Electrical Engineering Stack Exchange is a question and answer site for electronics and electrical engineering professionals, students, and enthusiasts. It only takes a minute to sign up.
Sign up to join this community
What are the disadvantages of using a 7805 (with a heat sink) to use in a car USB phone charger?
Another option: It gives a 1 A current, so could I maybe wire two of those and give a 2 A output?
In past 10 years have tried more than a dozen chargers and they all break and stop working in a month. Have already spent lot of money and now tired.
Whilst you can make a power supply out of a 7805, you need a small difference between the input and output voltages and a small current draw.
Automotive electronics needs to work with around 14V - that's the voltage from the alternator. So you have a 9V drop on the 7805, and 2A. That's 18W. Suppose (round numbers) you could tolerate a 56 degree rise in temperature, that'd need a 3 degC/W heatsink. Looking online, you're going to need a heatsink that's around 10cm x 10cm.
This is going to be very inconvenient to put in your lighter socket! It's also going to cost you around £5 for the heatsink, before you even start thinking about the cost of anything else like an enclosure.
Automotive electronics needs to be protected against spikes of high input voltage, reverse input voltages, reverse current draw, output short circuit to 0V, output short circuit to 12V, overtemperature, and those are just the ones I can remember. Linear regulators exist which have all this protection built in for not much more money, but the basic 7805 doesn't. It is a very bad choice of component in a harsh environment like a car, and without all those protection features you could easily kill your phone or the 7805 if the 12V is less than perfect - and in a car it often isn't.
But even with an automotive grade voltage regulator, you still have the problem of needing a huge honking heatsink.
Because it gets too hot. A linear regulator like the 7805 reduces voltage by turning that excess voltage into heat.
So if you feed it 12V from the car to produce 5V@1A, it is turning 7V@1A worth of power into heat. That's 7W which is a lot. Even if you had a huge heatsink that rises 5C per watt, that would still be a 35C rise above ambient. At ambient temperature of 25C, that much temperature rise is starting to burn skin.
With a more reasonable heatsink it would be easily be a rise of more than 100C rise above ambient.
A 7805 is a linear regulator. It is inefficient and can not provide 1A output if you want 12V in and 5V out. For getting that 5W out from a 7805, there will be 12W going in, so 7W wasted as heat by the 7805. The 7805 can't handle that. Also you can't just parallel two linear regulators by connecting them together, so paralleling needs some additional circuitry to be designed, and since one 7805 can't provide 1A, then two of them can't provide 2A.
So for the money to spent on designing how to make it work with a 7805 taking account the power dissipation and cooling, it is cheaper to just design a 3A switching mode power supply that can provide 5V at e.g. 3A and be 95% efficient, so for 15W output, only 15.8W input is required, and only 0.8W is wasted as heat.
A cheap-as-possible phone charger will use something more like an MC34063 which does not require a heat sink and is very inexpensive in quantity ($0.05 maybe).
As others have noted, there are several reasons why a linear regulator is not a good idea for a car charger.
It boils down to efficiency, which is horrible for linear regulators in general, especially when there's a large input-output voltage difference.
Add to this the cost to provide adequate heat-sinking when the current is above 100-200mA or so. Good heat sinks do cost quite a lot compared to other components.
Moreover, your linear regulator choice is also quite debatable: the 7805 chip is an old dinosaur. A much sturdier beast (almost as old, but much more rugged in my experience) is the LM317, as far as linear regulator goes. But still it can't handle more than 1A without external circuitry.
All in all switching power supplies are the right choice, and they can be quite rugged, if designed correctly, and still remain affordable.
Until now I more or less said what others have already said. However, what I want to point out specifically is the following statement of yours and some possibly related issues nobody has mentioned yet:
In past 10 years have tried more than a dozen chargers and they all break and stop working in a month.
Yes, indeed cars are a harsh environment for electronics connected to the car electrical system, but more than a charger a year is quite suspicious, especially if they lasted just a month or so.
If you purchased low-quality, el-cheapo brands this could be barely justifiable (I have bought some of them in the past, and none failed, although I admit I never used them for very long periods). However, if you purchased quality ones I would start questioning your specific situation regarding "electrical cleanliness".
