Author Topic: EEVblog #761 - Pebble Time Smartwatch Teardown  (Read 16436 times)

0 Members and 1 Guest are viewing this topic.

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 38720
  • Country: au
    • EEVblog
EEVblog #761 - Pebble Time Smartwatch Teardown
« on: July 01, 2015, 01:00:27 pm »
What's inside the new Pebble Time Smartwatch?
Dave does a teardown after taking it through the torture test, and gives it a close up view with the Tagarno microscope.
How easy is it to disassemble and replace the battery?
How much do the components cost?
What is the manufacturing quality like?
How does the Bluetooth antenna work?

 

Offline ivan747

  • Super Contributor
  • ***
  • Posts: 2046
  • Country: us
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #1 on: July 01, 2015, 02:42:04 pm »
You asked for something to glue the LCD back on.

A year ago I bought a roll of this:
http://www.amazon.com/YUYIKES-Adhesive-Digitizer-Supersonic-Incredible/dp/B004WIQW3U/ref=sr_1_14?ie=UTF8&qid=1435761437&sr=8-14&keywords=touchscreen+tape&pebp=1435761786500&perid=1593MA2Z6DEQQTQPNVA8

That roll is very long and very cheap.



It's great double sided tape for flat surfaces. It sticks quite hard if you clean the surfaces flat with IPA. The tape is very thin and you can cut it with a sharp knife. The tape itself is transparent when it's stuck right. Just store it in a very dry place (like a shoe box with silica gel bags) because if you don't, it degrades (but then I live in the tropics, so you might get away without special care in your aircon'ed room).

The material is 3M 300LSE tape. You can find a supplier that sells it in patches and those patches can be machine cut to whatever shape you need (you would need an apropriate machine. I haven't researched if a laser cutter will cut it):
https://www.google.com.do/search?q=3m+300lse&es_sm=93&biw=1281&bih=755&source=lnms&tbm=isch&sa=X&ei=jvuTVfbWI8jPsAXzj4_wBw&sqi=2&ved=0CAYQ_AUoAQ

Datasheet:
http://multimedia.3m.com/mws/media/142939O/adhesive-transfer-tapes-with-adhesive-300lse.pdf

P.S. large patches are available in Amazon, it seems:
http://www.amazon.com/super-strong-double-sided-discounts-digitizers-3M9474-all/dp/B00IQWOA50
« Last Edit: July 01, 2015, 05:32:27 pm by ivan747 »
 

Offline Laertes

  • Regular Contributor
  • *
  • Posts: 58
  • Country: de
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #2 on: July 01, 2015, 05:13:42 pm »
The processor actually has an internal 32khz RC oscillator... probably not exactly accurate, but when it's connected to a smartphone, it's got access to an accurate reference time anyway...
 

Offline andyturk

  • Frequent Contributor
  • **
  • Posts: 895
  • Country: us
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #3 on: July 01, 2015, 05:50:59 pm »
The bluetooth chip (CC2564) requires an external 32768Hz clock, so there's got to be one on there somewhere. The STM32 does have an internal RC oscillator, but it's too inaccurate for either timekeeping or bluetooth.

The 26MHz source is pretty obvious. Maybe the black thing next to it is the 32k? My guess is that the design uses oscillators modules (instead of crystals) so the outputs can be shared between the STM32 and the CC2564.
 

Offline eV1Te

  • Regular Contributor
  • *
  • Posts: 186
  • Country: se
  • Your trusted friend in science!
    • richardandersson.net
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #4 on: July 01, 2015, 06:44:09 pm »
I'm guessing the extra passives (resistor and capacitor) close to the switches are for debouncing. For this kind of very low power application you do not want extra interrupts or software wait timers to check for bounces. You want the micro to sleep all the time and only wake when it receives a real interrupt from one of the switches.
 

Offline Yansi

  • Super Contributor
  • ***
  • Posts: 3898
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #5 on: July 01, 2015, 08:34:21 pm »
Nice teardown, Dave!

