LM7815 power supply question

[jmaf]

Why produce heat? Instead, add a switcher between the 24VAC output and the rectifier:



Adjust it to switch at around 20 VAC peak.

[mikepukmel]

My gut tells be that might be too low in general. But we'll see. I can't answer the heatsink question w/o knowing a bunch of other things.
The first question is what is your required max load current?
Who is your preferred transformer vendor (so I can look up the specs)?

Thanks Nick. After some toil, I found these: https://www.mouser.com/ProductDetail/546-187E16 16v 56va, 3.5a, a little bigger than needed, I think, for a 1a PS. Allows for some voltage drop and still stay above 17v min input.

[mikepukmel]

When I'm sizing up for small PSUs using 78XX regulators I get the closest VA rating (higher rather than lower) to my maximum load and the closest AC voltage to my output voltage. Assuming 10% regulation this usually works out fine for dropout voltage. So for a 15V regulator I'd go for a 16v transformer. Usually the rectified output is a little higher than calculated due to primaries being wound for 2x115v or 1x230v and the mains being higher (mine is 248v today).

Thanks Mick, that's what I did (more or less by accident or trial and error), 16v and a little higher VA than needed 56va. Mains are a little higher here but Im not quite sure what that means all the way at the input of the 7815, Id have to measure.

[mikepukmel]

The secondary current will be much higher (about 1.7x) than the output current plus the secondary voltage is higher than the output meaning the VA rating needs to be at least twice the output power (as a starting guess). For such a transformer the regulation is be around 25%.

Here are my estimates for a 16V 32VA transformer.

[ATTACH=CONFIG]49404[/ATTACH]

Plugging the numbers into a simulation to be more precise I get,
Vo min = 18.3V this is above the 17V 7815 requirement, OK
Vo max = 28.2V this is below the 35V 7815 requirement, OK
Pd = 7.9 Watts dissipated in 7815
=> Heatsink better than 7.8 C/W, so yes best use thermal compound.
Secondary current= 1.64A rms
Min VA rating - 1.64 x 21.5 = 35.3 VA ( very worst case of continuous max load and max input voltage).

Cool, thanks for the calc's. As far as I can tell: the transformer should never go to full load, its a 3.5a rated, and the board should not ever draw more than about 1a. So, I don't think the supply will ever get to that point. The board manual only says "10 - 15V DC power supply, 500 mA min." I believe that is to program and run one chip, without having to drive any additional hardware, i.e. only a few LED's on board., outside the boards power supply.

I got the duncan software, but spending some effort to figure out how to use it.

[mikepukmel]

Why produce heat? Instead, add a switcher between the 24VAC output and the rectifier:

[ATTACH=CONFIG]49405[/ATTACH]

Adjust it to switch at around 20 VAC peak.

Thanks JMAF! Reading up . . .

[mikepukmel]

My gut tells be that might be too low in general. But we'll see. I can't answer the heatsink question w/o knowing a bunch of other things.
The first question is what is your required max load current?
Who is your preferred transformer vendor (so I can look up the specs)?

Nick, sorry, missed your question on first read: The board manual says 10-15v DC min 500ma. I think it doesn't state a max since you could be programming a microcontroller chip to drives other stuff, so I think they assume the user knows something and will calculate that.

[mikepukmel]

A more general question about these small regulated power supplies: I searched and found a bunch of different but very similar circuits for LM7815 regulated power supplies. It looks like, or at least what I found, most have an electrolytic cap on either side of the regulator, one on the input side and one on the output side. Say 470uf on the input side, 100uf on the output side. A few had a much bigger cap on the input side, like 2200uf and none on the output side.

Do both of these caps serve the same purpose: if the current draw increases rapidly by the load, then the cap will make up the difference until either the LM7815 or the transformer + rectifiers, have time to compensate?

If we know the max output is, say 1a, it seems like we also need to know the rate at which the load could change current draw to know how big capacitors to use?

[jmaf]

A more general question about these small regulated power supplies: I searched and found a bunch of different but very similar circuits for LM7815 regulated power supplies. It looks like, or at least what I found, most have an electrolytic cap on either side of the regulator, one on the input side and one on the output side. Say 470uf on the input side, 100uf on the output side. A few had a much bigger cap on the input side, like 2200uf and none on the output side.

Do both of these caps serve the same purpose: if the current draw increases rapidly by the load, then the cap will make up the difference until either the LM7815 or the transformer + rectifiers, have time to compensate?

If we know the max output is, say 1a, it seems like we also need to know the rate at which the load could change current draw to know how big capacitors to use?

