bottles with water, you wouldn’t install a special tap that delivers a trickle, you’d just use an ordinary tap turned on just enough to deliver a trickle.

HTH

For Bedside model users:

Wilson, SleepyHead no longer supports Zeo, but you can get a CSV file that matches the old Zeo export by using Woodinblack’s viewer here: http://www.alienrat.net/software/zeomonitor.html

The primary differences as far as the CSV export is concerned is that the original contained the 5 minute hypnogram (used by SleepyHead) and Woody’s contains the 30 sec hypnogram data only. If there is enough demand, perhaps the author of Sleepyhead will incorporate the 30 sec data hypnogram into the program. The 30 sec hypnogram presents much more detailed information and is better suited for use in SleepyHead than the 5 minute hypnogram was.

BTW, it is difficult to test Zeo by wearing the headband while awake because it does not start recording until after a sleep stage has been determined.

I can not determine from your post which method you are using to retrieve your Zeo data, but if you’re not using Woody’s viewer referenced above, you should.

Thanks, I have teh bed side model. I was able to get a full nights data and the ze is working normally. I do have woody’s viewer and it looks good. I got in tough with the sleepy head developer and asked if he could look at the difference in the CSV files. I will pass on to him that the main difference is that the new file had data at 30 second intgervals, not 5 minutes, that make it an easy fix.

Thanks for the help. For now woodies printouts will do what I need for working with my neuro. - hormones messed up since a pitituary tumor and radiation to finish it off,

Thanks again.

I sent Mark an example of both the old and new CSV files. I’ll send you a copy by email.

Thanks guys for your input on the battery issue. I did measure the voltage at a number of places, and every reading was exactly zero. I used another 3.7V battery as a reference to make sure there was no issue with the voltmeter. I haven’t tried to measure the voltage when the unit is charging or immediately after taking the battery off the charger. I think trying to rejuvenate a zero volt battery would be an expensive and difficult tasks, considering I don’t have the required equipment or skills. It should be much cheaper and easier to find another battery that fits into the unit. For that I’ll have to do an estimate of the thickest battery the unit will possibly take and see if one is available. If I do find one, I’ll write an update.

Hi Gary,

I’m sorry it took a little time to reply.

ResMed did acquire some of the Zeo assets including intellectual property. The anonymised sleep data were used to make comparisons with our extensive research on the sleep norms and various meta-analyses of sleep data during development of S+ by ResMed.

We have not reconstituted the Zeo platform. However, we are planning to offer a unique program to former Zeo customers (available in the US initially) so that they can build new data quickly and easily with S+. In the meantime, we love the vision that you and the QS movement have shown and want to do all we can to support healthy sleep.

Best wishes,
Colin.

1 Like

Thoughts about releasing the Zeo data? From the sound of it, you have no further use for it now that you’ll be getting additional data from S+ users.

Gary, would it be possible to PM me the contacts for Gary?

I use a resmed CPAP machine, since this is infering sleeping patterns from breathing, I want to see if they could build some if into their CPAP machines which measure the resperiation directly.

Thanks

Ted Wilson

you can also send it to my yahoo email:
theodore_a_wilson@yahoo.com

I also understand if this is not possible.

Thanks

I think you can buy this battery here:

http://www.aliexpress.com/store/product/Original-PP312122AB-3-7V-Li-polymer-Rechargeable-Battery/318150_2032438855.html

US $7.99 + $5.26 shipping

Total: $13.25

1 Like

I made an interesting discovery of possible interest to Zeo Mobile users. Upgrading to iOS 8.1 on my iPad 3 brought back “Sleep History” functionality on the Zeo iOS app. Tapping Sleep History crashed the app in iOS7, but on my device the histograms are back in iOS8! I had prepared myself to lose all functionality, so this was a happy surprise.

I have a similar problem with 3D Vision glasses, and a dead battery. These are LiPo batteries as well.

Based on my reading and understanding of batteries (EE background, not totally guessing. :s) you should be able to replace the LiI/Po battery there with something used for RC planes. All the batteries are very similar, the 3.7V is the key.

Do not under any circumstances replace the battery with a coin-cell or anything not LiI/Po because there is a charging circuit there that will interact badly with non-chargeable batteries. NiCd, NiMH are the wrong voltage and will not work.

