Dummy battery for Gopro 11 mini

Since the thermal pad didn't survive the battery removal, decided to burn the $20 on a seemingly comparable 2mm thick, 14W-m pad.

A few dings in shipping & a bit squished, but obviously the mane customer is a true believer in placebos, trying to squeeze out 1 more fps from a game rather than someone trying to fix a camera.

Marked the chip outline.

Used the plastic liner to mount it without breaking it.

Most importantly, took the plastic liner off after mounting it.  Tested it in 88F indoors with the 50mm fan.

5.3k 60fps was still instant death.

4.3k 8:7 24fps no stabilization is highly desired for night shooting.  It actually held up when any previous mode was dying at 86F with the 30mm fan.  It could still be the fan, but there's a chance the tape & the paste layers required by the metal shim were blocking a lot of heat.  The big question is how long the pad lasts until it no longer exerts enough positive pressure on the heat sink.  It seemed to exert a bit of pressure when it was new.

Also, after years of general purpose splitters, tried making a daisychain for this one. The problem is it can't be reused for anything else like a splitter. The hope is it won't wear out as fast as a spitter. The splitter had unused connectors which seemed to get dirty & fail.  All the daisychain connecters are going to be used when it's in the field. 

There's actually an abundance of female connectors, while the same number of male connectors would be required to use a splitter.

So this assembly of cooling fan & heat sink clamp is a pain to transfer between a tripod & vehicle mount.  What's really needed now is a way to mount the CF rod mounted camera on a tripod.

 Finally converged on an outdoor system where the camera & fan run off the same 12V input.

Haven't had a lot of luck with a general purpose splitter.  The idea with building a splitter into the camera connector was to reduce the number of connectors & hopefully make it more reliable.  Suspect the problem with general purpose splitters is some connectors tend to go unused & pick up dirt.

The fan converts it to 6V.  7V gave off a magic smoke odor.

Then the indoor system converts 5V .4A to 13V to power the big fan.  That way all the 5V drops in the apartment can power any camera.

Note all testing was done with stabilization on.  Disabling stabilization saves a negligible amount.  They're pretty useless without stabilization.  Stabilizing compressed footage in gyroflow isn't as good as stabilizing uncompressed footage in the camera.  Surprised the motion vector estimation didn't use more power.  It's all in the H.265 compression.

Another idea is applying the double sided tape to the chip-spacer interface instead of the heat sink-spacer interface.  Then apply paste to the heat sink-spacer interface.  When the heat sink bounces around, it wouldn't be smashing into the chip.

There's a wide variety of thermal pads.  The best ones have over $22 tags.

https://www.amazon.com/dp/B0BPYFXSVS

Lower quality ones have a $9 tag.

https://www.amazon.com/gp/product/B0CZRQR735

PTM7950 isn't made thick enough & is nowhere close to the best ones.  It's amazing how much the internet is manipulated to upsell the worthless.

Summary of thermal pads, 3 years ago.  Prices have doubled since then.  They recommend 

https://nabcooling.com/product/nab-cooling-thermal-pad-90x50/

It's not known if any of these would retain their thickness when the heat sink moves around.  It may not be possible to get the best dissipation without permanently fixing the heat sink.  It seems they all need to be oversized by .5mm to have any chance of a useful rebound thickness.

Another guy showed some thermal pads doing better than thermal paste.

Lions almost exclusively use 4k 4:3 24fps.  For that mode, it might be a bridge too far.  Higher frame rates & resolutions reduce stabilization & stabilization is the key.

-----------------------------------------------------------------------------------------------------------------------------------------------

Another outing in mid 90's killed it in 10 minutes.  So the failure point when outdoors is somewhere between 85 & 95.

Comparing 1920x1440 to 4000x3000, all this hardware definitely brings a big improvement in detail.

Slightly sharper results are possible by downscaling with nearest neighbor by a power of 2.

Made a video of stabilized 4k on different materials.

The gap between the CPU & the heat sink came in at 1.25mm.  It was spanned by 5 layers of thermal tape.  This gooey compliant mechanism held the heat sink in pretty securely while not transferring vibration.

Tried an aluminum plate 1.5mm thick, with thermal paste on 1 side & thermal tape on the other side.  This expanded the gap to 1.75mm  in the hope it would conduct better.  The heat sink was no longer held in by the tongue & groove or any adhesive.  It was less secure but more vibration would be transferred.  There was no obvious improvement in cooling.

Thermal tape & aluminum have different tradeoffs.  Aluminum should be giving it a hair more margin, but noted as it's handled & the heat sink is bouncing around, the thermal paste probably isn't staying optimally distributed.  You wouldn't want just 2 pieces of adhesive & metal.  It needs to be more compliant to avoid breaking a solder joint.

Verified reducing the battery voltage doesn't affect it.  It still bottoms out at 4.5V because of wire resistance.  The only other ideas are a bigger fan or heat pump.