Maybe the electrical system of your car puts out particularly nasty spikes from the 12DC outlet, and this could cause the damage. If the outlet wiring was routed too near the high voltage circuits that power the spark-plugs, this could happen. You could search the Internet for your car model and see if this is a recurring problem.
Another issue could be ESD (Electro-Static Discharge) problems. Especially if you live in a very dry climate, the synthetic upholstery of the car, together with your clothes may act as a very high voltage generator (thousands of volts) when rubbing together. In turn this could cause yourself to being charged to a substantial electric potential.
Touching the charger while you are so charged-up could potentially damage the internal circuitry, especially if the case of the charger is light plastic and the circuit has little ESD protection. If you ordinarily experience your fingers being "zapped" when you enter or exit your car, that could be a hint of an ESD problem.
Of course the El-chepo charger brands are probably more susceptible to ESD damage, so the two issues may compound.
I know that this is possible, because I did use 7805 as a phone charger. It took breaking 2 devices (one for the cable and the other for the 7805 regulator itself), but it was kind of emergency.
The chip failed to effectively dissipate the heat when mounted on a painted steel sheet (this is what we had available) and its thermal protection triggered before it had some meaningful effect on the phone battery. So I ended up immersing it in a glass of water. The positive leads (in&out) corroded a bit for the few hours of use, but everything worked fine.
Using two 7805 to get 2A output is not going to work - except maybe by using equalizing resistors of ~0.1 ohm at their output terminals and living with some voltage loss. 7805 are not created equal so one of them will always have few mV higher voltage and will bear the whole load. When overloaded, most 7805 don't simply limit the output current to 1A, but almost completely shuts down itself, leaving less than 100 mA output.
This generally doesn't matter because when you miss the circuity that advertises the charger capabilities to the load, phones and other USB-charged devices generally limit themselves to either 0.5A or 0.9A .
Why factory-made chargers get burnt in you car?
I cannot replicate your experience. Even the cheapest, grocery-store sourced chargers work flawlessly in my cars.
I can propose few reasons:
Most car chargers are happy to run off either 12V or 24V power.
On the other hand, some of them are 12V-only and fail (not necessarily right away) when the voltage goes 20V or higher.
Even a 24V-capable charger may find itself at its limits when the engine is running, the weather is cold (or the voltage regulator is wonky) and the socket actually gives off 30-33V. 33V (or 16.5V in a 12V system) is bad, but still not fatal for the car batteries so one may not be immediately aware of some problem there.
In past 10 years have tried more than a dozen chargers
Calculate how much you've spent on them, in total. Then buy a car charger that costs that much. It will last more than 10 years on average I bet :)
Stick to brand names, and claim warranty benefits when needed. Some brands in some markets will offer warranty longer than a year, so that's worth looking into as well.
and they all break and stop working in a month.
I know you're a bit hyperbolic, but even if those chargers lasted a quarter each, that would be 40 chargers. I don't think I have bought that many over the last decade for my entire household, never mind just to keep in a car.
A car charger dying in a month is somewhat typical for bottom of the barrel stuff that, for example, is typically sold in gas station convenience stores, or at the grocery store checkout lanes. Don't waste your time and money on that. It's usually junk made to work long enough that the customer is a long way away by the time the thing fails. It's usable if you're "in a pinch" and absolutely need a charger and there's nothing better available nearby or there's no time to investigate or drive to get it. Otherwise, those cheap things are best avoided.
It's not particularly hard to design a one-off switching power supply charger that will last longer than the USB connector on it. It will cost in the $50-$100 ballpark due to low quantity, if you're experienced in such designs.
As other answers have well argued, a linear regulator is not desirable. It may be "just a 7805" and yet it may not last long anyway. I've seen plenty of 7805s fail due to running hot and heat cycling in office applications, never mind in a car. A 7805 may well be wiped out just by transients in an automotive application - if you don't protect it sufficiently.
This is a bit of an X->Y problem. The purely linear regulators are not suitable for your objective as the other answers have suggested.
However, when I was faced with a similar problem - 5V power from a 12V vehicle with little heat generation - I ended up with switching regulator modules by Recom: https://recom-power.com/en/products/switching-regulators/switching-regulators-sip/rec-s-R-78K-2.0.html?2
I only required 500 mA in my project, but there's a 5V, 2A version of it as well:
R-78K5.0-2.0
You should probably still use a capacitor on the input and outputs though.
