LM7815 power supply question

[J M Fahey]

Wow, didn't think it would drop 15%. Is that a smooth drop off, or a hockey stick thing that goes down quickly near the limit?
EDIT: Oops, sorry, you wrote clearly above "loss is proportional to I squared".

Yes, itīs a smooth drop.
Commercial Amp designers know this of course, since forever; Booteeq ones or homemakers often do not and either write crap about the transformers they get OR overrate specs 2x or 3x above whatīs actually needed.

Most ampmkanufacturers donīt write supply or rail voltages any more, just check it yourself ... except +/- 15V which is regulated, of course.

Tube amps often do not change that much, but only because at idle, no signal, they are *already* pulling a lot: filaments and too much idle current through power tubes, plus low efficiency all over the place, but SS amps which pull feeble current at idle, drop rails a lot when actually played loud.

VERY FEW Commercial amps tell the whole story on schematics, (which I find VERY annoying and unprofessional) , only Roland (and maybe scant others) tell all.

Just check this Roland KB100 posted here a couple days ago:


a) rail voltage: +47 to +54V , meaning: +54V at idle, dropping to +47V at full power.

b) something which causes MUCH confusion, it took me a long time to decode, and had to check a ton of Roland schematics to self confirm it was so after all: transformer primary rating is AC (mains) **BUT** in Rolandspeak secondary voltage is NOT RATED IN VAC but in "DC rectified voltage achieved with that transformer" .

A normal tech and even more a somewhat advanced end user will see that schematic and, if neded, will buy a 47VAC+47VAC transformer ... and blow everything in its path.

That label must be read as: "transformer whose rectified secondary gives you +/-49 Vdc under a 1.7A load" which "will swing from 54V at idle down to 47V under full load and still be under spec".

Which looks complex but in fact is a very accurate way to fully describe the transformer which will fit there.

LAB Series amps, "by Engineers for Engineers", usually stated transformer secondary VAC but stated two rail voltages, for example: "+54 (+49)V" meaning: "+54 at idle, +49 full load"

Others just post nominal rail voltage and "forget" to mention it will drop a lot.

And as I said before, MANY plain do not post any rail voltages, period.

[nickb]

Take manufacturerīs regulation with a pinch of salt.

The "number" they quote is for AC regulation into resistive loads (and maybe it holds for choke input supplies) , but capacitor input ones get charged by narrow pulses with quite high peak current and narrow duty cycle.

Since loss is proportional to I *squared*, 1A 50% of the time requires 2A 25% of the time just to keep *average* current the same, but RMS current will be twice as high, with matching doubled loss.
And even worse with narrower pulses.

My point being that rectified voltage drop is higher than expected.
I have used overbuilt 5% regulation transformers where DC under full load dropped some 15%

The point of knowing the regulation (and it's such a basic and important parameter that I can't believe they didn't include it) is I can calculate the effective resistance and from that I can accurately allow for the high current pulses and determine the min output voltage. It's a bit more complicated than that, but I think we said that way back...

[mikepukmel]

As an aside, how do you test a power supply: i.e. OK I know how to test no load, just ... don't hook it up to any load. But is there a reasonable way to test various current draws, then check the voltage? Some kind of resistor network?

[nosaj]

As an aside, how do you test a power supply: i.e. OK I know how to test no load, just ... don't hook it up to any load. But is there a reasonable way to test various current draws, then check the voltage? Some kind of resistor network?

Enzo was telling me he used to have a big (I think 1 ohm resistor) in his drawer and he would use that to figure current draw. Kinda of like the 1 ohm resistor people will use on power tubes.

nosaj

[mikepukmel]

This will probably be shocking to you all, and I hope I didn't ruin anyone's day/evening, but these little regulator board kits I got for 1.60 USD each, have the + and - outputs labelled reversed. And the heat sink looks, rather small. Will post photos. The sinks are about the height of the installed regulator.

I got these kits mainly for the boards, thinking some of the components might need to be updated.

[Enzo]

I used to repair piles of SMPS for the arcade video game industry. Black boxes with +5, +12, and -5 outputs +12 at an amp or two, +5 at anaything from 5A to 20A. -5 was usually an amp.

First test it unloaded, then indeed I had some resistors for loads. They were mostly ceramic cylindrical ones. I had a 5 ohm one - to draw 1A from 5v. I had a 1 ohm one to draw 5A from 5v. I had a 12 ohm one for.. oh you can guess.

Plus I had a whole selection of power resistors in my stock, I could grab.

Put an appropriate resistor across the output and see if the voltage holds up. If I were in a lab setting I could use manay load values and chart the complete performance, but that is rarely necessary here. If the thing works at essentially full out, and at a gentle load, chances are real good it will also work at points between.

[mikepukmel]

I used to repair piles of SMPS for the arcade video game industry. Black boxes with +5, +12, and -5 outputs +12 at an amp or two, +5 at anaything from 5A to 20A. -5 was usually an amp.