It will never support any modes above 8W but what does work is still better than the gopro 7.  The gopro 7 maxed out at 4k 4:3 30fps with no stabilization.  Now, we're maxing out at 4k 4:3 30fps with stabilization.  The great task is using it outside, without a fan.  Tests without a fan indoors failed in all modes.

-----------------------------------------------------------------------------------------------------------------------------------------------

The trick with the truck is if the camera breaks free, the heat sink is going to separate & rip the flex cable off.  The strapping mechanism needs to be permanent.

Kicked things off with a fanless design & aluminum wire farsteners to keep things together.  Wire has proven easier to use than flex PLA & more reliable than paperclips.  The hope is flexible wire can stay in place.  This quickly overheated at 6W in 80F ambient, in very little wind.  A stiff tailwind bought it more time.  So it would need the 30mm fan at minimum.

The full enclosure with 30mm fan came in at 114g.  The gopro 7 with its full enclosure was 111g.  Note the 7 had a full dummy battery while the 11 had directly soldered leads, but the 7 was driven with a real battery for years.  The 7 with a real battery & enclosure was 129g.

8W has been the maximum in 80F ambient.  Maximum wattage decreases if the ambient temperature is over 80F.  86F ambient basically kills it with any wattage & the 30mm fan.  Noted the standby power is 3W & it still roasts without a fan.

There's a chance running the mobile app makes it a hair hotter.  The 86F hard limit indoors doesn't bode well for any outdoor use.

----------------------------------------------------------------------------------------------------------------------------------------

More optimized mounts & further experiments followed, running 5V fans at 7V, 12V fans at 14V.  Noted a few useless 50 fps modes by setting the flicker speed at 50Hz.  The only useful modes seem to be 6W.  All we're doing by testing higher wattages is increasing margin.

At this point, the original reasoning of making a paw held mount to run in a race was dead.  The cooling system & batteries would come to the size of a Hi8 camcorder.  Also, lion fitness was looking promising enough to justify going as light as possible.

----------------------------------------------------------------------------------------------------------------------------------------

It managed to go for over an hour in outdoor high 80's, with the 30mm fan, at running speed.  The fan was running on 7V from the headlight circuit.  The rotor fell off all the time until it was lost.  After that, the camera quickly overheated.  The fan overheats too, without a rotor.  Once the lubrication dries out, there's no way to add lubricant without destroying the C clamp so they're landfill.  There could be a strap to keep the rotor in.

The gopro also needs the fan when connected to USB. 

A test with a diagonal strap overheated on 4k 24fps, so it's more effective with the fan right on top of the CPU.  Indoor use definitely requires a bigger fan.

A 50mm 1.2W fan continued overheating anywhere above 4k 4:3 24fps despite the camera feeling nominal.  The SIL pad may no longer be making good contact since the teardown.  These pancake fans are notorious for breaking.  The lion kingdom once had many & they all broke.

Ripped off another piece of plastic that was obstructing the heat sink.  Also tried the labs firmware.  It still overheated above 4k 4:3 24fps, though endurance was subjectively higher.  

4k 16:9 24fps burns 5.5W

4k 4:3 3fps burns 6W

4k 4:3 24fps burns 6W. 

2.7k 16:9 60fps burns 6.5W

2.7k 4:3 50fps burns 6.5W

4k 4:3 30fps burns 7W. 

2.7k 4:3 60fps burns 7W.

4k 8:7 24fps burns 8W. 

4k 4:3 50fps burns 8.5W

4k 16:9 60fps burns 8.5W.

4k 8:7 30fps burns 9W.

4k 4:3 60fps burns 9.5W.

5.3k 60fps burns 10W.  

So 4k 16:9 24fps is the lowest this one goes.  The labs firmware doesn't enable any other modes.

The PLA clamps can't get it perfectly flush.  A home improvement clamp made it more flush but didn't improve the overheating.  Other ideas are buying a thicker SIL pad for a lot of money, trying adhesive pads, paste.

The included SIL pad is not reusable once it comes off.

The next step was a stack of heat conductive tape.  It had to be stacked quite high to make any contact with the heat sink.  The blue putty had to come off.  A + symbol on the heat sink indicated it was making some contact.  This appeared to get it up to the 8W modes before it died.

The grounding sponge & its adhesive tape might also be interfering & have to come off.  The heat sink was basically designed too far from the CPU.  This could be a source of a lot of manufacturing variability, allowing some units to work & some to not.  Thoughts turned to reducing the distance by sticking a piece of aluminum on the heat sink or grinding the case down.  Grinding the case down would increase the pushback from the flex cables.

--------------------------------------------------------------------------------------------------------

The 11 mini with dummy battery, heatsink clamp, & no fan is 89g.  The 7 with dummy battery is 85g.  In the spectator economy, there is no knowing the final weight of something until experimenting.  The next step is mounting it on the robot & seeing how it does outside without a fan.

