Jolla Solar Half prototype

[skvark]

I remembered last night that I had solar panel in my scrap box. The size was perfect for Jolla and a quick test proved that it can be used to charge Jolla too:

- voltage is about 6 V (measured with multimeter)
- panel outputs 40-80 mA under a lamp (measured with multimeter)

I haven't seen sun in a month or so here in Finland so the test was performed under 70 W halogen, distance ~10 cm.

http://relativity.fi/jolla_solar_toh/solar_toh7.JPG

More pictures:

http://relativity.fi/jolla_solar_toh/

Today I modeled new other half for the panel and printed the new model. The panel fits nicely. I dit not take yet pictures from the inside, but I soldered some wire to the panel outputs and glued wires so that they are in contact with Jolla power pogo pins.

UPDATE: Jolla's CSD tool shows charging current of -10 mA (at winter in Finland). This means the panel outputs about 100-110 mA at winter in Finland (the idle discharge current is about 100 mA). There's still room for 60-70 mA more, but this can be tested here when the sun is shining a bit higher at summer.

Under the 70W lamp CSD tool shows +50-60 mA, so the panel is giving about 50 mA out but the battery is still discharging. The phone still says that charger is connected.
UPDATE END

I have to test the Solar Half more, but it indeed seems to be working rather nicely :cool:

EDIT: .stl file available here: https://github.com/skvark/TheOtherHalf/blob/master/thesolarhalf.stl
EDIT2: I misinterpreted the CSD tool. Yes, the phone says it's charging and yes, the CSD tool says "Charging" but under a lamp the panel is just limiting the disharge current. If the battery is actually getting more juice, the current is shown as negative value. However, the original 70-100 mA is still realistic.

[clovis86]

wow amazing other-half :D

I imagine your 70W lamp is really lower than the direct sun exposition ?

[Chinoman10]

What a coincidence that your solar panel from your scrap box was a perfect fit for Jolla :D

I'd love to see something like this commercialized.

Say...how much did you pay for that solar panel? (just a rough guess, I understand if you can't remember exactly)

Thanks in advance.

~Chinoman10.

[pichlo]

I imagine your 70W lamp is really lower than the direct sun exposition ?

I'd imagine so. Direct sun exposure is about 1kW per square metre. At high noon, with the receiver exactly perpendicular to the incoming sunlight.

By my VERY approximate calculations, his solar panel received about 5W from his 70W lamp. Most of it in the wrong wavelenght. So getting 360-480mW out of it (6V x 60-80mA) is actually not too bad a result.

[aironeous]

Next Jolla phone should have audios tech on the touch screen. That would eliminate need for a solar other half.

[skvark]

What a coincidence that your solar panel from your scrap box was a perfect fit for Jolla :D

I'd love to see something like this commercialized.

Say...how much did you pay for that solar panel? (just a rough guess, I understand if you can't remember exactly)

Thanks in advance.

~Chinoman10.

I bought the panel from Adafruit (if I remember right) over a year ago. And yes, it was perfect :D

I think it was 10 USD. Panel: http://www.parallax.com/product/750-00030

So it outputs in direct sunlight max. ~167 mA.

[pavlaras1]

Try to contact with some Solar Panel Factories to send you some test samples in the same size but with higher efficiency panels with specs such us 6V - 3W.
I already use 6V - 2W Solar Panel for my mobile phone Nokia 1020 (130mm*71mm).
I measure in direct sunlight at 14:00 about 220mAh.
Cost me 25 euros.

[kimmoli]

hmm.. seems that this is available in many places, though parallax is out of stock. e.g. Mouser.. Need to get one :)

just google "Solar Panel 6V @ 1W, 125 x 63 mm"

[dirkvl]

so, no components between the panel and the phone? so the phone has regulator that limits current for preserving 5v?

[minimos]

so, no components between the panel and the phone? so the phone has regulator that limits current for preserving 5v?
Looking at what rainisto said in the other thread about wireless chargers ( http://talk.maemo.org/showpost.php?p=1404322&postcount=42 ) a 'dumb' 5V source should be the best option.

[skvark]

so, no components between the panel and the phone? so the phone has regulator that limits current for preserving 5v?

No other components than the panel and wire. And yes, I think there's regulator (because the connectivity & TOH i2c specs are still missing can't be sure) :rolleyes: There's more information about the power input in wireless charging thread: http://talk.maemo.org/showthread.php?t=92059

EDIT: And you have to glue a plastic pin like in official TOH's to the panel next to the battery so the microswitch will trigger. Won't charge without it.

Actually that's not true. It charges without the pin too. Hmm..

[dirkvl]

microswitch, good to know!!

[droll]

so what happens if you plug the phone into the wall charger with this toh and pput it under the sun at the same time?

[skvark]

so what happens if you plug the phone into the wall charger with this toh and pput it under the sun at the same time?

