Yet another cheap thermal imager incoming.. Seek Thermal

[mikeselectricstuff]

Since these data are presumably unprocessed raw data, why does the image change significantly?

It may be wrong assumption while there is quite powerfull IC MCU ARM ROMLESS 100TFBGA digikey: NXP LPC4320FET100 

Latest datasheet linked from Digikey: LPC18X0,LPC43X0 IRC Update 22/Aug/2014
Update: Above some production notice of IRC details changed.
Its NXP page there:
http://www.nxp.com/products/microcontrollers/cortex_m4/lpc4300/LPC4320FET100.html
Its datasheet: http://www.nxp.com/documents/data_sheet/LPC4350_30_20_10.pdf

It would be very interesting to.... disassemble its ARM program code, but where it is -on external flash IC? 

BTW: It costs on Digikey <$10 and arount $4 for bigger volume. Quite nice MCU-maybe it is time to do some ARM development 

Code is stored in an external SPI flash, so should be pretty accessible, probably via the test pads.  ISTR someone said there was a firmware image file in the .APK package.
Souldn't be too hard to write oyur own code to run on it - Once I have all the units I have on order I may sacrifice one to trace out a schematic.

[miguelvp]

I got it today, doesn't work on my 1st generation NVidia Shield but it works on the wife's phone.

Picture of one of my dogs.



Question: How do you open that protective case? its like a jigsaw puzzle to me.

Edit: got it, was just hard to release the latch

[sacherjj]

Got mine in today.  Picture is a little noisier than I hoped, but about what I expected.  Tried to measure a 650F soldering iron.  Set the emissivity to stainless steel.  Highest I saw was 350F.  Though the spec went to 330 C ~ 626 F.  Figured this would max it.  Hmm..

[sgstair]

You can turn off the watermark in the app options.

[firehopper]

on the one image of the internals, it looks like the pins for that coil are not soldered to the board. am I the only one to notice that?

[bktemp]

on the one image of the internals, it looks like the pins for that coil are not soldered to the board. am I the only one to notice that?

I noticed that too, it makes sense: This board is build for lowest price. The coil seems to be pressed into the board. This can be done with a machine, soldering coil contacts is often done by hand.
If done properly, pressfit contacts aremore robust than soldering. This technique is widely used in automtive electronics because it is much more tolerant to vibrations than solder connections.

[marshallh]

on the one image of the internals, it looks like the pins for that coil are not soldered to the board. am I the only one to notice that?

I noticed that too, it makes sense: This board is build for lowest price. The coil seems to be pressed into the board. This can be done with a machine, soldering coil contacts is often done by hand.
If done properly, pressfit contacts aremore robust than soldering. This technique is widely used in automtive electronics because it is much more tolerant to vibrations than solder connections.

It could also be pin in hole reflow (on top only).

[sacherjj]

Watermarking the images like that is a very obnoxious move from Seek Thermal.

The app has options to turn this off and to turn on both time/date and location watermarking.

[mikeselectricstuff]

on the one image of the internals, it looks like the pins for that coil are not soldered to the board. am I the only one to notice that?

I noticed that too, it makes sense: This board is build for lowest price. The coil seems to be pressed into the board. This can be done with a machine, soldering coil contacts is often done by hand.
If done properly, pressfit contacts aremore robust than soldering. This technique is widely used in automtive electronics because it is much more tolerant to vibrations than solder connections.

It could also be pin in hole reflow (on top only).

No, it's definitely pressfit (I've had it out) - the pin is split in the middle to provide tension

[marshallh]

[eneuro]

Picture of one of my dogs.
...
img_thermal1915628406.jpg (35.53 kB, 624x832 - viewed 193 times.)

Why this image has bigger resolution than Seek Thermal sensor capability?
It is resized to the size of phone screen size, so no way to get oryginal not resized thermal image on native sensor resolution? 

[miguelvp]

it's exactly 4 times wider and 4 times taller than the output from the sensor.

If it was only 156x208 it would be too tiny to display, also I'm not sure what are they doing with the extra 2 rows that don't give you actual sensor output.

[marshallh]

process/batch variation? or die revs?

