Jump to content

iPad Explodes - caution users


Recommended Posts

Guest Andys@coffs
OK, I'm looking at a Belkin micro adapter, which is so small there's barely room for the electrons. So I have no idea whether it's a transformer ("linear") type or a switch-mode type, I can't see how there's room for either. But if what you're saying is, it's unlikely that form of DC / DC step-down converter can produce much in the way of voltage spikes, the next question is, if it has a dirty input supply (as from the switch-mode voltage regulator of a permanent-magnet battery charge regulator), how effective is it likely to be in preventing voltage spikes from passing through it? And if it is effective, what happens to the resulting heat? The thing is so small there's hardly any heat dissipation capability there.

A TO220 packaged LM7805 is about 1cm x 1.5cm x 0.5cm and requires an additional 2 to 5 discrete components the largest of these will be electrolytic capacitors which size in microfarads determines physical size.

 

Regarding noise linear regulators are very good with regard noise providing that the transients spikes which swing around the 13v supply don't have the negative swing penetrating lower than say 5.5v. If that is the case you'll get pretty clean supply of 5v

 

If your supply does have transients of that size then im pretty sure it will have made itself known when with engine on suddenly intercom /radio noise is terrible. Bottom line is noise on the input size isnt simply scaled back to 5/13ths and reproduced on the output side such as would occur if we were using a resistive voltage divider.

 

Your question of heat is a real issue. Switchmode supplys are very frugal with heat generation and linear regulators by comparison aren't. Looking at the pdf datasheet for the device shows it has a thermal overload protection so presumably it will shutdown if it gets too hot. Bottom line is that linear regulators of this type will certainly run hotter than a typical wall mount plugpack switchmode, for the power that they produce 1A so 5w, which is way below the 12w that apple requires for latter model ipads. This has an effect that even plugged in and charging the onboard battery will slowly discharge because device power usage will generally exceed 5w. But for usage life 5w in is still miles better than 0w in and in the event that the device is off (well screen is off) the device will slowly recharge from a 5w capable input. I have said the device is capable of producing 1amp of output and that is true when the device has suitable heatsink connected to it. If it doesnt then the output capable will be lowert

 

Andy

 

 

Link to comment
Share on other sites

  • Replies 89
  • Created
  • Last Reply

Top Posters In This Topic

Thanks, Andy. One hears occasional complaints about intercom noise from Jabiru owners, which suggests to me that the built-in supply supression in intercoms etc is occasionally marginal for the six-pole permament magnet alternator on the jab. Most of those things would, I expect, be designed to deal with whatever comes out of a typical field-current regulated automotive-type alternator - but the voltage regulators on those can produce quite a bit of noise, too.

 

I certainly don't fancy a cigarette lighter socket in the aeroplane, no matter what's plugged into it. A hard-wired converter is definitely the way to go. By the term "linear" converter, do you mean a resistive voltage-divider type? I do not have the alternator capacity for such a luxury. I've seen this term applied to transformer style converters, which, for a DC input, must have a switch-mode device upstream, but presumably rely on some silicon iron to round-off the corners of the waveform, before feeding it into a full-wave rectifier.

 

I'm looking at the Rotax 582 lighting-coil output, which I would expect to be an order of magnitude worse in regard to ripple than the Jab alternator, and also bad in regard to voltage spikes from the regulator. So I was thinking more along the lines of trying to provide a 12 volt avionics bus with about 6 amp average capability, with the filtering / suppression at the take-off point of that bus. I'm not too keen on electrolytic capacitors for this; I've had electrolytics explode before. However power-factor correction capacitors for fluorescent lights are typically around 3 microfarad, 600 volt - and they are quite light; eight of them, giving 24 microfarad, would not be impractical - I can accommodate the bulk, and they would be easy to keep cool. So would a suitable choke using ferrite rings, presumably one would use a pair of them, one for each side of the circuit (I'm planning a 2-wire system, not a ground-return system). So you can see that miniaturization is not important to me, but reliability is. One of the compensations of an aircraft, is that cooling air is readily available. That sort of setup should go some way towards calming the 12V supply down to manageable proportions for the power supply filtering built into the avionics.

 

I will be running the ipad, a GPS, a VHF COM and a transponder off that bus. Whatever 5 V converter I use for the ipad, would draw from that bus; perhaps I should be looking at two LM 7805s in parallel.

 

 

Link to comment
Share on other sites

Guest Andys@coffs

David

 

To avoid boring those not interested in electronics I wont go into huge detail, other than point you here http://en.wikipedia.org/wiki/Linear_regulator this gives a better explanation than I could put together. the thing to remember is that it uses the example of a resistive voltage divider but uses the work "like" and infers that the device is in real time adjusting the resistance values so that supply voltage (including noise and spike) are always divided down to exactly the output voltage we need, and without the negatives like having to design for a known current draw and dealing with the wasted current lost in the divider network..