First test it unloaded, then indeed I had some resistors for loads. They were mostly ceramic cylindrical ones. I had a 5 ohm one - to draw 1A from 5v. I had a 1 ohm one to draw 5A from 5v. I had a 12 ohm one for.. oh you can guess.

Plus I had a whole selection of power resistors in my stock, I could grab.

Put an appropriate resistor across the output and see if the voltage holds up. If I were in a lab setting I could use manay load values and chart the complete performance, but that is rarely necessary here. If the thing works at essentially full out, and at a gentle load, chances are real good it will also work at points between.

Cool, thanks Enzo. I will give these cheapies a run at 1/2a and 3/4a before going full bore with the programming board.

[jmaf]

As an aside, how do you test a power supply: i.e. OK I know how to test no load, just ... don't hook it up to any load. But is there a reasonable way to test various current draws, then check the voltage? Some kind of resistor network?

Adding to what Enzo already replied, this guy has a cool Youtube channel and he's done a test on Chinese batteries: http://www.instructables.com/id/DIY-...-current-load/

Similar gadget can be used to test the power supply, just place it where the battery would be and build the test rig to match your supply voltage (15V?).

[mikepukmel]

Adding to what Enzo already replied, this guy has a cool Youtube channel and he's done a test on Chinese batteries: http://www.instructables.com/id/DIY-...-current-load/

Similar gadget can be used to test the power supply, just place it where the battery would be and build the test rig to match your supply voltage (15V?).

Thanks for the link Jmaf! (Just read the page, cool project, I could build that!)

[mikepukmel]

Best Epoxy to glue heat sink to circuit board

Hi All!

So I got these way cheap kits (very cheap), came with all the parts. The one on the left is almost done, but has the smallest heat sink I've ever seen. I dug around on Mouser, and found better ones that I thought would fit ok.




The pins on these bigger heat sinks don't line up with the holes that the board came with. No surprise there. An hour with my hand drill and I got the bigger sinks into the board reasonably well for stone knives and bear skin tools I have. One fits reasonably tight. One fits a little loose. Either way, Id like to put some epoxy on the pins under the board.

I started reading about epoxy, found some high rolling stuff, but the links only have "Get a quote" so they want me to buy gallons of the stuff. The opposite end of the spectrum is the cr@p available at the big box stores. <frown>.

Can anyone suggest some kind of epoxy I can use: put a drop on the pins on the bottom of the board, that won't emit something corrosive like the home fixit epoxy and silicone?

Its not going into a high vibration environment, but I will bolt the board and transformer into a box and move it around, so I'd hate to have the thing loosen up the solder connections on the LMXX.

Thanks!

[jmaf]

Hi All!

So I got these way cheap kits (very cheap), came with all the parts. The one on the left is almost done, but has the smallest heat sink I've ever seen. I dug around on Mouser, and found better ones that I thought would fit ok.


[ATTACH=CONFIG]49514[/ATTACH]

The pins on these bigger heat sinks don't line up with the holes that the board came with. No surprise there. An hour with my hand drill and I got the bigger sinks into the board reasonably well for stone knives and bear skin tools I have. One fits reasonably tight. One fits a little loose. Either way, Id like to put some epoxy on the pins under the board.

I started reading about epoxy, found some high rolling stuff, but the links only have "Get a quote" so they want me to buy gallons of the stuff. The opposite end of the spectrum is the cr@p available at the big box stores. <frown>.

Can anyone suggest some kind of epoxy I can use: put a drop on the pins on the bottom of the board, that won't emit something corrosive like the home fixit epoxy and silicone?

Its not going into a high vibration environment, but I will bolt the board and transformer into a box and move it around, so I'd hate to have the thing loosen up the solder connections on the LMXX.

Thanks!

IMO the best solution would not be epoxy or any kind of filler.

Maybe you could bore a hole into the heatsink and affix it with a screw or pop rivet through the board.

[Enzo]

Amen, I agree.

[mikepukmel]

Well, I got holes in the board that let the bigger pins come through. I gota few extra heat sinks, could pull off the pins from one and try to put a screw in the same place. The pins look like they're friction fit into the fins, so a similar size screw could work.

Thanks! !! I'll try that! And I was dreading fooling around with messy epoxy.

[jmaf]

Well, I got holes in the board that let the bigger pins come through. I gota few extra heat sinks, could pull off the pins from one and try to put a screw in the same place. The pins look like they're friction fit into the fins, so a similar size screw could work.

Thanks! !! I'll try that! And I was dreading fooling around with messy epoxy.

Those heatsinks are aluminum and the pins are tin, so they can't be soldered together. They're usually rivetted. To remove the pins just drill a hole into the pop rivet.

[mikepukmel]

Thanks Jmaf. I finally got some time today, and the pins are press fit into the bottom of the sink. its not a closed hole, but more than "U" shape, so that it fits around the screw or pin slightly so it won't slide out. So I could pill the pin out with a vice grip, (its got a little knurling to make it stick) and use a screw in place. I'll try to get a photo, its tough to focus in there. Pretty cool design, I think. Have to trek out to the hardware store tomorrow to see what screws I can find. Hope to get at least a couple of these running tomorrow.