I don't know, but I hope there's some circuitry or logic which selects only one of the inputs when they both are available.

[droll]

maybe check dbus messages and see if you can differentiate between the two different power sources. if it is distinguishable, there might be hope.

[pichlo]

I expect bad things to happen if you plug in a wall charger and do not put the solar panel under the sun at the same time. Or even without the charger, depending on how it is wired.

The reason is that the photovoltaic cells suffer when subjected to reverse voltage. Which means, if you connect the panel to a voltage source higher than the voltage produced by the panel itself.

So some sort of protection is essential. At least a diode if nothing else. This has been mentioned a number of times, for example here (http://talk.maemo.org/showthread.php?p=1352112#post1352112). Of course, your solar panel may have a diode built in, but I somehow doubt it.

[skvark]

I expect bad things to happen if you plug in a wall charger and do not put the solar panel under the sun at the same time. Or even without the charger, depending on how it is wired.

The reason is that the photovoltaic cells suffer when subjected to reverse voltage. Which means, if you connect the panel to a voltage source higher than the voltage produced by the panel itself.

So some sort of protection is essential. At least a diode if nothing else. This has been mentioned a number of times, for example here (http://talk.maemo.org/showthread.php?p=1352112#post1352112). Of course, your solar panel may have a diode built in, but I somehow doubt it.

Answer from #jollamobile:

13:48 < skvark> does someone know how jolla reacts if there are 2 chargers connected, another to the usb an another to the back power pins?
13:50 < Jonni> skvark: back power pins are disabled when usb cable is connected.
13:51 < Jonni> (and yes, I've tested it with my wireless charger toh)

And I'll add a diode there to be sure that nothing bad happens if the voltage gets reversed.

Edit: If there is a reverse protection already in Jolla, then the diode would be useless?

[droll]

excellent! the pins are disabled when USB is connected. aaaaaaaah.

slightly off topic question: if the back power pins are disabled, will a TOH with a spare battery be impossible to charge when you plug in the wall charger? or is the reference in IRC only for input power pins?

[juiceme]

How about printing the 2ndhalf so that there's an overhanging flap that covers the USB port when the solar kit is in use :D

[juiceme]

slightly off topic question: if the back power pins are disabled, will a TOH with a spare battery be impossible to charge when you plug in the wall charger? or is the reference in IRC only for input power pins?

You would naturally build the extra-battery-capacity 2ndhalf so that it connects to the battery terminals of the device.
(of course you need to design it so that it does not add to the existing battery but replaces it)

[skvark]

excellent! the pins are disabled when USB is connected. aaaaaaaah.

slightly off topic question: if the back power pins are disabled, will a TOH with a spare battery be impossible to charge when you plug in the wall charger? or is the reference in IRC only for input power pins?

The answer from IRC only covers the wireless charger + usb cable test so I don't know. This is why we really need the official specs from Jolla.

Edit: juiceme has the answer for you :)

[skvark]

...
EDIT: And you have to glue a plastic pin like in official TOH's to the panel next to the battery so the microswitch will trigger. Won't charge without it.
...


Actually that's not true. It charges without the pin too (took it off to test again). Hmm..

[Markkyboy]

This is brilliant stuff! I'm really impressed and excited by this!
@skvark - I downloaded and printed a copy of your 'other half', it came out really nice, but sadly, it doesn't quite fit the phone in the length.
It was that tight that it started to bow the phone so off it came!
I used Netfabb Basic to cut a section from the lower half of the model and stitched it back together in Blender.
I'm going to my friends tomorrow to print it off, hopefully it won't be so tight!, perhaps we have some settings wrong on my mates printer, but when we measured the model the dimensions were as you set them.
Has anyone else experienced the same thing?, I've added 1mm to the length of the model to compensate but I will also check the software for the printer again.
Any info appreciated.
Regards
Mark

[skvark]

This is brilliant stuff! I'm really impressed and excited by this!
@skvark - I downloaded and printed a copy of your 'other half', it came out really nice, but sadly, it doesn't quite fit the phone in the length.
It was that tight that it started to bow the phone so off it came!
I used Netfabb Basic to cut a section from the lower half of the model and stitched it back together in Blender.
I'm going to my friends tomorrow to print it off, hopefully it won't be so tight!, perhaps we have some settings wrong on my mates printer, but when we measured the model the dimensions were as you set them.
Has anyone else experienced the same thing?, I've added 1mm to the length of the model to compensate but I will also check the software for the printer again.
Any info appreciated.
Regards
Mark

I'll look into this, but it seems there's about 0,4 mm error in the model length and PLA (if you used it) tends to shrink max. 1%. I'll keep you updated.

[Markkyboy]

I'll look into this, but it seems there's about 0,4 mm error in the model length and PLA (if you used it) tends to shrink max. 1%. I'll keep you updated.