 

If you need more than 1A output then instead of the TO220 style 7805 there are larger packages available in TO3 package that can supply 3A or even more if you use a large power transistor in series. If you look at the Fairchild datasheets (first provided in google with a search for "LM7805 datasheet pdf") then at the back are a number of example circuits that can be used to vary the output voltage from the stock 5V or as you need, increase the current that the regulator can supply.

 

National Semiconductors was, from memory, the developer of these regulators and the associated family, there will without doubt be a regulator that will meet your need in the family. A higher current one that comes to mind without look is the LM317K but in that case its a variable regulator and judicious use of resistors determines the final output voltage. I seem to recall that its a 3A regulator or a LM117K which is 5A out (Both need external heatsinks and not insignificant at that. In fact at higher currents its then that switchmodes start to make sense. Linear regulators in laymans terms produce heat because the spare 7 to 8V is an unwanted side effect, a switchmode supply is constantly switching between full input voltage and zero volts (and hence there is no wasted 7 to 8 v that produces heat) and uses an inductor and a series of capacitors to produce an average voltage. obviously if the pulse width of the full output becomes narrower and narrower the average voltage drops but its getting rid of the pulses that relates to the effectiveness of reducing noise at the output.

 

Andy

 

 

Link to comment
Share on other sites

DavidTo avoid boring those not interested in electronics I wont go into huge detail, other than point you here http://en.wikipedia.org/wiki/Linear_regulator this gives a better explanation than I could put together. the thing to remember is that it uses the example of a resistive voltage divider but uses the work "like" and infers that the device is in real time adjusting the resistance values so that supply voltage (including noise and spike) are always divided down to exactly the output voltage we need, and without the negatives like having to design for a known current draw and dealing with the wasted current lost in the divider network..

 

If you need more than 1A output then instead of the TO220 style 7805 there are larger packages available in TO3 package that can supply 3A or even more if you use a large power transistor in series. If you look at the Fairchild datasheets (first provided in google with a search for "LM7805 datasheet pdf") then at the back are a number of example circuits that can be used to vary the output voltage from the stock 5V or as you need, increase the current that the regulator can supply.

 

National Semiconductors was, from memory, the developer of these regulators and the associated family, there will without doubt be a regulator that will meet your need in the family. A higher current one that comes to mind without look is the LM317K but in that case its a variable regulator and judicious use of resistors determines the final output voltage. I seem to recall that its a 3A regulator or a LM117K which is 5A out (Both need external heatsinks and not insignificant at that. In fact at higher currents its then that switchmodes start to make sense. Linear regulators in laymans terms produce heat because the spare 7 to 8V is an unwanted side effect, a switchmode supply is constantly switching between full input voltage and zero volts (and hence there is no wasted 7 to 8 v that produces heat) and uses an inductor and a series of capacitors to produce an average voltage. obviously if the pulse width of the full output becomes narrower and narrower the average voltage drops but its getting rid of the pulses that relates to the effectiveness of reducing noise at the output.

 

Andy

Thanks, Andy.

 

 

Link to comment
Share on other sites

Thanks for trying to keep it simple, Andy, but all that electronic talk made my head spin.

 

I have a couple of cigarette lighter sockets to run add-on devices. Now it seems the Jabiru-type regulator may not protect them. Can any of you knowledgeable types tell me if the charging adaptors used by my radio, iPhone and iPad protect those devices from voltage spikes? If not, can you suggest a filter I can fit upstream of these sensitive items?

 

 

Link to comment
Share on other sites

I know you are worrying a lot about this charging issue: you can buy external rechargeable batteries for ipads with usb output ports. If you don't trust them rechargeable batteries you can even buy ones that run on alkaline batteries... try ebay

 

 

Link to comment
Share on other sites

Agree re dc converters, surely with the wide uptake of ipads in commercial use theres someone making isolated or high quality 12vdc power supplies......... Not a ciggy lighter typeNarva do have hardwire USB port available but no idea whats inside.

Did you say Narva and high quality in the same post.....thats imaginative.

 

 

Link to comment
Share on other sites

Guest Andys@coffs
Thanks for trying to keep it simple, Andy, but all that electronic talk made my head spin.I have a couple of cigarette lighter sockets to run add-on devices. Now it seems the Jabiru-type regulator may not protect them. Can any of you knowledgeable types tell me if the charging adaptors used by my radio, iPhone and iPad protect those devices from voltage spikes? If not, can you suggest a filter I can fit upstream of these sensitive items?

