** Bad idea.
** Not true - read the data sheet.
It is at least 1.5 amps.
** No, that is the rated RMS current for continuous operation.Conversion to DC reduces the number to under 1.5 amps.
... Phil
Yeah the voltage is only good to a few percent so one will hog the current. If you look at the spec sheet there are circuits with a pass transistor. Make it should be big and beefy to handle the power. It might be much easier to reduce the input voltage.
George H.
Yup. The 78xx dumps a poorly controlled current of several milliamps out of its ground pin, but works fine at no load. When dinosaurs walked the earth (you know who you are) ;) they used to use 78xx regs in LM317-mode. It sorta worked, but the regulation was the pits because of the GND current variation. A diode or two in the ground lead of a 7805 was good medicine for some things.
As we've discussed before, the classical LM317 circuit from the app notes uses a 249-ohm resistor from OUT to ADJ, and whatever resistance is needed from ADJ to ground to set the output voltage. As somebody here pointed out, though, that doesn't meet the worst-case LM317 spec for minimum output current, which is 10 mA not 5 mA. I've switched to
120ish ohm resistors instead.(I think we've all been bitten by the LM317 minimum output current problem at one time or another.)
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
I have gotten different stories from different data sheets -- I think this varies by manufacturer. At the very least, there are TO220 parts out there (KM78xx?) with different current ratings than the usual.
So, yea, read the data sheet of the _specific part_ you're using.
-- www.wescottdesign.com
Of course, dropping 20-odd volts at 2A + 2A is going to require quite the heat sink to keep it out of thermal limiting for long, e.g.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
I mentioned heat sinking once, but I don't know if it sunk in.
A honkin' big heat sink is necessary.
-- www.wescottdesign.com
The 78xx datasheets specify load regulation from 5 mA to 1 amp, so anything goes at light load.
The 7815 is rated for 1A and not 2A. The transformer voltage rating is RMS, not peak, so that a 30VAC is 42V peak, which will annihilate the 7815. The Transformer RMS current is nonlinearly related to the regulator load curre nt, in your case it is in the 1.6x range, meaning you will overheat the tra nsformer and destroy the 1N4007's. Almost everything in your "circuit" is d estroyed.
S, not peak, so that a 30VAC is 42V peak, which will annihilate the 7815. T he
It's 42v at full load and nominal mains voltage. Offload it'll be however m any percent higher - if 10% that's 46.2v. If mains voltage is 10% high then you've got 50.8v. You can take a little off for diode drop, but not much. Less than 1v each at moderate load.
ent, in your case it is in the 1.6x range, meaning you will overheat the tr ansformer and destroy the 1N4007's. Almost everything in your "circuit" is destroyed.
Not exactly. Transformers have a ton of thermal mass, you can exceed curren t ratings massively for short bursts and still stay within design temps.
You can also exceed design temps and still have a transformer that lasts ye ars. Higher temps begin by affecting long term survival rate.
The latter point is true for silicon to a more limited extent. You can run past heat specs, upto a point, and it'll still work for a while.
15-0-15 would be nearer the mark than 30-0-30.NT
** With the disadvantage that you lose current limiting & thermal protection.
I have seen T220 regs wired in parallel in PSUs to get more amps, but the makers add 0.1ohm ballast resistors at the outputs and do a bit of selection too.
.... Phil
tabb
** Winding temp rise depends on average heat dissipation and that goes with RMS current squared times the duty cycle. So for double load current, the duty cycle need to be 0.25 to come out even.For triple, it needs to be not more than 0.11.
Then there is the voltage regulation, which depends on the peak value of the load current. With a small DC supply overload by a factor of 3, the voltage on the filter electro will drop by half or more compared to no load.
... Phil
One approach is to use a PNP transistor, emitter resistor R2 from +V to emitter, base resistor R1 from +V to base and to input of three-terminal regulator. That makes a (roughly) emitter current equal to regulator input current times R1/R2, or less. So, you can set the gain so that the transistor has a scaled current limit from its drive regulator.
***bad ASCII art follows***(+)-----+----R2------+ +--------+----- Vreg output | V____/ | | | | +-----R1----------------+------[7815]----+ | GND The regulator output and PNP collector are connected together, so you get I_output
Yup... but you don't need to stay within temp spec all the time to still have a reliable device. Occasional brief excesses aren't normally a problem. This is why transformers can be thermally protected rather than fused.
What it drops to depends on the transformer's regulation, the reservoir capacity, and the various voltage margins you've built into the supply, so its usually possible to get more current than one would expect without brownout.
NT
** It is much safer to use an internal thermal device, so it cannot be easily replaced with a higher rated one like you can a fuse.
However, once the trip temp is reached they do open fast - like a few seconds.
** A small transformer (say 50VA) has a regulation of 12 to 15% under resistive load. With a rectified and filtered DC supply much worse and with 3 times overload what I posted becomes true. ** Only good for about 20 milliseconds.Baring the use of one Farad caps.
... Phil
but trip temp can be above rated operating temp, and they have a good chunk of thermal capacity.
There's always V drop on the reservoir, every single half cycle, unless load drops to zero. A fat reservoir can reduce this usefully.
One always designs in voltage margins to deal with worst case specs and mains V variations. These can generally be eaten up too. But as you say it all only goes so far.
NT
What can happen with these dual supplies is that one side comes up to voltage before the other one, depending on when the AC is applied. That can cause one of the regs to experience reverse voltage and latch up, which will apply twice the voltage you expect to the other one. I've fixed the problem in the past with some clamp diodes, can't recall the details. IIRC google helped, or maybe some kind assistance from this group.
-- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus
Drat; the ASCII art is a mass of typos, and my silly text editor wasn't dispaying them.... Second try:
***bad ASCII art follows***(+)-----+----R2------+ +---------------------+---- Vreg output | V____/ | | | | +-----R1-------+------[7815]----+ | GND The regulator output and PNP collector are connected together, so you get I_output
You don't get the thermal limit in any meaningful way, because just mounting them in common is too slow and doesn't track well anyhow.
An auxiliary transistor will help.
(+)-----+----R2-+-----+ +-----------+---- Vreg output | | V____/ | | | | | | ----- | | +-----A \_____+ | | | | | R3 | | | | +-----R1--------+------[7815]----+ | GND
But it's really a lot easier to use an LM338 instead, at least for the positive side. The LT1185 is good for the negative side. They both have too low a Vin spec for the job, of course: 35V abs max for the LM338, 40 abs max for the LT1185.
One good use for an auxiliary transistor would be as a power zener or coarse pre-regulator: a 24V zener from the base to ground, base to collector with a resistor, collector to unreg supply, emitter to the voltage regulator. That still isn't great, because you have to use an oversize heatsink to account for the sloppy specs of the vreg's current/temperature limiting.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
The LM7815 thermal limits at 150C, and a pass transistor might be OK to 200C (MJ15004, for instance). So, poor tracking is still good enough. It's easy, too, to predict the heat generated at high loads, and balance the heatsinking so that temperatures track even if the coupling of the devices has significant thermal resistance or time delay.
Higher-current linear regulators, and this kind of trickery, are both kinda obsolete (switchmode converters have taken over). Heatsinks are VERY EXPENSIVE.
regulator 7815:
Good point. A 78xx won't start up if you pull its output below ground. The OP's supply would be better off with a 1N5822 from each output to ground.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.