Thanks for replying. Yes, PLA was used and I understand what you're saying about shrinkage/tolerances.

I will post back later tomorrow after another printing session! :)

[skvark]

I did some measuring with the CSD tool (= phone is in airplane mode and I have turned the tohd.service off). Charging current under the 70 W lamp stabilized to 50-60 mA and some spikes go to 70-80 mA. If the CSD tool can be trusted (and the panel is actually charging the phone, like the phone and the tool says), there should be about 1% more battery every ~30 mins. To confirm this, I ran a short test:

I started testing at 18:00 (battery level was 85 %). Strange thing is that the UI does not update the battery level without a reboot (bug?). So I rebooted the device couple of times during the loading period. After every reboot there's was always 1% more battery level. Reboots were at 18:45 and 19:45. After the second reboot device reports battery level of 87 %. So getting 1% takes more than 30 mins.

Some conclusions:

- the 1W (125x36mm) panel is enough to upkeep the battery level and to charge the phone very very slowly (under a lamp, sunlight it would be maybe 2-3 times faster)
- for faster charging more efficient panel is needed

[pichlo]

So getting 1% takes more than 30 mins.

Or not but the reboot sucks some juice. Not sure about Jolla but on N900 a reboot always does.

- the 1W (125x36mm) panel is enough to upkeep the battery level and to charge the phone very very slowly (under a lamp, sunlight it would be maybe 2-3 times faster)
- for faster charging more efficient panel is needed

125x36mm = 0.0045 m2. At full sumlight that should receive about 5W of energy. The best solar panels have about 20% efficiency. So it is not even theoretically possible to get more than 1W from a solar panel of that size without revolutionary technology advances. Sorry.

[captainofiron]

very cool stuff

[skvark]

Or not but the reboot sucks some juice. Not sure about Jolla but on N900 a reboot always does.



125x36mm = 0.0045 m2. At full sumlight that should receive about 5W of energy. The best solar panels have about 20% efficiency. So it is not even theoretically possible to get more than 1W from a solar panel of that size without revolutionary technology advances. Sorry.

Reboot propably sucks some juice yes. I was referring to this post: http://talk.maemo.org/showpost.php?p=1404978&postcount=7

The panel size is actually 125x63 (made a typo, not 36mm), so it gets close to 2W.

[matimilko]

Just for imformation:

http://www.adafruit.com/products/1485#Technical_Details

http://www.adafruit.com/images/medium/1485_MED.jpg

This one is only 0,77 mm thick! :D

.

[skvark]

Sunny day today :) It seems that the CSD tool is showing negative current when the battery is loading even though there's is text "Charging" when it's positive too (I should have noticed this earlier...). The tool shows now -10 - +10 mA current when the phone is in sun (there's some snow in the air).

So the earlier results are ok but they have been just limiting the discharge current (though I don't still get how the battery can go from 85 to 87 % in 2 hours during the last test when it should have been going down to 83% :confused: (EDIT: this is because the phone was most of the time in stanby though I'm not sure did I have the NFC bug workaround in use during testing). So the panel gives at winter in Finland about 100-110 mA output, which is still 60-70 mA short from the max ouput. If the phone discharges at ~100 mA in idle (EDIT: Idle as in "screen on" and no software running. Standby current is about 10 mA when screen is of) , ~160-170 mA from the panel should result in 60-70 mA loading current. Better and a bit bigger panel should get to over 100 mA easily.

EDIT: Of course if the phone is shut down you'll get the ~100-110 mA more juice per hour.

[Markkyboy]

Hi skvark,
I have a solar panel prized out of a solar charger and as I don't have a soldering iron (yet), I'm using copper tape to 'replace' wires.
My main confusion and question, is which of the GPIO pins am I to connect to on the rear of the phone?, I first thought it was the GROUND and 5V+ (which are side by side), am I right or wrong here?, I feel I'm so close to getting this working but am a little afraid of damaging the phone due to lack of 'electronics' knowledge - any info gratefully received,
Regards,

[skvark]

Hi skvark,
I have a solar panel prized out of a solar charger and as I don't have a soldering iron (yet), I'm using copper tape to 'replace' wires.
My main confusion and question, is which of the GPIO pins am I to connect to on the rear of the phone?, I first thought it was the GROUND and 5V+ (which are side by side), am I right or wrong here?, I feel I'm so close to getting this working but am a little afraid of damaging the phone due to lack of 'electronics' knowledge - any info gratefully received,
Regards,

Yeah, connect the solar panel's '+' to 5V DC input and '-' to GND. Take a look at the page 14 in TOH technical specification. (http://jolla.com/the-other-half-developer-kit) I'm myself just a SW dev with limited knowlegde about electronics :)

[Markkyboy]