I think the small FPGA is probably for driving the display-specific interface. It would consume less power thatway, than bitbanging the interface with the MCU.

andyturk: The STM32 has an RTC_OUT to provide 1Hz, 512Hz or 32kHz (through MCO1) buffered output from the crystal. The internal lowspeed oscillator (LSI: 17-47kHz, typ 32kHz  RC oscillator) is not anywhere near useful for maintaining accurate time. The LSI is used mostly for WDG timers and/or AWU (autowakeup) unit.
 

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17236
  • Country: 00
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #6 on: July 01, 2015, 08:50:45 pm »
I'm guessing the extra passives (resistor and capacitor) close to the switches are for debouncing. For this kind of very low power application you do not want extra interrupts or software wait timers to check for bounces.
Hardware debouncing? Never!    :palm:

 

Offline Goober_in_CA

  • Newbie
  • Posts: 6
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #7 on: July 01, 2015, 10:51:16 pm »
Hi Dave
Thanks for the great teardown of the Pebble Watch!
You mentioned the name of the microscope you were using during the video, but I was not able to locate this manufacture. Could you please tell us the model number and manufacture of the microscope you used during this review?
Thanks
Daryl
PS: if you are ever here in Los Angeles area and have time for a tour at SpaceX, please let me know.
 

Offline adcurtin

  • Contributor
  • Posts: 29
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #8 on: July 02, 2015, 03:11:58 am »
Dave, the Pebbles are all designed for continuous submersion (though not with button pressing while submerged, at least not on the old Pebbles). http://help.getpebble.com/customer/portal/articles/1850071-water-resistance

I'm disappointed that you talked about the water resist so much when you've got it misunderstood. At least you got the mic and the LCD correct this time around.  :)

The FPGA is the display controller (nice job). See this comment from one of the developers on the reddit post about ifixit.

Quote
Display controller. You can get an idea of its function from the emulator.

In the ifixit teardown, they have a list of all the major parts in step 9 and step 10. The ifixit teardown confirms a lot of what you said.
« Last Edit: July 02, 2015, 03:29:29 am by adcurtin »
 

Offline Barny

  • Frequent Contributor
  • **
  • Posts: 311
  • Country: at
  • I'm from Austria, not Australia ;)
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #9 on: July 02, 2015, 05:41:09 am »
About the 30m at Watches.
Din8310: Water resistant for 30minutes at 1meter water depth  :-DD
(Only for factory new watches. Older watches could leek earlier.)
 

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 38720
  • Country: au
    • EEVblog
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #10 on: July 02, 2015, 06:02:06 am »
You mentioned the name of the microscope you were using during the video, but I was not able to locate this manufacture. Could you please tell us the model number and manufacture of the microscope you used during this review?

Tagarno
http://www.tagarno.com/products/magnus-fhd-zip-0
 

Offline bitwelder

  • Frequent Contributor
  • **
  • Posts: 973
  • Country: fi
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #11 on: July 02, 2015, 06:10:45 am »
The processor actually has an internal 32khz RC oscillator... probably not exactly accurate, but when it's connected to a smartphone, it's got access to an accurate reference time anyway...
How much accurate could it be if you use it for a long time as a stupidwatch(TM) away from the paired companion phone?
 

Offline adcurtin

  • Contributor
  • Posts: 29
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #12 on: July 02, 2015, 06:27:17 am »
About the 30m at Watches.
Din8310: Water resistant for 30minutes at 1meter water depth  :-DD
(Only for factory new watches. Older watches could leek earlier.)

It's not Din8310.

According to a kickstarter update, it's ISO 22810. from here:

Quote
For example, if a watch is said to be “water resistant” to 30 meters (100 feet), this means that for all aquatic activities down to a depth of 30 meters, the watch case should not leak. The manufacturer of a water-resistant watch may offer ratings to help consumers determine the sort of environment the watch can handle.

The kickstarter update also talks about swimming, and says you can do anything up to 30m (it'd be 50m if it weren't for the microphone) in the water, but no diving.
 

