Need a 40V input rated version of the lm7815 regulator. 1

[PoyntExH]

Hello all,

I am working on a power supply board which has produces 15V, and 24V outputs. The transformer we're using outputs roughly 36-37VDC unfiltered which gets sent into a lm7815 regulator to make the 15V output, we can't use a transformer with lower output because we need the higher voltage to feed the buck regulator which outputs 24V. We're exceeding the 35V limit for the lm7815 regulator (it hasn't died yet) and it would be nice if I could find a fixed 15V output 3 terminal regulator, either in the to-220 packadge or something with three terminals spaced 100 mils apart. I need one that is rated for 40VDC or higher.

So far all I could find is the Fairchild KA78T15 regulator, which WOULD be perfect, however from what I've been able to tell its an obsolete part and only Rochester Electronics is selling those parts and this board is expected to last 10 years, so we don't want to be limited to one supplier.

Does anyone know of a 15V fixed output 3 terminal regulator that can handle up to 40V or more input voltage, preferably in the to-220 package? Current requirements aren't too high, 500mA would probably suffice.

I have seen a high voltage version of the LM317 that can do up to 60Vin, I may try to use that if I have to, but we're trying not to have to modify this board any more to fit the extra components required for an adjustable regulator.

Does anyone have any ideas? We may be able to use a 3 pin buck regulator chip, provided it has a reasonably clean output.

-Robby

 

[itsmoked]

Good luck. I think you're SOL. Best bet is the LM317.

Sounds like you might end up with a smoking crater in the board anyway. Dumping 50% of the raw voltage on the regulator will usually try to roast things.

If your 15V load current is fairly fixed you can proceed the regulator with a power resistor to skim off half the voltage on a 'dumb' part instead of toasting the semiconductor regulator. That will remove voltage that would otherwise be 'seen' be the regulator. You could then stick with the LM7815.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Skogsgurra]

I have saved my bacon (from being fried) by using a power zener in series with the input. Not very elegant but it works and it moves the problem from finding exotic 40 V parts to handling the heat produced in the zener. A string of 1N400x can also be used. They do not get hot.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[ScottyUK]

Can you feed the 15v and 5V regulators from the 24V regulator output? It makes the 24V regulator work harder but gets around the immediate problem. I don't know how much spare capacity is on the 24V rail.

 

[IRstuff]

I'm perplexed why you might use a buck converter for the 24V, but not the 15. Something like

http://www.maximintegrated.com/en/products/power/switching-regulators/MAX17521.html

could supply both 15V and 24V in a single package, assuming they were both less than 1A.

TTFN
faq731-376
Need help writing a question or understanding a reply? forum1529


Of course I can. I can do anything. I can do absolutely anything. I'm an expert!
There is a homework forum hosted by engineering.com:

http://www.engineering.com/AskForum/aff/32.aspx

 

[PoyntExH]

I'll have to check to see if we can run the 24V buck into the 15V reg. I am hesitant to do this since it may produce unwanted transients into the circuitry that runs off the 15V rail.

We were thinking of using a 30V zener diode to bring the voltage back down, though it would be nice if we didn't need to. We shall find out tomorrow!

 

[MacGyverS2000]

I'm with Gunnar... i hate the idea of wasted power, but a few diodes in series to drop the voltage will solve your problem (albeit in a eye-twitch-inducing manner to any engineer). With that much drop (and assuming more than a few tens of milliamps in current), I would use a buck converter and save a ton of power.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[zappedagain]

Are you avoiding the switcher due to noise? I've had good luck with a Texas Instruments TPS7A4901DGNR driving my analog circuits from a switcher. I'm going to reqire a lot more SNR before I'll move back to a linear supply.

Z

 

[itsmoked]

A TPS7A4901DGNR is a linear reg isn't it?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Comcokid]

I would just run the 7815 off the output of the buck. Buck noise on the 15V output would be eliminate enough by the regulator, unless you are powering super-low-noise PPL or LNA devices from the 15V.

Otherwise, use a zener as suggested, or even put another 7815 with a 15V zener in the ground leg (makes for a 30 volt regulator) ahead of the 7815 for the 15V. The first 7815 with a ~30V output will drive people unfamiliar with the circuit crazy.

I had a similar situation for a simple low current supply running from 28VDC down to 12 V which needed to withstand 2 msec transients up to 88V, so I just put a TL783 (100V rating) regulating down to 24 volts, and then a 7812. The TL783 runs out-of-regulation sometimes with insufficient input to output headroom when the 28V input is at absolute minimun but it has done the job for several years in a product. Could have also used a zener regulated pass transistor to drop the voltage as well.

 

[zappedagain]

Keith,

The TPS7A4901DGNR is a low drop out linear reg with very good PSRR across a wide bandwidth. it is designed to be put on the output of a switcher to knock down the switcher noise. I'm using one on a TIA with over 120 dB of gain and power supply noise is not an issue.

Z

 

[PoyntExH]

I had forgotten when I originally posted this that the 24V rail gets turned on separately since its for a completely different circuit and can't be used with the 7815.

The TPS7A4901DGNR that I found on digikey can only handle up to 35V and since this is a through-hole board it would be rather awkward to jumper in the 8 pin SMT package.

I pulled the 15V regulator out of the board on order to use a separate 15V lab power supply to measure the current draw. The circuit draws 200mA at max load. Also I tested a few of our zener's in stock and the best onewe had was only 1.3W at like 3.3W which just wouldn't cut it.

What we did instead was we found a drop in replacement buck regulator on digikey:

http://www.recom-power.com/pdf/Innoline/R-78HBxx-0.5_L.pdf

Now I'm currently in the process of probing the board and trying to catch any issue that may be caused by the switching transients on the power rails.

 

[itsmoked]

Thanks for the lesson zapped!

And Rob thanks for showing us that absurd part by an unknown company! I shall put that PDF in my "interesting parts" directory. It could come in handy someday.


Keith Cress
kcress -

http://www.flaminsystems.com

 

[PoyntExH]

Hahaha, yeah goodness I didn't know such things existed either! My boss was bound determined to find something. At first he had a part in mind that he used back at college - it was a similar buck regulator, but its switching transients were so bad that it required an awful LC filter on the output to keep it stable and useable....neither of us liked that thought....so luckily he managed to find this handy little guy instead!

I'll be taking a better look at the switching transients on the rails tomorrow to see if they're causing me any issues. So far I'm really enjoying the 96% efficiency and not needing a heatsink.

 

[roydm]

How about two 7815s in series. the first one with the adj pin tied to its output would drop from 40 - 20, the second from 20 to 15.
No doubt you would need some resistors in series to balance the Voltage.

 

[itsmoked]

Hi roydm; a missed point here by a lot of folks was "drop-in replacement so no board cycle would be needed".

Keith Cress
kcress -

http://www.flaminsystems.com

 

[JG2828]

Those DC-DC converters have been around for years and are amazing. You don't need a heatsink and they are fully protected, so even if you short the output they won't burn up. I used a couple of them in a doorbell camera project I did a few years back. 5v to power the Raspberry Pi and 3.3v to drive the GPIO pin. Project came out great and is still running strong.