The regulators can oscillate in some circuits. These caps are mostly meant to clamp oscillation, in precision circuits you usually place the cap very near the regulator and very near the load, with short copper tracks from the cap to the regulator and to the load.

Your concern with the max current is not the caps, forget the capacitors, the regulator needs no capacitor to maintain constant voltage, ripple rejection is absurd as long as your source can keep a voltage above 15V RMS + a offset (see the datasheet).

The problem is power dissipation, this is a passive regulator, it turns excess power into heat.

The 7815 can source max 1A but not continuously, all the LM78xx's fry if you do 1A 100% duty cycle.

If you have too large a voltage offset like you do in your project you may be better off with the switching version of the LM regulators. You didn't say what your application is, so gotta observe that the switching ones are noisy and may require special shielding and extra filtering.

[Mick Bailey]

Why produce heat? Instead, add a switcher between the 24VAC output and the rectifier:

[ATTACH=CONFIG]49405[/ATTACH]

Adjust it to switch at around 20 VAC peak.

I may be wrong, but I would have though that an audio circuit powered this way would be subject to switching noise - like having a light dimmer in the same enclosure.

[J M Fahey]

Why produce heat? Instead, add a switcher between the 24VAC output and the rectifier:

[ATTACH=CONFIG]49405[/ATTACH]

Adjust it to switch at around 20 VAC peak.

The OP might need a switcher but *definitely* not that one, which can only handle *resistive* loads (itīs the classic cheap lamp dimmer) or universal series wound motors, both of which integrate that horrible output waveform (because of thermal or mechanical mass) into something useful.
--------------------------------

The board manual only says "10 - 15V DC power supply, 500 mA min."

Then it can be fed by a 7815.

You need at least 3V more than +15V feeding the 7815, including ripple, so aim at some 20/22V DC RAW to account both for ripple and drop under load.

If you can get a 15/16VAC transformer, rated 1 to 1.5A, it will be perfect, so you donīt need more than, say, 22/25VA .

All of this is very conservative, the board may not even take 500mA by a large margin, if a "500mA supply" was suggested.

[jmaf]

The OP might need a switcher but *definitely* not that one, which can only handle *resistive* loads (itīs the classic cheap lamp dimmer) or universal series wound motors, both of which integrate that horrible output waveform (because of thermal or mechanical mass) into something useful.

I wouldn't feed the output of that to motors or other inductors. It can definitely be rectified and regulated, no problem.

[nickb]

The 7815 can source max 1A but not continuously, all the LM78xx's fry if you do 1A 100% duty cycle.

If you have too large

I really don't think that is right. If the thermal issues are managed properly i.e. you don't hit the internal thermal or safe are operating limits limit then it will be fine. Some versions will even do 1.5A. The data sheet I'm looking at specifies a max operating junction temperature of 150C and a thermal resistance of 5C/W junction to case. In my calcs I used 125C as max and determined the heatsink size using that for continuous 1A. BTW I see I used 3.5C/W for chip plus interface - that should be 5.5C/W making the heatsink 7.2C/W.

All that will be conservative if the max load is 500mA anyway.

Mikep, the best resource for reference circuits is the data sheet. The National LM7815 shows 0.1uf on the output and 0.33uF on the input. It states the input one is only required if the regulator is located an "appreciable distance" (whatever that is) from the main filter cap and the the output one improves stability and transient response. Just use both and locate close to the regulator.

[jmaf]

I really don't think that is right. If the thermal issues are managed properly i.e. you don't hit the internal thermal or safe are operating limits limit then it will be fine. Some versions will even do 1.5A. The data sheet I'm looking at specifies a max operating junction temperature of 150C and a thermal resistance of 5C/W junction to case. In my calcs I used 125C as max and determined the heatsink size using that for continuous 1A. BTW I see I used 3.5C/W for chip plus interface - that should be 5.5C/W making the heatsink 7.2C/W.

I'm not giving advice based on datasheets. In fact I haven't opened a single PDF today.

[nickb]

You may have experienced a non-representative sample at some point.

We can test the hypothesis easily enough. Here we have have a L7815CV with 25V in and 1A out. It's been running happily for hours.

I don't have an LM7815 here but I'm sure you'll agree it would dissipate the same amount of power so we should expect the same result.

I do have an LM7805 if you'd like me to test that.

[jmaf]

Here we have have a L7815CV with 25V in and 1A out. It's been running happily for hours.

I don't have an LM7815 here but I'm sure you'll agree it would dissipate the same amount of power so we should expect the same result.

Since you're choosing more powerful components, why not go with one of these just to shoot the shit?



I'm sure you'll agree it would dissipate the same amount of power so we should expect the same result.