Thanks Skvark,
that is how I have it connected, but the CSD tool isn't indicating that any charge is being created.
The panel is definitely working, as I have a very powerful high intensity discharge lamp to work with (400 watts), which send my multimeter off the scale!, but I don't know what the specifications of the panel actually are.
I think perhaps I should find a different way to hook up to the inputs/outputs.
How have you made your connections?
Regards,

[skvark]

Thanks Skvark,
that is how I have it connected, but the CSD tool isn't indicating that any charge is being created.
The panel is definitely working, as I have a very powerful high intensity discharge lamp to work with (400 watts), which send my multimeter off the scale!, but I don't know what the specifications of the panel actually are.
I think perhaps I should find a different way to hook up to the inputs/outputs.
How have you made your connections?
Regards,

What's the voltage of the panel?
Do you have the panel datasheet or link to manufacturer page? Are you sure the connection to the pogo pins is ok (there's voltage in the copper tape, usually only the other side of the tape is conductive because there's glue at the other side)?

[psonek]

Hi,
i'd like to ask if these panels would be suitable for solar TOH:

http://www.flexsolarcells.com/PowerFilm-Solar-OEM-Components.php

I'd probably used two SP3-37 - but they make 6V - could this damage the phone?

Would the voltage drop enough (to 5V) when the battery starts charging?

Would it be possible to cover part of the panel to drop the voltage to 5V?

Thanks!

Radek

[Markkyboy]

What's the voltage of the panel?
Do you have the panel datasheet or link to manufacturer page? Are you sure the connection to the pogo pins is ok (there's voltage in the copper tape, usually only the other side of the tape is conductive because there's glue at the other side)?

Hi,
sorry for a late reply and thanks for getting back to me.

PANEL DETAILS - The panel was removed from a solar charger, of which I believe is; 5.5v 100mA http://www.paramountzone.com/solar-mobile-phone-charger.htm

POGO CONNECTION - As far as I am able to see, my strips of copper tape are in contact and aligned with the pogo pins and the polarity is also correct. Yes, you're right the copper tape has adhesive.
https://scontent-a-lhr.xx.fbcdn.net/hphotos-ash3/t31/1553158_10201564336012513_1328313292_o.jpg

COPPER TAPE - As you can see, I have run the copper tape in one length strips to the corresponding pins.
I realise this is not the best way to make the conenction, but it's all I have at my disposal, for now.
The original wires that were soldered onto the solar panel had left quite a blob of solder behind, so I feel quite certain, coupled with meter readings, that voltage is being carried along the tape, not that the Sailfish CSD tool seems to show any changes in voltage supply.
Regards,

EDIT: It only bloody works!! (My connection, where the old blobs of solder are just not quite making contact), but with reasonable pressure on the panel, it connects and charges! Yay!

https://fbcdn-sphotos-f-a.akamaihd.net/hphotos-ak-frc3/t31/1932498_10201564465495750_584424932_o.jpg

What do you glean from the reading on the phone's screen, Skvark?, good, bad?
Regards,

[skvark]

What do you glean from the reading on the phone's screen, Skvark?, good, bad?
Regards,

You need to compare the reading without panel and then with panel. That reading in the picture indicates that your phone is still discharging at a rate of 136 mA.

That 136 mA is imho very high discharge rate for the phone. My phone gives about 100 mA when running the discharge tool without the solar panel.

With the panel CSD tool shows discharging rate of 50-60 mA (lamp) / 0 - -10 mA (sunlight) (= panel gives 40-50 mA (lamp) / 100-110 mA (sunlight)). Do you have multimeter to measure the panel real mA output?

[skvark]

Hi,
i'd like to ask if these panels would be suitable for solar TOH:

http://www.flexsolarcells.com/PowerFilm-Solar-OEM-Components.php

I'd probably used two SP3-37 - but they make 6V - could this damage the phone?

Would the voltage drop enough (to 5V) when the battery starts charging?

Would it be possible to cover part of the panel to drop the voltage to 5V?

Thanks!

Radek

I myself use a 6V panel (fluctuates between 5-6V depending on light source), but I think that a step-down converter to 5V is necessary for non-proto stuff. I'll have to check what the phone reports as the voltage, those are multimeter values. Covering the panel partly may be possible, I have to test that too.

EDIT:

cat /sys/devices/platform/msm_ssbi.0/pm8038-core/pm8xxx-adc/dcin

Result: 4522986 Raw: 32421

So phone says 4,52 V (cloudy sky). This may get a bit closer to 5 V if I make better connections :D Take this with a grain of salt but it seems to be safe to use 6 V panels.

[Markkyboy]

You need to compare the reading without panel and then with panel. That reading in the picture indicates that your phone is still discharging at a rate of 136 mA.

That 136 mA is imho very high discharge rate for the phone. My phone gives about 100 mA when running the discharge tool without the solar panel.