I think you'll be fine as is. My J230 cig lighter drives a couple of cig lighter chargers without any issue. Imho i think we are worrying about low risk items and probably ignoring much higher risk of likelihood issues. If your battery is good quality and not end of life, and the high current wiring and terminals are in good condition then the battery itself is a pretty reasonable noise filter in its own right. In reviewing the incidences and issues so far this year in the ATSB 1/4yr reports no out of control device fire was listed as an incident. There were plenty of incidents however to focus our attention on other causes however

 

Andy

 

 

Link to comment
Share on other sites

to put those risks in perspective, there are currently over 150M ipads running at any one time, that 3.6B hours of operation every 24 hours or 1314B hours of operation every year. Plan for the worst and manage your risks appropiately

 

 

Link to comment
Share on other sites

to put those risks in perspective, there are currently over 150M ipads running at any one time, that 3.6B hours of operation every 24 hours or 1314B hours of operation every year. Plan for the worst and manage your risks appropiately

How many of them are running on power from the lighting coil of Rotax 582?

 

 

  • Agree 1
Link to comment
Share on other sites

take the car for gods sake

I often do; but when I want to fly, it really does not serve. This thread has provided quite a bit of information for me, thanks to Andy (plus a bit of research). I suspect he's perfectly correct insofar as the "large" Jabiru alternator on the 3300 engine is concerned; and I now understand what he means by a "linear" adapter; it will likely give adequate performance on that alternator. It will be dissipating about 18 watts of heat in the process. I'm using a Rotax 582 on my motorized Blanik, can't fit a jab 3300, so my question is apposite. There are lots of Trikes etc flying with that motor. It will have an order of magnitude greater ripple in its power output, unless I miss my guess, so what works for a Jab 3300 may not work for it. It may indeed need a filter of some sort.

 

 

Link to comment
Share on other sites

Just to add a little confusion (though hopefully not too much), I have to disagree with Andy's comments that the 12->5v converters are likely to be linear. The power dissipation (9W for a 1A converter working from 13.8V) is just too high for the cigarette lighter plug packs. They'll almost certainly be switchmode devices, running in the megahertz range. These high frequencies allow the inductor and caps to be very small, and they easily run at >90% efficiency so they're dissipating less than 500mW (and they're still as cheap as chips).

 

Their safety basically comes down to the components used, and to put a rough rule of thumb on it, the more expensive, the better. Why? Two factors. The first is that the switching FET - one of the more expensive parts @~20-30 cents) is subject to all the spikes that exist on the main bus, and in any automotive or aircraft system, they can be very large indeed. If the FET fails, it typically fails short-circuit, not open. That then puts the full 12-15V of the bus onto the 5V input on the electronic equipment. Most of the time, it'll blow a fuse (either in the equipment or on the panel), but if you combine it with the large spikes that took out the FET, they could take out the battery too (with extreme prejudice). Given that these spikes tend to occur at engine start and shutdown, it's wise to switch off the power to the converter before starting or stopping the engine (just like you always do to protect the rest of the avionics, don't you!).

 

The second factor (though it doesn't affect the battery safely) is RF conducted & radiated emissions. It's not hard to suppress the emissions from a power supply if you do it at design time, but if you're either slap-dash in your design or looking to shave every cent you can from the BOM cost, then it's easy to 'accidentally' leave off the components that make the unit meet the international specs. It is illegal to sell electrical equipment in Australia that doesn't meet the requirements, but if you're buying cheap (and especially if you're buying off the web), there's every chance it will radiate like crazy. There's no easy way to check this your self, but basically, if you use a converter and you start hearing noise in the intercom or radio, chuck it and go and get a better one. Life's too short to stuff around with making a $5 converter work in your system.

 

Bottom line? If you're powering your $800 iPad, don't use a $2 converter.

 

 

  • Agree 1
  • Informative 1
Link to comment
Share on other sites

Guest Andys@coffs
Just to add a little confusion (though hopefully not too much), I have to disagree with Andy's comments that the 12->5v converters are likely to be linear. The power dissipation (9W for a 1A converter working from 13.8V) is just too high for the cigarette lighter plug packs. They'll almost certainly be switchmode devices, running in the megahertz range. These high frequencies allow the inductor and caps to be very small, and they easily run at >90% efficiency so they're dissipating less than 500mW (and they're still as cheap as chips).Their safety basically comes down to the components used, and to put a rough rule of thumb on it, the more expensive, the better. Why? Two factors. The first is that the switching FET - one of the more expensive parts @~20-30 cents) is subject to all the spikes that exist on the main bus, and in any automotive or aircraft system, they can be very large indeed. If the FET fails, it typically fails short-circuit, not open. That then puts the full 12-15V of the bus onto the 5V input on the electronic equipment. Most of the time, it'll blow a fuse (either in the equipment or on the panel), but if you combine it with the large spikes that took out the FET, they could take out the battery too (with extreme prejudice). Given that these spikes tend to occur at engine start and shutdown, it's wise to switch off the power to the converter before starting or stopping the engine (just like you always do to protect the rest of the avionics, don't you!).

