Does Coldness Really Fix Electronics or is There Another Trick?

[snoopies622]

TL;DR Summary

leaving an old wifi device in my fridge a few hours fixed it.

This is more of a consumer question from someone who knows nothing about electronics: Last night my old Amazon Fire TV device apparently died, but after leaving it in the fridge all day it's now working again. What happened?

 

[Baluncore]

Semiconductor leakage currents increase with age and with temperature. Cooling the device may reduce leakage currents.

Warning: If you put electronics into a refrigerator, condensation of water onto the circuit board may cause problems due to leakage. Breakdown of high voltage supplies is likely.

 

[snoopies622]

Thanks Baluncore, will now read a bit about semiconductors.

Warning: . . Breakdown of high voltage supplies is likely.

Do you mean within the device or outside of it, in my house? I don't mind risking the device since it's obviously on its last legs anyway.

 

[Baluncore]

Do you mean within the device or outside of it, in my house?

Internal to the device.

If the device is poorly vented and contains warm humid air, then refrigerating it will condense some of that water internally as it cools.

If it is open and well vented then when the cold device is returned to a warmer high humidity environment it may condense water on internal cold surfaces.

Is it the heat that stops it working? Are you in an unusually hot place?
Why did you think of cooling it?

 

[snoopies622]

The device isn't in an especially hot place, but strangely it seems to have no ventilation holes at all. In the past when it's stopped working I've unplugged it for a few days, then it worked fine after plugging it back in. That didn't work this time but, other than cooling off, I couldn't think of anything that would be happening during those unplugged days, so I tried the refrigerator. I wrapped it in a paper napkin and plastic before putting it in the fridge to keep out moisture, but yes I imagine there's internal moisture trapped inside it, too.

I ordered a new one because I expect that this will only be a temporary fix. I didn't realize that electronics "wear out" like mechanical parts do.

 

[snoopies622]

I think they're no longer made, instead Amazon has the "stick" which hopefully works as well.

 

[Jody]

I would not stick that into the television after putting it in the fridge.

I think viewing the stick on its last legs and taking the risks is okay, but sticking it into a known good television... you're risking the television too. Best case scenario if it damages the television is only the port is broken, but I've seen during my years a technician... broken ports damaging the entire device (this I did not have an engineering background at the time to understand it, but I am thinking in hindsight a short related to the power could be the cause). My guess is your Fire Stick is worth $40-50 USD ish... how about your television?

I'm not going to suggest that is happens every time, but it's unnecessarily higher risk than usual.

I think the damage is done and it is too late for your Fire Stick (due to the possible humidity that Baluncore is talking about), but maybe the next electronic consider using a fan instead. I've had laptops that get super hot and would not turn on anymore... lift the laptop with the screen against the table and a regular fan blowing against it's bottom for a few minutes (or more convenient since it's available for a laptop just a laptop cooling pad)... it might turn on and boot into its operating system long enough for me to gather whatever the customer was trying to recover. I believe the same concept would work for other electronics.

 

[snoopies622]

Thanks, Joshy. I wonder if blowing a fan on the ports (there are two or three on the back) could possibly blow out the moisture.

 

[Baluncore]

Moisture is usually condensed onto cold surfaces, or dissolved in the air as humidity. You must lower the relative humidity by raising the temperature of colder drier air. I use a warm hair drier to dry equipment.

The most significant danger of refrigeration comes from insulation breakdown within switching power supplies, where several hundred volts DC can damage the PCB insulation before the internal fuse blows. I see in the photo the device has an external switching supply. Keep that away from the refrigerator.

Refrigeration is a quite unnecessary risk when air cooling can achieve the same result.

 

[snoopies622]

At this stage it might be impossible to keep the semiconductors both cool enough to minimize current leakage while also warm enough to prevent damaging condensation. I could keep a fan blowing on the unit while I use it, but that would create unwanted noise. Guess I'll just replace it now that the new one has arrived in the mail. Thanks, all!