With the panel CSD tool shows discharging rate of 50-60 mA (lamp) / 0 - -10 mA (sunlight) (= panel gives 40-50 mA (lamp) / 100-110 mA (sunlight)). Do you have multimeter to measure the panel real mA output?

Oh, okay, seems I misunderstand how these things work. I see what you are saying now, I think. A quick peak now at discharging rate, shows it is around 115mA without solar panel.

I do have a multimeter. I'm no expert at using it. With a 20 watt lamp over the panel, I get an output of around 5mA (again I maybe reading this wrong), or do I add a zero and it becomes 50mA?
My apologies for being a dickhead where this is concerned, I did all this at school, many. many, years ago and have forgotten ALL OF IT! :D :(
Thank you for your help so far :)

[skvark]

Oh, okay, seems I misunderstand how these things work. I see what you are saying now, I think. A quick peak now at discharging rate, shows it is around 115mA without solar panel.

I do have a multimeter. I'm no expert at using it. With a 20 watt lamp over the panel, I get an output of around 5mA (again I maybe reading this wrong), or do I add a zero and it becomes 50mA?
My apologies for being a dickhead where this is concerned, I did all this at school, many. many, years ago and have forgotten ALL OF IT! :D :(
Thank you for your help so far :)

The value depends on which scale the multimeter is. For example my multimeter has mA and µA and A. There's 1000 mA in one A etc. (see: http://www.adafruit.com/index.php?main_page=popup_image_additional&pID=308&pic=1&products_image_large_additional=images/large/308wire_LRG.jpg This is very basic stuff :)

[Markkyboy]

Thanks skvark. I found my skytronic multimeter and put a fresh battery in!

Right new figures, which show much more promise! I set the multimeter to 200mA setting and showed the panel the light, the reading shot off the scale, leaving a number 1 on the left side of the meter screen, this represents that it went past the given rate!

https://fbcdn-sphotos-a-a.akamaihd.net/hphotos-ak-frc1/t1.0-9/10004017_10201574630789876_502687288_n.jpg
https://fbcdn-sphotos-g-a.akamaihd.net/hphotos-ak-ash3/t31/841333_10201574768193311_750739175_o.jpg

I thought i'd see what voltage the panel is outputting. I read this as 5.7v or perhaps I'm a wally!, am I reading this right?, I believe I am.

So, evidently, my panel is outputting nicely. Based on a figure of 200mA, surely I can assume now, that is is sufficient to charge the phone?, none of this really tallies with the figures CSD Tool is giving! (because of my lack of knowledge I suppose!) :D

By the way, how is your panel connected to your pogo pins?

[juiceme]

@Markkyboy, what are you using as a load?

You cannot just put your ammeter across the voltage source and get a meaningful reading, heh :D:D:D

[Markkyboy]

@Markkyboy, what are you using as a load?

You cannot just put your ammeter across the voltage source and get a meaningful reading, heh :D:D:D

Hi juiceme,
yep, as you have guessed, I don't really know what I'm doing! :D
I'm crap at all this, but keen to learn!
No load on the panel, just a 20 watt halogen lamp at 40mm distance.
I need to get it back on the phone and test again, but as I have clearly displayed, its all a lot of guess work! :)
Feel free to rib me all the way, I gotta learn somehow! ;)

[skvark]

Thanks skvark. I found my skytronic multimeter and put a fresh battery in!

Right new figures, which show much more promise! I set the multimeter to 200mA setting and showed the panel the light, the reading shot off the scale, leaving a number 1 on the left side of the meter screen, this represents that it went past the given rate!

I thought i'd see what voltage the panel is outputting. I read this as 5.7v or perhaps I'm a wally!, am I reading this right?, I believe I am.

So, evidently, my panel is outputting nicely. Based on a figure of 200mA, surely I can assume now, that is is sufficient to charge the phone?, none of this really tallies with the figures CSD Tool is giving! (because of my lack of knowledge I suppose!) :D

By the way, how is your panel connected to your pogo pins?

First, read this: http://learn.adafruit.com/multimeters/overview

It's a very nice tutorial covering a lot of stuff how to use a multimeter.

The voltage value seems to be right.

About current sensing.. you have to get the current to flow trough the multimeter. Basically the multimeter has to be connected between the phone and the panel.

I have just two wires soldered/glued to the backside of the panel. They are positioned so that they touch the pogo pins when TOH is in place. I'm going to change the wires to copper tape later because they are too thick and prevent TOH to lock fully from the other side.

[Markkyboy]

http://learn.adafruit.com/multimeters/overview"]http://learn.adafruit.com/multimeters/overview[/URL]

It's a very nice tutorial covering a lot of stuff how to use a multimeter.

Thanks for this, page saved/pdf downloaded. :)

[aironeous]

WYSIPS google it.

[psonek]

Yesterday i have assembled my solar other half using those flexible solar panels (link posted earlier in the thread).