 

The second factor (though it doesn't affect the battery safely) is RF conducted & radiated emissions. It's not hard to suppress the emissions from a power supply if you do it at design time, but if you're either slap-dash in your design or looking to shave every cent you can from the BOM cost, then it's easy to 'accidentally' leave off the components that make the unit meet the international specs. It is illegal to sell electrical equipment in Australia that doesn't meet the requirements, but if you're buying cheap (and especially if you're buying off the web), there's every chance it will radiate like crazy. There's no easy way to check this your self, but basically, if you use a converter and you start hearing noise in the intercom or radio, chuck it and go and get a better one. Life's too short to stuff around with making a $5 converter work in your system.

 

Bottom line? If you're powering your $800 iPad, don't use a $2 converter.

So I had one of the cheapies, it was an iphone 3 generation and it was a linear regulator and I opened it up when a fuse blew a few years back....

 

I also had a 3rd party logic 3 charger from big W that was bought for a iphone 4s and opened that one up tonight after reading your post to find the telltale toroid inductor and a TO92 FET.....

 

So your right it is a switchmode based. Clearly my history with the iphone 3 3rd party charger is out of date....

 

If people are concerned about switchmodes then a linear regulator despite the heatsink requirements may well be a better choice for a regulator where RF and or switching transients are of concern. If you want to be sure, try and get one that has crowbar protection. what that is, is circuitry that is separate functionally from the regulator that monitors the output voltage and in the event that it exceeds the designed safety limit of the crowbar circuit the circuit will introduce a switched shortcircuit across the output rails which then leads to a blown input fuse....

 

Apologies to those I have mislead!

 

Andy

 

 

Link to comment
Share on other sites

Well, there you have it . . . .

 

One small point could perhaps be (I'm speculating here) that many of the cigarette-lighter style adapters offer two USB outlets - which may be a way of exporting the "wasted" power and thus reducing the heat, provided a load is connected to each of the outlets; added to the liklihood that there's a thermal shutoff built into the things - which would mean that they will likely to be charging less than half the time. My reading of all this is that if your cig. lighter adapter is charging full-time and staying nice and cool, it's likely to be a switch-mode device, and therefore likely to be putting voltage spikes into your ipad. If it's getting hot & bothered & only charging part-time, especially if it's only driving one outlet, it's a "linear" type, and probably not putting voltage spikes into your ipad. The 2.1 amp ones for the ipad will be far more obvious under this crude test than the 1 amp variety for a GPS or whatever.

 

As Malcolm Frazer said, "Life was not meant to be easy". Why not use a pair of 6 volt batteries in series, and run an LM317K linear regulator from each of the 6 volt halves, for the ipad & GPS? That gets the waste heat way down - to about 3 watts, instead of 27 for the two outlets. I'm planning to build my own L-C filter for the 12V supply to the radio & transponder, including crowbar protection; ferrite toroids are remarkably cheap, and I have some 25 MFD 450 volt polyethylene film capacitors laying about . . .

 

 

Link to comment
Share on other sites

I'd like to thank everybody who has made a useful contribution to this thread; especially Andy and SfGnome; we've gone around in circles a bit, but I think we've unearthed a potential problem for panel-mounted ipads etc; and I for one, now see a potential solution. It remains to get an oscilloscope on the job and find out whether - or to what extent - the issue is a real problem. I suspect it will be, for Rotax 582 users.

 

 

Link to comment
Share on other sites

Lets face it! I have an ipad and I have paper charts. The ipad is an excellent backup for the paper chart! To which I might add - I've never had to charge a paper chart and it has never let me down when I most need it! Even if you use an ipad you still have to have your chart with trip info up to date don't you!!

 

 

Link to comment
Share on other sites

Guest Andys@coffs

There's always 2 sides to every story.....

 

Try and make that claim if you fly in an open cockpit.......I have lost maps overboard, followed by an instantaneous Brrrrp as the prop dealt with the offending map when trying to refold to cover the ground im currently over..... I'd need to be going a whole lot faster to have the cockpit wind grab hold of the ipad ....and in any event don't need to take out of the kneeboard cover or pick it up because it constantly always reflects the ground Im over..... but if it did go through the prop then navigation beyond my immediate area at that point of time is likely to suddenly become a whole lot less important:lost:

 

Andy

 

 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×
×
  • Create New...