 

[Tom.G]

Once you get the new one working, open the failed one and look around. It will help satisfy your curiosity if nothing else, and you may learn something too!

For instance, often the failure is one or more electrolytic capacitors gone bad. Those are the small cylindrical things, with a usually black plastic sleeve on them. Sometimes they show as a bulging body or white stuff leaking out of them, often nothing visible though.

Occassionally the problem is a bad solder joint. The tiny expansion with temperature may disconnect a component. Cooling causes stuff to shrink and can reconnect the component... until it gets hot again. These are quite hard to spot.

Have Fun!

Cheers,
Tom

 

[darth boozer]

Back in the days of replaceable discrete components, an aerosol can of freezer spray was a common part of the electronics repair toolkit. It was particularly useful for heat-related faults because individual components could be suddenly cooled, which often restored normal operation, to discover the cause of the fault. Poor solder joints and failing transistors were most common. Freezer spray is still available.

 

[Baluncore]

Freezer spray had a couple of problems. Firstly it contributed to the destruction of the ozone layer. Secondly, the cooling was so rapid that some expensive chips did not survive the rapid differential contraction, and so never worked again.

 

[darth boozer]

Freezer spray had a couple of problems. Firstly it contributed to the destruction of the ozone layer. Secondly, the cooling was so rapid that some expensive chips did not survive the rapid differential contraction, and so never worked again.

Freezer spray which used damaging chlorofluorocarbons was discontinued from the mid 1970s. The era that I posted about was discrete components, i.e. transistors, diodes, resistors . capacitors and inductors, NOT integrated circuits. The post was about the original question, "Coldness fixes electronics?", and specifically mentioned discrete components. Modern circuits are so small and integrated that they have become disposable instead of repairable. Cold temperatures can cause some equipment to function after failing under hotter conditions, but in almost every modern case replacement, not repair, is the only option.

 

[Baluncore]

Freezer spray is still available.

It was that last line that worried me, and now you are talking about the B&W TV era.

Real freezer spray is available only for applications where it is essential and where there is no environmentally safe substitute.

A better solution today is a supply of compressed air and a https://www.ebay.com.au/itm/VORTEX-COLD-AIR-GUN-DRY-COOLING-AIRFLOW-MACHINE-WITH-FLEXIBLE-TUBE-HEAT-GUNS/333113642280.

 

[snoopies622]

Follow up: Perhaps no one knows the answer to this but any input would be appreciated. I just hooked up the new Amazon Fire device, and I intend to use it 2-4 hours a day, like the old one. Will it last longer if I keep it unplugged when I'm not using it (to keep it cooler) or will the act of unplugging it and plugging it back in every day somehow wear it out faster? I kept the old one plugged in most of the time and it lasted . . I don't even remember, at least five years.

 

[Tom.G]

any input would be appreciated.

The IC life is reduced by temperature swings (think power on-off), and to a lesser extent by elevated temperature.

Electrolytic (filter) capacitor life is reduced by time-at-elevated temperature, especially if power is on during the elevated temperature.

In my experience, it is usually the Electrolytic caps in the power supply that fail first. That's because they operate at a higher ambient temperature.
(Try a different power supply for your old unit... that may be all that's needed.)

Then there is the mechanical cycling of the power connector. The wall outlet is probably good for several thousand cycles. The full size USB connector is rated for 1500 cycles, the "Mini B" is rated for 5000 cycles and the "Micro" is rated for 10 000 cycles. (I think the Fire Stick uses the Micro, not sure though.)

Four years of one plug/unplug cycle per day is 1461 cycles. All that manipulation will likely cause the wire to break before the connector wears out.

Actually, four years continuous for a low cost consumer electronic item isn't all that bad; it's 35000 hours.

Personally, for anything that costs less than about $100 and is used most days is just left plugged in. (I am a little biased though, many things I repair when needed)

Anyhow, that's my 2-cents worth.

Have Fun!

Cheers,
Tom

 

[Jody]

I would just leave it plugged in.

Even though it might be better in terms of electronics (it's always on a little bit) the connector and its pins are more likely to wear out more quickly.