All seems to be working except that the phone is not charging :mad:

I think the problem might be that the panels are not delivering enough energy and voltage drops too much so that the phone stops charging from the other half.

I used multimeter and on unattached TOH i can see 6V. But once i attach it drops to 0.6V

I also tried:

# cat /sys/devices/platform/msm_ssbi.0/pm8038-core/pm8xxx-adc/dcin
Result:37554 Raw:24767

Quite unexpectedly the "Result" value raises as in darkness...

Any other ideas what might be wrong?

[skvark]

Yesterday i have assembled my solar other half using those flexible solar panels (link posted earlier in the thread).

All seems to be working except that the phone is not charging :mad:

I think the problem might be that the panels are not delivering enough energy and voltage drops too much so that the phone stops charging from the other half.

I used multimeter and on unattached TOH i can see 6V. But once i attach it drops to 0.6V

I also tried:

# cat /sys/devices/platform/msm_ssbi.0/pm8038-core/pm8xxx-adc/dcin
Result:37554 Raw:24767

Quite unexpectedly the "Result" value raises as in darkness...

Any other ideas what might be wrong?

That 3,75 V is too low. Could you measure current with multimeter? Also double check connections.Those copper stripes seem to be very close to each other.

[psonek]

Hi,
oki, i will try to check current but need to lead the wires out of the case, so it might take a few days until i get to it...

The copper stripes are close, but as far as i could check they seem to be ok. There is 6V on them when measured with multimeter and when i placed a piece of plasticine the pogo pins made two holes on the right places...

[tvicol]

You need an energy harvester ic in order to make this work.

Have a look here https://www.youtube.com/watch?v=lrELRVFEx_s

[tvicol]

I would personally use LTC3105
http://www.linear.com/product/LTC3105

[psonek]

Hi,
it seems that my first version had swapped + and - although i have checked several times i did it wrong :confused:

Yesterday i redid the whole thing from scratch again. This time i used plasticine for precise copper stripes position and verified with 4.5V battery that charging works - and it worked.

After attaching solar panels it showed no activity under lamp, but later on direct sun the phone vibrated and displayed charging dialog + the white led was shining.

This is ok, but i havent seen that battery charge raised. In idle my jolla normally eats ~7mA, the panels should deliver 25mA so the charging current should be ~18mA. But i wonder if the phone is really idle while charging. And i wonder how to measure it...

I would personally use LTC3105
http://www.linear.com/product/LTC3105

Thanks for the link - interesting, but i wonder if it will fit somewhere under the cover...

[skvark]

This is ok, but i havent seen that battery charge raised. In idle my jolla normally eats ~7mA, the panels should deliver 25mA so the charging current should be ~18mA. But i wonder if the phone is really idle while charging. And i wonder how to measure it...

To get 1% more charge (21 mAh), you have to keep the panels in direct sunlight for over an hour.
18 mA is very very slow charging rate.

[pichlo]

To get 1% more charge (21 mA)

Did you mean 21 mAh :)?

18 mA is very very slow charging rate.

Indeed. Assuming 6V (being generous here, the real voltage must be much lower), it means only 150 mW (6V x 25 mA). That's about an order of magnitude less than what I would expect. Either the light level was very low or the panels are too small (covering less than 10% of the back space), particularly poor quality or damaged by the previous experiment with reversed polarity.

[skvark]

Did you mean 21 mAh :)?

Yeah :)

@psonek, could you look what the phone says about the voltage now that the polarity is right (cat /sys/devices/platform/msm_ssbi.0/pm8038-core/pm8xxx-adc/dcin)?

[rob_kouw]

I'm a bit afraid the whole thing will become hot. When I put my WakaWaka in the sun, it will be hot in an hour. Make the solar panel or the OH detachable so the phone can get some shade? :cool:

[dirkvl]

I'm a bit afraid the whole thing will become hot

ehhh

http://www.quickmeme.com/img/e5/e564cd644f440286750bca0b1c532e77c951346086c3aeb04f b11edf619d4b84.jpg

isn't this the case for every solar oh??

either
-accept it (join the sunny side)
or
-make a detachable third-half with solar panel and battery

[psonek]

Yeah :)

@psonek, could you look what the phone says about the voltage now that the polarity is right (cat /sys/devices/platform/msm_ssbi.0/pm8038-core/pm8xxx-adc/dcin)?

[root@jolla nemo]# cat /sys/devices/platform/msm_ssbi.0/pm8038-core/pm8xxx-adc/dcin
Result:4314054 Raw:32057
[root@jolla nemo]# cat /sys/devices/platform/msm_ssbi.0/pm8038-core/pm8xxx-adc/dcin
Result:4514790 Raw:32399

Mostly the values seem to be between 4.3V and 4.5V

[psonek]

Hi,
after a few days i still have not seen the battery being really charged with my solar toh.