It’s possible battery itself has a circuit board, that’s how the 3D Vision glasses work. But your description of the tabs being soldered to the board itself suggest that it’s a stock LiI battery, and that the overvoltage protection is provided by the headband circuitry. This is what I would expect design-wise.

Here is a post regarding battery replacement in the glasses, and has some detail about sourcing the LiI battery.

http://3dvision-blog.com/8323-again-about-replacing-adead-battery-in-3d-vision-glasses/

I think Martinus nailed it with that link above. You should be able to just solder a new one on the board.

http://www.aliexpress.com/store/product/Original-PP312122AB-3-7V-Li-polymer-Rechargeable-Battery/318150_2032438855.html

In terms of soldering in a new battery, that is going to be very easy for someone with a little experience. Without experience, you’d be better off to find a local tech-shop of some form, or DIY group, or MakerFaire type people who can do this in their sleep.

1 Like

A useful contribution by @bo3bber, above. In the last few pages of this thread overall in we seem to have gone from no information at all, to quite good information on replacement of the Zeo transmitter battery when it fails, with information from @Tolerant on opening the device, and a likely source of batteries from @Martinus. The particular tiny batteries available from the radio-controlled model supplier in China (linked from the URL that bo3bber gives) are a bit too thick for the Zeo, at 5mm. They had a huge range of batteries, but, from a quick glance, nothing suitable. The battery topic here starts at post #362. Unless I’ve missed something, it’s probably the only information available on the topic.

Maybe one of the first posts could say that there is a lot of information on battery replacement starting with the post mentioned above?

1 Like

BTW, based on my research, the LiI batteries like these have a working life of about 5 years. That’s whether they are used or not. They don’t have a better shelf life than about 5 years. Apparently they start to degrade chemically as soon as they are made.

So, that means that all the batteries in headbands will die at some point, even if left on the charger. Or even if disconnected. And that if you find a replacement, try to make sure it’s not been sitting on a shelf, or get an idea of date of manufacture.

To contradict my earlier statement- you could use a coin cell Lithium battery as a replacement, but you’d have to make an on/off switch and never put it on the charger. As long as you can get 3.7 Volts the head band should work. I think you would be better off working with the correct chemistry though.

As a former Zeo user, I recently got an email from Resmed offering a discount on the S+. I’m really curious to see how these compare. Are there any current Zeo users out there who have used the S+ in addition to the Zeo? I’d love to see some feedback on whether the data returned by each device matches.

Any data points yet?

I’ve been using the Zeo Mobile for a few months, and recently purchased a Resmed S+. It’s taken me ~3 weeks to fine tune its position such that my wife’s movement doesn’t interfere, but I think it almost there now… I have done a couple of comparison tests with the Zeo and it seems to roughly match up but not as well as I’d expect. However I dont know how much of the discrepancy is due to the S+ not being positioned properly!

I’ll gather some more data over the Christmas break and post it here.

Some general comments on the S+ from my experience so far:
PROS:

  • Great user experience, flawless Bluetooth connectivity (Android phone)
  • Microwave sensor is very impressive for a consumer device
  • Resmed have obviously put a lot of work into it

CONS:

  • Incredibly difficult to find ‘just the right’ positioning of the pod. I haven’t done a great deal to isolate the root cause of the issue, but it consistently thinks I’m awake for a total of ~1hr during the night when the Zeo (and my memory) says otherwise
  • Setup and troubleshooting documentation from Resmed is very poor. They provide a ‘not to scale’ diagram of the ideal setup which is of little help, and the help section regarding setup issues is very scarce. In their defense I did receive a couple of good tips after contacting customer support, and they did offer to help fine tune setup if I send photos etc. My suggestion would be: provide detailed diagrams of setup requirements, a detailed troubleshooting guide, and some live feedback to the user as to the quality of the signal. The most frustrating thing is that you have to wait until the NEXT DAY to figure out how the change in setup affected sleep detection. Hence the incredibly cumbersome fine tuning process…
  • No access to any raw data beyond the hypnogram and sleep stage breakdown (and temperature/light level)

Seems like a lot of cons, but really I’m quite happy with the S+. It has a lot of potential, so hopefully some of these issues can be fixed with firmware updates.

1 Like