Two questions about 7815/7915 V-regs

[Emperor-TK]

While working on my second GSSL, I had the x-former hooked up wrong originally and was getting some strange voltage readings, even after correcting the hook-up. I at least blew out my bridge rectifier, since it was passing only AC. I decided to rebuild the power supply on perfboard and check all the components as they are re-installed. I now have everything together, with the exception of the V-regs. I get +/- 24V DC and everything seems fine, but when I hook up the V-regs, I start blowing fuses. The Fairchild data sheets seems to suggest that it is nearly impossible to damage one of these. Could my problem be that I am probing the supply without any load, or does it seem like the v-regs are likely damaged?

Secondly, I am going to build a Seventh Circle kit and am interested in adding a few other DIY modules to the chasis. The SCA power supply runs at +/-27VDC. If I were to put a green pre into this chasis, would 7815/7915's be OK to flip the voltages up/down to 15? Are there any trade-offs, such as reducing the total current capacity of the SCA power supply (through heat)? It almost seems like magic that a V-reg can just be thrown in to a power supply without any perfomance trade-offs, so there must be something, no?

-Chris

 

[Svart]

well you *should* be able to just put a Vreg on a DC rail as long as it doesn't go above the max input V for the reg. However, in reality you HAVE to have proper decoupling and smoothing caps on the inputs and outputs or sometimes they oscillate and sometimes they are just plain noisy, both of these waste energy and can lead to failure. you should aim the feeding voltage for about 3v above what you want the output to be to minimize heat and maximize efficiency. you also want a fairly high VA rating for the PSU trafo. It sounds like those Vregs are dead. it's REALLY easy to kill them regardless of what the manufacturer wants you to believe, usually they also make statments like that assuming that you installed everything correctly and have proper layout and design. Running them without a load won't damage them.

as for using the 15v regs for voltage reduction, you should be fine if you use proper decoupling. the current rating will be set by the rating of the Vregs themselves, the to220 parts are usually around 1-2A total current @ some ambient temp.


that's a layman's explaination.. if you need/ want more depth just say so and one of us will be around to give it to you!


:thumb: :thumb:

 

[CJ]

What a coinsidence! I popped two 7915's oer the last few days trying to hook up a high school level pwr supply.

I think the pinout is different for the 79xx series.

Going to find data sheet.

 

[BYacey]

78xx -
pin 1 input
Pin 2 Ground
pin 3 Output

79xx
pin 1 ground
pin 2 input
pin 3 output

I don't know why they are manufactured different, but there must be some reason involving layout or something.

 

[bcarso]

"I think the pinout is different for the 79xx series. "

Yes definitely!

 

[maneco]

...from my experience,i strongly suggest that you use lm317 and lm337...they are less noisy,and can stand higher input voltages...there are many examples of bipolar supplies which use this pais,and they can even regulate for phantom power

good luck!

 

[chrissugar]

I agree with Maneco. I use the Lm317/337 series for all my projects, even for phantom power and they work great. No problems so far.

chrissugar

 

[bcarso]

"And nothing beats the erection you get from trimming your 15v to 15.0000000000000001 volts."

Wow Jaakko you must have a considerably bigger... errr... voltmeter than I do.

I think we both need to get out more too....

 

[CJ]

Ah ha! Thanks Mr. Bill!
Now I can get that VP in action!

So if you have the typical plus/minus circuit, and 37 volts plus and minus feeding the regs, and a faulty hookup, the 79xx probably sees 74 volts! Nice BB-Q!

:guinness:

 

[Emperor-TK]

Thanks guys for all the info. Getting a solid answer that I can understand is even better than trimming voltages to 10 decimals :grin: :thumb: .

There is only one part I don't understand:

[quote author="Svart"]However, in reality you HAVE to have proper decoupling and smoothing caps on the inputs and outputs or sometimes they oscillate and sometimes they are just plain noisy...[/quote]

So what is a decoupling cap in this example? I always thought that this term was another word for an interstage dc blocking cap. I guess that's wrong? The GSSL schematic has electros and film caps in parallel tying both the v-reg inputs and outputs to ground. I thought they were all for shunting AC and noise to ground. Is one of these a decoupling cap and the other a smoothing cap? And what's meant by "decoupling"?

While I'm at it, how do any of these differ from a "filter" cap?

Thanks again,
Chris

 

[CJ]

it's just that 0.1 that never does anything. I have never solved one problem with a 0.1. not in 30 yrs.

 

[clintrubber]

I have never solved one problem with a 0.1. not in 30 yrs.

Hooking up that Yngwie-type by 0.1 to some HT-AC would solve a lot though ! :twisted: :twisted: :twisted: :twisted: :twisted:

Oops, wrong thread.

 

[clintrubber]

Svart wrote:
However, in reality you HAVE to have proper decoupling and smoothing caps on the inputs and outputs or sometimes they oscillate and sometimes they are just plain noisy...


So what is a decoupling cap in this example?

I haven't looked it up, but I thought the 79XX needed a bit more decoupling than the 78XX. So while a 0.1 would keep the '78 silent, it wouldn't work well enough for the '79.

Depending on brands the '79 can be a nasty beast. I've used it less than '78's yet I've had more problem-occurences with it.

 

[BYacey]

The reason they oscillate is that the regulator is just an amplifier stage. It constantly monitors the output voltage and generates an error correction when it deviates from the reference. The cap on the output might be better called a "stabilizer" cap. It shunts the high frequency noise from the rapidly occuring correction signal that shows up on the output. You could limit the high frequency response of the regulator amplifier with feedback, but it wouldn't compensate fast enough to current load changes, and the voltage output would wobble around the reference creating more lower frequency noise.

 

[bcarso]

Chris wrote: "So what is a decoupling cap in this example? I always thought that this term was another word for an interstage dc blocking cap. I guess that's wrong?"

A decoupling cap is a power supply cap local to a section of a circuit, from the power rail to ground (or sometimes the opposite power supply rail). The purpose is to provide a low impedance for that circuit section, both to filter fluctuations out from upstream and reduce the variations of the rail due to currents of the circuit section. Often there will be a small R between the main power supply feed and the decoupling cap to form a lowpass filter. Sometimes the inductance and resistance of the trace/wire is enough. Once in a great while an inductor, most likely a ferrite bead, will be used with the resistor as well or by itself.

The reason it is called a decoupling cap: it reduces the coupling of one circuit section to another through the power supply, which is usually an unwanted and often unanticipated part of the "real-world" circuit, leading to misbehavior of the overall system. If you had a perfect power supply, that is a perfect voltage source distributed to each circuit section, you wouldn't need such networks. Many times with low frequency op amps etc., moderate to low gains, and not-too-long runs of trace/wire, you don't need them.

An interstage coupling cap is serving a very different role, as you say, blocking d.c. and low frequencies between stages in the signal path, while passing the desired signal.

 

[Emperor-TK]

Thanks! :green: :thumb:

 

[PRR]

> 78xx - pin 1 input Pin 2 Ground pin 3 Output
79xx pin 1 ground pin 2 input pin 3 output

I don't know why they are manufactured different, but there must be some reason involving layout or something.

For cooling, the most-negative part of the die is be soldered to the heatsink tab.

For historical reasons, the heatsink tab is always Pin 2 (the center pin, the one that was missing on the TO3 package that TO220 derives from).

On the 78xx positive regulator, "Gnd" is most negative.

On the 79xx negative regulator, "Input" is most negative.