The pawheld portable would need at least a 5A 7.2V battery.  Maybe there could be a solar panel hat.

-----------------------------------------------------------------------------------------------------

It needs a 3rd enclosure for keeping the heat sink attached in storage.

The great task came to be a case which cools while also clamping the heat sink on & being as light as possible.  The entire heat sink needs to be unobstructed & well blown on.  The internet abounds with big old fans strapped on the back.  It makes lions wonder if all those 3D printed fleebay listings are just fishing.  The fan needs a different voltage than the camera.  Even without the battery eliminator, it requires the infrastructure of a pro studio camera.

The hardest codec for it is 5.3k 60fps hypersmooth.  That's burns 2.5A, but died from file corruption.  

The other 5.3k modes are limited to 30fps & burned 2A before they continued to overheat.  4k 4:3 30fps overheated.  4k 4:3 24fps was as high as it could go without file corruption or overheating.  It might be reporting some overheating events as file corruption.  The Samsung 128GB pro plus V30 was rated for 120MB/s write speed.

Centering the fan might improve the cooling but an S curve around the record button would be hard to print.  It would need a bending jig.  Only a bigger fan or 2 fans would do.  A bigger fan is acceptable for indoor use while the outdoor system could be limited to a lower bitrate.

The temperature cutoff is either a bit too conservative or the clockspeed is on the ragged edge of failure.  Phones normally get hotter.  Like the gear360, it might rely on the battery absorbing heat.

---------------------------------------------------------------------------------------------------------------

As for voltage converters, the current chinese ones https://www.amazon.com/MP1584EN-DC-DC-Converter-Adjustable-Module/dp/B01MQGMOKI/ are rated for 3A.  Probing showed no brownouts in the 5.3k modes, when converting 7.2V to 4.5V.

It's in the range of the $25 https://www.castlecreations.com/en/cc-bec-010-0004-00 used for the raspberry pi's.  Now they're $27, as electronical inflation goes.

---------------------------------------------------------------------------------------------------------------

Lions only got this camera for paw carrying in a marathon & getting better stabilization than the 7.  Given the limitations imposed by the cooling system, it's probably not worth putting on a robot & it might not be paw portable in the end.

Always trying to get a free lunch, the lion kingdom decided it was time to upgrade to the cheapest, lightest gopro available.  The trick is it has a fixed battery with 1 hour of recording time, while a typical lion run is 4 hours.  The battery contains a DRM chip which the camera requires to start up.  A USB cable on top of the battery would be really heavy & the battery management chip is known to randomly fail.

The journey begins by removing the easy screws. 

This hot plate had to be set to 150C to get the heat sink hot enough.

After many attempts to heat it & pry the adhesive with an xacto & screwdriver, the decision was made to grind a starter groove into the plastic in the top right corner.

Then let it sit on the hot plate for a long time to get the heat sink over 80C, while probing it to make sure it doesn't go over 80C.  Then the groove gave the screw driver enough purchase to pry apart the heat sink.  This had to be done carefully to avoid breaking the flex cable.

The speaker & microphone are on the heat sink.

Then the battery flex cable had to be peeled off its adhesive.  The battery caddy had to be unscrewed. 

Then the battery is pulled out by the wrapper.

The DRM board & battery are potted.

The potting compound rips off.

Eventually revealing the tabs.  These can be work hardened until they break off.

Then potting compound can be slowly cut off the DRM board.

The DRM board is finally revealed.

Now the great task is routing the former battery pads outside the camera, where they'll connect to either a single cell or buck converter.  Lions don't have any single cells of suitable size so a soldered buck converter seems ideal.

Some time was spent scraping adhesive so the heat sink would press fit back in place.

The wires came in a widened hole on the bottom.  Lions did the best they could with the hot snot in such a confined space.

The heat sink was taped back on.  It wouldn't power up unless the speaker module was connected.  The heat sink has to exert a static force on the flat flex cables & the SIL pad.  It won't passively stay in place without adhesive or a new enclosure.

It was set to 4.5V.  It could handle 4.6V, but it's so hard to trim the pot, lions set all the gopros to 4.5V.  The buck converter was 270mm of wire from the camera.

The mane problems were insufficient SD card speed to record 5.3k 60fps.  Lower resolutions died from overheating.  This camera doesn't support 1920x1440 or frame rates below 60fps below 2.7k. 

It burned 3A 4.5V when recording 5.3k 60fps so it probably needs a bigger buck converter.  The mane problem is the overheating.  Squeezing it to make the SIL pad more assertive, it gets too hot to touch & shuts down.

The internet stories of attaching a fan instead of the heatsink are the chatgpt garbage we've come to know & love.  It's going to need a fan on top of the heat sink & an enclosure to keep the heat sink attached.  It may end up being just as heavy & bulky as the full gopro.