 

[snoopies622]

. . the connector and its pins are more likely to wear out more quickly.

Well, there is the possibility of plugging/unplugging it indirectly with an extension cord, thus protecting the connector and pins, but I'm getting the impression that it's six in one and half a dozen in the other.

 

[sysprog]

I let monitors go to sleep to prevent burn-in, and outside of that concern, if the machine is not put in a 'presently on vacation' condition, I leave it on like an incandescent lamp.

 

[darth boozer]

It was that last line that worried me, and now you are talking about the B&W TV era.

Real freezer spray is available only for applications where it is essential and where there is no environmentally safe substitute.

A better solution today is a supply of compressed air and a https://www.ebay.com.au/itm/VORTEX-COLD-AIR-GUN-DRY-COOLING-AIRFLOW-MACHINE-WITH-FLEXIBLE-TUBE-HEAT-GUNS/333113642280.

I did not suddenly change to "the B&W TV era". Discrete components were extensively used in the 1970s and 1980s. Freezer spray was effective in finding the faulty IF transistors in some Sony Trinitron colour TVs as well as dry joints in some Philips colour TVs, along with numerous other heat-related faults I had to locate and repair.

Extract from the Jaycar Electronics website (https://www.jaycar.co.nz/freezing-spray-can/p/NA1000 ), "Non CFC Ozone safe propellant. Instantly freezing spray for rapidly cooling components to detectintermittent thermal faults, dry joints and overheating problems. Non-flammable. 250g size ". Why are you worried?

 

[Baluncore]

Why are you worried?

Because venting 1,1,1,2-Tetrafluoroethane to the environment is quite unnecessary.

R-134a may not destroy the ozone layer as quickly as R-12, but it is hard to contain, long lasting, and significantly increases global warming. It has been banned in the EU and now in some North American states.

Physical vibration, or a vortex tube cooler, is sufficient.

 

[Tom.G]

a vortex tube cooler

Were they readily available 40 years ago?

Currently: $50 + compressor + hose + hearing protection = $$$
For a business, could be (and a good idea); for personal use, not likely.

 

[Baluncore]

Were they readily available 40 years ago?

Yes, but they were too expensive and CFCs were too cheap.
The history goes back half over 70 years.
https://en.wikipedia.org/wiki/Vortex_tube#History

Liquid coolant cannot be used with ceramic inserts and some carbide cutting tools because they shatter with high differential temperature fluctuations. Air cooling can be used, so the vortex tube found an application. The price dropped from $750 to $50 a few years ago once China started using them in manufacturing industry and production increased significantly.

A small portable compressor now costs less than $100 and will do more than run a vortex tube.

 

[trurle]

Summary:: leaving an old wifi device in my fridge a few hours fixed it.

This is more of a consumer question from someone who knows nothing about electronics: Last night my old Amazon Fire TV device apparently died, but after leaving it in the fridge all day it's now working again. What happened?

Kindle Fire and subsequent Amazon Fire series have a "poorly designed" battery charging circuit which is likely not a "bug", but an intended feature to make device non-functional after certain degree of battery wear. Refrigerating allows to temporarily increase battery load capability, but effect is likely to vanish after ~10 more charging cycles. Proper fix actually would be battery replacement.

 

[sysprog]

Kindle Fire and subsequent Amazon Fire series have a "poorly designed" battery charging circuit which is likely not a "bug", but an intended feature to make device non-functional after certain degree of battery wear. Refrigerating allows to temporarily increase battery load capability, but effect is likely to vanish after ~10 more charging cycles. Proper fix actually would be battery replacement.

(emphasis (by underlining) added)

Can you present any evidence for this very serious accusation? I have suspected that kind of thing before of some manufacturers, but without even so little as reasonable suspicion level evidence or some other kind of prima facie evidence, I would not publicly call it "likely". Why do you think it's ok to make such a claim?

 

[Tom.G]

Why do you think it's ok to make such a claim?