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 38720
  • Country: au
    • EEVblog
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #13 on: July 02, 2015, 06:58:32 am »
Dave, the Pebbles are all designed for continuous submersion (though not with button pressing while submerged, at least not on the old Pebbles). http://help.getpebble.com/customer/portal/articles/1850071-water-resistance
I'm disappointed that you talked about the water resist so much when you've got it misunderstood. At least you got the mic and the LCD correct this time around.  :)

I do know what I'm talking about, I simply hadn't read all the detail, which apparently is only available in a Reddit feed.
That page you pointed to says nothing about any standard, nor does their website it seems.
It seems Pebble are using the new ISO2281:2010 standard, not the usual (older) ISO2281 standard.
Basically in the watch industry when you advertise something as having "30m water resistance" (you should use the term "proof") it means exactly as I stated based on the ISO2281 standard practically the entire industry uses. It's not hard to misunderstand huge info graphics on their main page screaming "30m water resistance", and then assume that means the same thing as when practically everyone else in the industry uses the same term.
And based on my 10+ years experience designing underwater marine electronics I reckon their claim of 30m continuous immersion rating is suspect.
« Last Edit: July 02, 2015, 07:40:54 am by EEVblog »
 

Offline adcurtin

  • Contributor
  • Posts: 29
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #14 on: July 02, 2015, 07:20:55 am »
Check out this article, ISO22810 has been the standard since 2010 (and water-resistant is the term it uses. Waterproof is actually frowned upon by both the ISO ( says cnet, couldn't find a better source :( ) and FTC (pdf))

Unfortunately the Pebble website is lacking info and specs on the time, which is disappointing (the pebble time isn't available for sale yet though, just preorder. they should still have info up.). It is hard to find a good definitive source (kickstarter update #12). They do briefly talk about their testing in that update, but not really any solid details.
« Last Edit: July 02, 2015, 07:23:33 am by adcurtin »
 

Offline tombi

  • Regular Contributor
  • *
  • Posts: 163
  • Country: au
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #15 on: July 02, 2015, 07:26:10 am »
Before you glue it back together it might be cool to see what the power consumption is under different circumstances.

Could break out a uCurrent and a scope?

Tom
 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8420
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #16 on: July 03, 2015, 01:35:41 am »
The 14960B is a MAX14690BEWX+T which is a "wearable charge management solution". Maxim themselves have very little information about the MAX14690 except a mention here, but you can find a bit more info in Chinese:

http://solution.eccn.com/uploads/solution/201501/20150104094404713.pdf
http://design.eccn.com/design_2015060214124260.htm
http://design.eccn.com/uploads/article/201506/20150602141117277.gif

Relevant translated quotes from the above
Quote
For wearable device power design challenges and difficulties, reducing the burden of design engineers design, Maxim has developed a variety of wearable devices, integrated power chip, which MAX14690 / MAX14676 power-management chip is one of two typical products, features 9 and Figure 10

Figure 9 MAX14690 ultra-low quiescent current of integrated power management chip

There is a bit more info on the related MAX14676 but the only datasheet I can find has most of its pages missing. :--
 

Offline janekm

  • Supporter
  • ****
  • Posts: 515
  • Country: gb
Re: EEVblog #761 - Pebble Time Smartwatch Teardown
« Reply #17 on: July 03, 2015, 06:12:58 am »
The 14960B is a MAX14690BEWX+T which is a "wearable charge management solution". Maxim themselves have very little information about the MAX14690 except a mention here, but you can find a bit more info in Chinese:

http://solution.eccn.com/uploads/solution/201501/20150104094404713.pdf
http://design.eccn.com/design_2015060214124260.htm
http://design.eccn.com/uploads/article/201506/20150602141117277.gif

Relevant translated quotes from the above
Quote
For wearable device power design challenges and difficulties, reducing the burden of design engineers design, Maxim has developed a variety of wearable devices, integrated power chip, which MAX14690 / MAX14676 power-management chip is one of two typical products, features 9 and Figure 10

Figure 9 MAX14690 ultra-low quiescent current of integrated power management chip

There is a bit more info on the related MAX14676 but the only datasheet I can find has most of its pages missing. :--

Looks like the full datasheet requires an NDA and a statement regarding expected volumes, as they reckon you'd need their help to use the part... A bit obnoxious for my liking. I suppose that means the part won't be available through regular distributors anyway...

It's a bit of a shame as it'd be a useful part to me, but those kind of conditions make me vary about availability.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf