Smoking those 7805s
LoopyByteloose
Posts: 12,537
Hello,
I have a project that I wanted to use my model r/c in.· It is a robot with two 6volt lead acid gel cells in series.· The output is 12volts and eventually I want them to drive 12 volt motors.· So, I built a little voltage regulation board for my 4 channel r/c reciever and have the whole thing set up with servo motors on a trial basis.
I had some troube with connects and so forth and my first regulator seemed to become inoperable.· So, I built another regulator setup with a few improvement.·
Again I was having trouble because I really need to inventory what batteries are charged and what are not.· The result is that at one point I know I pluged something in backwards on the supply side of the 7805.· Now,·the 7805·seems to only output about 1.13volts.
Did I once again smoke the regulator in this fashion?·
I bypassed the regulator and have checked the reciever w/·servos·and everything else is functional.
I really think the regulator is damaged, but I thought they were more robust that I am experiencing.
Is it a big 'no no' to input reversed polarity on the output side?· Or, could it be that I short-circuited the input or the output?
I would like to know so that I don't keep burning up the regulators.· As it is, I am onto building my third board.
It would not hurt to re-read the specifications after all this destruction.· I might understand them better.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
G. Herzog in Taiwan
Post Edited (Kramer) : 8/14/2005 9:28:40 AM GMT
I have a project that I wanted to use my model r/c in.· It is a robot with two 6volt lead acid gel cells in series.· The output is 12volts and eventually I want them to drive 12 volt motors.· So, I built a little voltage regulation board for my 4 channel r/c reciever and have the whole thing set up with servo motors on a trial basis.
I had some troube with connects and so forth and my first regulator seemed to become inoperable.· So, I built another regulator setup with a few improvement.·
Again I was having trouble because I really need to inventory what batteries are charged and what are not.· The result is that at one point I know I pluged something in backwards on the supply side of the 7805.· Now,·the 7805·seems to only output about 1.13volts.
Did I once again smoke the regulator in this fashion?·
I bypassed the regulator and have checked the reciever w/·servos·and everything else is functional.
I really think the regulator is damaged, but I thought they were more robust that I am experiencing.
Is it a big 'no no' to input reversed polarity on the output side?· Or, could it be that I short-circuited the input or the output?
I would like to know so that I don't keep burning up the regulators.· As it is, I am onto building my third board.
It would not hurt to re-read the specifications after all this destruction.· I might understand them better.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
G. Herzog in Taiwan
Post Edited (Kramer) : 8/14/2005 9:28:40 AM GMT
Comments
for the 7805 regulator, I normally use the components as shown in the attached schematic.
D3 makes sure that a wrong polarity of the input voltage does not kill the 7805. C8 and C9 should be ceramic caps, located as close as possible to the 7805 pins - they make sure that the chip does not oscillate. In case there is another large cap connected to the output voltage side, at power-off, this cap can force a reverse current through the 7805 when it discharges. Although D3 blocks this reverse current, as an additional safety measure, I usually connect another diode (D2) in reverse across the input and output pins of the 7805 to bypass any reverse current, as the 7805 can easily be killed by such currents.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Greetings from Germany,
G
I agree with you, it's a good idea to "keep cool" the 7805 at higher loads. On the other hand, the datasheet states that it has internal current limiting and thermal shutdown features for "safe operating area protection, making it essentially indestructible".
What can definitely "destruct" it is reverse current, wrong polarity, and oscillations. Therefore always attach the 100nF caps close to the in and out pins. For the in pin, the datasheet even specifies a 330nF cap, although I never had problems when using a 100nF cap instead.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Greetings from Germany,
G
Heatsink was already installed.
I may have had an oscillation problem, too.
It is hard to tell the origin. The batteries on the transmitter were low and it seemed to go away when I changed them. I am not ver familiar with the transmitter's low power behavior.
With servos and R/C involved I suppose that I have at least three and maybe four places that will cause the servo to move back and forth [noparse][[/noparse]an indication of oscillation].
1. Low power on the transmitter
2. Low power on the receiver
3. Low power to the servo
4. Improper set up of the 7805.
I have a wonderful like robot chassis here that I want to make work.
The top and the bottom are two Frisbees from the local Toys R' Us and the circular stainless steel decks with pre-drilled holes are what is commonly called hot plates or trivets with folding legs [noparse][[/noparse]a common local product in Taiwan]. They fit perfectly into the Frisbees and I can add as many decks as are needed in the future [noparse][[/noparse]that is why the heavy lead/acid gel cells for power].
Being round, it really gets less tangled up in the world around it and I hope to get it to turn on its own center.
Being expandable, I hope to get a lot more use out of it than another design.
As for the 7805, I do see all sorts of capacitor arrangements on the internet. The 100nF are commonly recommended by the manufacturers. The diode across the i/o is manufacturer recommended, but I have been to lazy to install one.
Since I have to build my THIRD supply [noparse][[/noparse]the other two are definitely dead and gone], being lazy was a complete waste of time.
The input protecton diode [noparse][[/noparse]D3] adds another 0.7 voltage drop. I have seen an alternative of having a reversed direction diode from Vdd to Vss input that will clamp a wrong input and protect the 7805 by short circuiting the power supply. That second configuration may be better for batteries where that voltage drop means less running time, but it could do real damage to wall warts and other supplies that are unfused.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
G. Herzog in Taiwan
Post Edited (Kramer) : 8/14/2005 9:26:21 AM GMT
The Manufacturer's sheets warn that using 10uF or larger ripple capacitors must have protection diodes. It is because these capacitors work as Reverse Votage Spikes on disconnect. [noparse][[/noparse]I happened to put a 50uF capacitor on the input side without the diodes].
So, it looks like the 3rd time is the charm. A lot of this stuff is very important to hobby electronics because you are constantly connecting and disconnecting breadboards. It is important to have your power supply, your regulation, and your project protected from errors and short circuits.
Thanks again.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
G. Herzog in Taiwan
well, using a reverse direction diode from plus to minus on the input side is a "brute-force" method, IMO. I once used this approach but with a fuse in series with the positive supply line, so this fuse would blow in case of wrong polarity. I've been using a small 500 mA fuse type which is normally used to go in series with the mains supply line. Just being curious, I measured the voltage drop across the fuse at full load, and it was almost 1 Volt, IOW more that the voltage drop across the series diode.
For battery-powered systems, you may consider using an LM2940 instead of the 7805. This type has a dropout voltage of 0.5...1V, i.e. the input voltage can be as low 6V with still 5V on the output side. The datasheet does not warn about reverse currents - nevertheless, I also use a diode for bypassing the reverse current, and the input protection diode for this device as well. Important with the LM2940 is the output filter capacitor. It must be at least 22µF with low ESR (Equivalent Series Resistance). Best would be a tantalum cap for this purpose. The dataseheet specifies an inut cap of .47µF. I normally use a higher value here, and also have the two 100nF ceramics installed. IOW, I use the same schematic for the 7805 and the LM2940.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Greetings from Germany,
Günther
Frankly, a lot of what is on the shelves in the few electronic supply stores here in Kaohsiung is production over-runs, overstocked items of older technology, and 2nd rate quality. The factories get all the really good stuff, but it goes into production and export. The good stuff stays in the duty-free zones, too.
I have to import the BasicStamp even though it is made here. It travels about 12,000 miles for delivery.
I would use the LM2940-5 if I could get them, but I haven't had any luck here. I have only seen the 12volt version [noparse][[/noparse]useless!] The people at the parts store are now giving me a professional discount because I ask for so many things that they have never heard of that they feel guilty. I imagine they think I am designing something very state-of-the-art and will later place a big order.
Since shipping and duties can add 33% to the cost, I am reluctant to import unless I buy a lot cheaply.
Since having a fuse would waste more energy that having the diode, I may take Gunther's suggestion. Still, all this protection and regulation eats into a micropower project.
What I really should get a a switching regulator with low drop out, but again I have to import. Nobody is stocking retail supplies.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
G. Herzog in Taiwan
too bad about the supply situation of - what I thought - such popular components in your area. Concerning switching regulators, I made very good experiences with the LM2574 from National Semiconductor. This chip, together with 2 capacitors, an inductor, and a fast Schottky diode makes up a nice step-down voltage regulator with input voltages between 7 and 40 Volts, and an output current of 500 mA max. You can get devices for 3.3, 5, 12, 15 V and an adjustable version. The devices have built-in thermal protection, and current-limiting. So far, I've used over 200 of them, and not one of them failed so far.
You can find a datasheet at the Natinal Semicondictor's WEB site (www.national.com). I only fear that you won't get those components either, locally.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Greetings from Germany,
G
Ryan
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Ryan Clarke
Parallax Tech Support
rclarke@parallax.com
Right, I believe that the IC mask is virtually the same for all of the 78xx regulators as well as the popular 314/317 regulators.
Essentially they are all adjustable regulators, the difference is the selected resistor values within the IC. This can easily be
done in a single mask layer by placing the right "jumper metal" to make the proper selection or you can do it externally.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
It took a few seconds for me to realize what was happening and pull the 9V battery that was powering the board.
It still works fine though.
I guess I got lucky.
Then again a rechargeable 9V doesn't have quite the kick of a 12V lead-acid battery pack.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I wonder if this wire is hot...
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
The regulator looses its magic smoke and stops working!!
·· The problem with internal protection is that under the right conditions it can fail.· The 7805 can go into thermal runaway under certain conditions, making the thermal protection useless.· Same with Amplifier Chips.· I have seen some with Thermal Runaway protection fail from Thermal Runaway.
·· One way to protect yourself is (I used to do this years ago) to use a lead from your supply with an in-line fuse holder, such as you find in automotive wiring harnesses.· You put in a fuse value rated to just above the anticipated current level.· Anything extreme such as a short will blow it and generally save your components.· This is, of course, not a fail-safe, but good protection for beginners, even when using batteries.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
How much current can one draw through a 7805 with a small heatsink?
Through a 7896 tin can bolt down version?
Rafael
·· As was posted earlier, many factors affect that answer.· A small heatsink will certainly help, but as I recall in the specs there is a thermal rating regarding heatsink surface area, thermal transfer material, etc.· Just count on getting close to 750mA from it with a standard TO-220 heatsink (Is there a standard?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
Post Edited (Chris Savage (Parallax)) : 8/26/2005 11:54:25 PM GMT
Rafael
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
Rafael
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
As I had mentioned earlier, the 78xx mask designs are internally ALL the same with the exception of a resistor tap that sets
the appropriate voltage. They do this for cost effectiveness when producing the chip. So it would make sense that ALL 78xx
series chips have the same thermal characteristics.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.