[miguelvp]

Back home I tried the shield gen 1 (Tegra 4) again, video preview doesn't work but that is common since Nvidia doesn't support screen recorders and access to the buffer directly it seems.

But I took a snapshot and it works. But, the camera is facing the clamshell so I need an extension cord, some phones that support the camera have the usb mounted in such a way that will face you, so you need a 180 adaptor or an micro usb extension cord.

But since I have a nook color laying around I'm going to install CM 11 (kit kat 4.4) and give that a try.

[ebeall]

Pretty quick for a preview of the tear-down to come from mikeselectricstuff! I got mine too (I'm pretty sure that everybody selling cores or making imagers got one), there's certainly more image noise than advertised, but then again, nobody else is selling at $200.

The nonuniformity onset is pretty fast, requiring a frequent shutter. Does anyone here know how long those types of actuators tend to last? Best case, worst case? Ideally, I'd like to know how many cycles the loop coil actuator is rated at, so I know how heavily it can be used (may have to ask Seek for the spec).

[eneuro]

process/batch variation? or die revs?

Did you tried to make more thermal image shots at different ambient temperatures just to see if these hexagonal sensor pixels patterns maybe changes somehow and it is temperature dependent or is consistent and unique to given Seek Thermal sensor-something like its footprint?

If it was only 156x208 it would be too tiny to display....

Operating system software could simply resize 208x156 it to fit into display, but we still could  be able to save oryginal one.

There must be something else behind this and it is interesting that this is 4 times bigger...
BTW: We are talking about iPhone version image only or Android versions has the same 4 times bigger thermal output images available to save?

[callipso]

@marshallh, @sgstair,

what are your order numbers?

[callipso]

Oh and hey Erik,

welcome to the forum!

[sgstair]

@callipso - It's between 350 and 375

[mikeselectricstuff]

Teardown delayed due to having to rent a different phone from Ebay to get it running.. hopefully this weekend.

[miguelvp]

Got 4.3 installed on Nook Color but no OTG support on the kernel
Then got 4.3 installed on a MyTouch 4G and again no OTG support

CyanogenMod drops support , last version for the Nook Color with OTG capabilities was Jellybean but not 4.3.

I guess I'm going to have to resort to Ebay after all, or tell my wife she can get a new phone so I can use her Galaxy S4.

I can use the Shield (Gen1) but no live update and for that I have to order a male to female OTG data cable extension, otherwise half of the image lands on the clamshell, unless I put a wireless controller and hook it via HDMI in console mode, but that's crazy!

[mikeselectricstuff]

First phone I tried was Galaxy S3 mini with Cyanogenmod, just received a Moto G which works

Bit busy building a 6,000 LED display at the moment so may take a couple of days before serious analysis....

[Fraser]

@Mike,

I own a Moto G so I am glad to hear that it works OK with the SEEK. 

Aurora

[Fraser]

When my SEEK arrives I will create some comparison images using the same thermal target and a selection of my thermal cameras. I have 160x120 and 320x240 resolution FLIR, NEC & TESTO cameras. It will be interesting to see how the images compare to the SEEK in terms of detail and noise. As previously stated, we are working with a new Generation 1 budget imager FPA and likely little better than BETA software for processing. Hopefully performance will keep improving. $200 is still great value for such technology 

Aurora

[bktemp]

on the one image of the internals, it looks like the pins for that coil are not soldered to the board. am I the only one to notice that?

I noticed that too, it makes sense: This board is build for lowest price. The coil seems to be pressed into the board. This can be done with a machine, soldering coil contacts is often done by hand.
If done properly, pressfit contacts aremore robust than soldering. This technique is widely used in automtive electronics because it is much more tolerant to vibrations than solder connections.

After thinking again, the press fit mounting has probably another reason:
In their bonding video you can see, all other parts are beeing already soldered when the sensor is beeing bonded. All other bonding videos I have seen, did bonding before soldering all other parts. Probably the sensor can not survive the high soldering temperature. Therefore wavesoldering the shutter coil connector is not possible. The USB cable is the only part soldered with the sensor mounted on the board, but this should not be a problem, because it is at the other end of the board and uses a different soldering technique.