Perhaps because other manufacturers have admitted using similar approaches. For instance Apple throttling the CPU on older iphones with an iOS update, supposedly around the time a new model came out. Of course a different reason was given.
https://www.vox.com/2017/12/22/16807056/apple-slow-iphone-batteries

 

[sysprog]

Perhaps because other manufacturers have admitted using similar approaches. For instance Apple throttling the CPU on older iphones with an iOS update, supposedly around the time a new model came out. Of course a different reason was given.
https://www.vox.com/2017/12/22/16807056/apple-slow-iphone-batteries

My LG V20 phone has a replaceable battery. I bought it (cheaply on the net) largely because of that. I found that I had to sharpen my 'jeweler skills' to replace the battery in my prior Samsung Galaxy S6 phone and my prior HTC One M9 phone. I strongly suspected in both cases that they had made battery replacement difficult in order to incent the customer to upgrade. I had no proof of either manufacturer having had or acted on such a motivation, so I consequently didn't publish the suspicion, and certainly not as if it were established fact.

 

[trurle]

(emphasis (by underlining) added)

Can you present any evidence for this very serious accusation? I have suspected that kind of thing before of some manufacturers, but without even so little as reasonable suspicion level evidence or some other kind of prima facie evidence, I would not publicly call it "likely". Why do you think it's ok to make such a claim?

Well, some of my customers (among other task i repair electronics) complained on "Amazon/Kindle Fire become impossible to charge". Therefore, out of curiosity, about 2 months before i took apart case of such malfunctioning Amazon Fire HD and soldered lab voltage source with current monitor to PCB at charging input, and voltage monitor to battery.
Initially i suspected USB micro-B connector may be faulty, but it turned to be more complicated.
Following picture was observed:
1) The current fluctuate between 0.4 A and 1.2 A if battery poorly charged (2.8-3.2 V open-circuit)
2) As soon as battery charges to 3.2V (~10% on indicator), the current become stable around 1.2-2 A.
3) If somehow the battery voltage drops below 2.8V, the controller refuse to charge battery at all. It may recover temporarily and allow charging if you refrigerate battery to increase open-circuit voltage.
4) Charging controller do not reject voltage over 5V (strong dependence of current on voltage - at 5.3V the current increase ~25%).
5) By the way, you cannot power Fire HD directly from charging input with battery removed. As long as battery below 2.8V, the gadget do not turn on, regardless of voltage on charging input.

From sum of the symptoms i observe, Amazon Fire HD seems to have either extremely poorly designed charging controller or controller specifically designed to destroy battery as soon as possible. Or may be controller which tends to fail in highly non-trivial way.

 

[sysprog]

Thanks for your cogent reply, @trurle; you make a pretty strong case; I'm glad to see that your assertions aren't baseless, but very displeased to see likelihood of deliberate misconduct -- I hope that another possibility turns up.

 

[Tom.G]

IIRC when Li batteries are deeply discharged they WILL NOT accept a charge.

Ahh, found it! There is a built-in protection circuit.
https://batteryuniversity.com/learn/article/low_voltage_cut_off

Cheers,
Tom

 

[trurle]

IIRC when Li batteries are deeply discharged they WILL NOT accept a charge.

Ahh, found it! There is a built-in protection circuit.
https://batteryuniversity.com/learn/article/low_voltage_cut_off

Cheers,
Tom

So the link above shows the absence of "boost" circuitry intended for revival of deeply discharged battery is common in mobile chargers. Therefore, more likelihood of "poorly designed charging controller" as diagnosis of initial problem.

 

[Svein]

Coldness fixes electronics? No way - what fixes electronics is a good hard slap! (experience from 15 years as a part-time electronics serviceman).

 

[Klystron]

Coldness fixes electronics? No way - what fixes electronics is a good hard slap! (experience from 15 years as a part-time electronics serviceman).

A hard slap to the top of the electronics enclosure actually helped vacuum tube devices. Apparently the jolt helped seat pins loosened by normal on/hot off/cold cycles. This was also when we waited for electronics to "warm up".