I wonder where is problem. Battery status is "Charging", but current_now is never negative. Even more interestingly while charging it seems to consume more power.

In dark on wifi ssh i get values like this:

[nemo@jolla ~]$ cat /sys/class/power_supply/battery/current_now
2595

[nemo@jolla ~]$ cat /sys/class/power_supply/battery/current_now
8415

Under lamp while solar charging:

[nemo@jolla ~]$ cat /sys/class/power_supply/battery/current_now
10618

[nemo@jolla ~]$ cat /sys/class/power_supply/battery/current_now
11876

Maybe while charging the phone wont enter deep sleep?

[kimmoli]

Maybe while charging the phone wont enter deep sleep?

Yep, it stays active when charging.

[skvark]

You need better solar panel(s). For example my panel gives out 4x more juice per hour.

[psonek]

It seems that my theory is right verified by simple test. First stop upowerd so that it does not blink with leds, vibra and undims screen:

killall -s STOP upowerd

Then i have simple script that blinks with red led:

[root@jolla nemo]# cat ./blink.sh
#!/bin/bash
while true
do
sleep 1
echo 0 > "/sys/class/leds/led:rgb_red/brightness"
sleep 1
echo 50 > "/sys/class/leds/led:rgb_red/brightness"
done

Run this script and put the phone in standby.

Normaly the led stops blinking soon because CPU enters idle state. While powered by solar TOH it keeps blinking which indicates that CPU is not idle.

Does this make sense? I wonder what to do so that the processor goes idle. Otherwise solar charges will be quite ineffective because ~30mA is wasted on CPU.

[psonek]

Oki, some more debugging:

cat /proc/wakelocks | sort > w1.txt ; sleep 5; cat /proc/wakelocks | sort > w2.txt ; diff -u w1.txt w2.txt | grep '+\|-'

After this command if i move jolla under lamp, the CPU wakes up and shows this interesting lines:

-"power-supply" 2097 0 11 0 1156191680 1098783785 17446964 2266241071209
-"power-supply" 218 0 0 0 187567735 165166352 9797039 2266240949161
-"power-supply" 239 0 0 0 1541130070 1139839364 44895468 2266244946246
+"power-supply" 2106 0 11 0 1169010084 1111602189 17446964 2326885012348
+"power-supply" 222 0 0 0 194434802 172033419 9797039 2326884890266
+"power-supply" 243 0 0 0 1578425861 1177104634 44895468 2326898532864

If i understad it well then CPU wont go idle because of "power-supply" wakelock. This wakelock is defined in linux kernel in power_supply_core.c The question is now how to suppress it.

[psonek]

You need better solar panel(s). For example my panel gives out 4x more juice per hour.

Sure, i even ordered stronger panel, but it's also thicker and wont be as nice as these thin stripes.

I'd still consider this as bug unless there is good explanation why the CPU does not enter idle state (my previous phones didnt have problem with this). The charging rate will slow down because of this very much.

As far as i can see the only workaround could be done by recompiling kernel, but i am afraid that this is quite dangerous becase we dont have kernel sources, flasher etc..

[juiceme]

Sure, i even ordered stronger panel, but it's also thicker and wont be as nice as these thin stripes.

I'd still consider this as bug unless there is good explanation why the CPU does not enter idle state (my previous phones didnt have problem with this). The charging rate will slow down because of this very much.

Now this is guesswork, as I do not know the HW internals of the Jolla device so well, but if it follows the design of N9 at all, then the device has to be alive to monitor the battery charging process.

Some devices implement the charging all in the HW, for example the older symbian devices. On some devices the algorithms are run by SW in the main CPU.

[shadowjk]

When details on Jolla's TOH interface were announced, and it was speculated it could be used for solar charging and such, I always assumed one would have to implement it as solar panel + battery, with the panel slowly charging up its own battery, and then dumping that battery quickly to jolla, in order to minimize extra energy use for phone.

[kimmoli]

hmm... could a thin supercap utilised here? to allow fast discarge/charging cycles.
Annoying thing here will be the constant notifications charging/not charging, unless some tricks could be made. maybe need to do some lab-supply-experiments, iirc if voltage is lowered to certain level, it will say it is charging, but draws minimal current.
I imagine this is place for small micro to take care of the charge cycling, monitoring charge level of supercap and discharge to the phone.

first google result: http://www.cap-xx.com/

[psonek]

Now this is guesswork, as I do not know the HW internals of the Jolla device so well, but if it follows the design of N9 at all, then the device has to be alive to monitor the battery charging process.

Some devices implement the charging all in the HW, for example the older symbian devices. On some devices the algorithms are run by SW in the main CPU.

Yup, we'll have to figure out how it's done in Jolla. If it's done in SW it could be maybe improved by rtc wakeups?

[psonek]

hmm... could a thin supercap utilised here? to allow fast discarge/charging cycles.
Annoying thing here will be the constant notifications charging/not charging, unless some tricks could be made.

The trick is quite simple:

killall -s STOP upowerd

it's more hack then trick, but it works :) Maybe the charging could be controlled by simple circuit on i2c pins - you could disable the charging notification before you start charging from the supercap.

[psonek]

Hi,
yesterday i hacked jolla's kernel so that CPU goes to deep sleep even while charging and got first success. I can see the phone being charged with current ~3mA (see attached screenshot). I have no idea if that will be useful for anything though :D

Now i am waiting for sunny day and more powerful solar panel...

[petaqui]

Anything new?? =) This would be awesome, I need this for my looong trips haha

[psonek]

Anything new?? =) This would be awesome, I need this for my looong trips haha

I need this for my bike trip too, but the bigger solar panel havent arrived after month of waiting :(

For now i am experimenting with crank charger from evolveo, but IMO there is same problem as with solar charger - the CPU does not go in standby while USB is plugged so the charging current is rather low and i was unable to charge single % of battery until now...

We should either somehow fix Jolla or it would be needed to charge some external battery/capacitor first and use it to provide high current.

[petaqui]

How many hours can take to charge the battery of the Jolla using our first proto of the solar charger?
And maybe could be possible some hack to make the CPU stand by and charge faster...

[skvark]

How many hours can take to charge the battery of the Jolla using our first proto of the solar charger?
And maybe could be possible some hack to make the CPU stand by and charge faster...

With my proto it would take propably ~15 hours in the best-case scenario but in reality it will never happen.

[kjokinie]

Hi,
yesterday i hacked jolla's kernel so that CPU goes to deep sleep even while charging and got first success. I can see the phone being charged with current ~3mA (see attached screenshot). I have no idea if that will be useful for anything though :D

Now i am waiting for sunny day and more powerful solar panel...

You, my friend, have invented phone with infinite standby time :)

BTW, would you happen to have that kernel hack in a git tree somewhere?

[psonek]

You, my friend, have invented phone with infinite standby time :)

BTW, would you happen to have that kernel hack in a git tree somewhere?

I have it on my computer at home. I can dig it up and put online in the evening. But even with this hack it's not usable, see:

https://together.jolla.com/question/47255/50ma-lost-during-toh-charging/

[dirkvl]

bump

had some interesting results with a panel, but still a lot of room for improvement in standby current and deep sleep while charging.

i would love to experiment with that deep sleep kernel hack, please dig that up!

[psonek]

Ok, here it is:

http://activationrecord.net/radekp/pub/ignore_charging_wakelocks.patch

How to build? I have dumped jolla's boot partition, unpacked it, assembled with my custom kernel and flashed to recovery partition. It can be then booted using POWER+VOLUME buttons. Maybe i could rebuild and upload the partition img for latest jolla release.

How to test? Make simple script which blinks with led. When CPU is alive it blinks, when CPU goes to suspend it stops blinking.

On jolla stock kernel:

- blinks during charging -> CPU awake
- stops blinking when not charging -> CPU suspended

On patched kernel:

- always stops blinking always -> CPU suspended

[kloumpt]

Ok, here it is:

http://activationrecord.net/radekp/pub/ignore_charging_wakelocks.patch

How to build? I have dumped jolla's boot partition, unpacked it, assembled with my custom kernel and flashed to recovery partition. It can be then booted using POWER+VOLUME buttons. Maybe i could rebuild and upload the partition img for latest jolla release.

How to test? Make simple script which blinks with led. When CPU is alive it blinks, when CPU goes to suspend it stops blinking.

On jolla stock kernel:

- blinks during charging -> CPU awake
- stops blinking when not charging -> CPU suspended

On patched kernel:

- always stops blinking always -> CPU suspended

Are you planning to release flashable image for the latest update ? :)

[Kabouik]

I just received my Solartoh today (thanks Dirk!), but unfortunately I have the exact same problem I had with my Diptoh, and same problem Clovis had with Solartoh: no network at all when mounted. I had no issue at all the days before with my prototype Toholed (not the same plastic) and my official TOHs. I removed my Solartoh, and pouf, back to 4G network. :/

[n900user259]

I just received my Solartoh today (thanks Dirk!), but unfortunately I have the exact same problem I had with my Diptoh, and same problem Clovis had with Solartoh: no network at all when mounted. I had no issue at all the days before with my prototype Toholed (not the same plastic) and my official TOHs. I removed my Solartoh, and pouf, back to 4G network. :/

Shouldn't this message be posted on the SolarJolla (http://talk.maemo.org/showthread.php?t=91933&highlight=solartoh) thread?

[Kabouik]

Of course it should! Sorry about the wrong topic, I mixed both. I can delete these two messages if necessary.