Universial Ac Dc converter circuit for small projects

[Rutger]

Hi, it is true that I have no experience with switches power supplies, but I have experience with regular power supplies.
I understand the safety concerns, but I am going by the design examples of the chip, nothing exotic here.

The schematic are online at the following link:

http://www.powerint.com/sites/default/files/product-docs/lnk302_304-306.pdf

Rutger 

[Simon]

and are you sure that your circuit only needs 225mA ?

[Psi]

...I'm unclear as to why there are so many source pins but only 1 drain pin..

my guess would be the source pins provide the best connection to the substrate, so for heatsinking to the pcb they've brought out extra source pins.

[Simon]

Yes that makes sense. I thought initialy this was a more powerful IC and that paralleled pins would be required to cope with the amperage but for a couple of hundred milliamps it's explained

[ShiftPlusOne]

Why are those XP Power all-in-one power supplies so expensive when compared to AC adapters?

[deephaven]

Why are those XP Power all-in-one power supplies so expensive when compared to AC adapters?

Probably mostly due to the volume they are made in.

[Rutger]

I will build a prototype first of the actual level control circuit and measure the max current draw so that I can actually know which exact power supply I will need. Is there a rule of thumb to add an additional percentage to the outcome of the result, so lets say my max current is 150 mA, would you recommend at least 165mA (10%) for a supply?

[Simon]

yes that's a good idea that way you know just how much power you need. um +10% ? at lower powers I'd be inclided to add a generous 100mA but that's just me if it was say 165mA I'd at least round it up to 200mA

[ShiftPlusOne]

Why are those XP Power all-in-one power supplies so expensive when compared to AC adapters?

Probably mostly due to the volume they are made in.

Ok, thanks.

[Rutger]

This might be a silly suggestion but could I not just use a linear transformer that is rated from 220V to 20V 60hz, a rectifier etc and then use a MC7812 fixed voltage chip? In the 220V line voltage case it works with a higher voltage drop across the 7812 and in a 110V supply it should have enough input voltage. My application doesn't need very much power so I don't think a need a heat sink for the 7812.
According the datasheet the 7812 can take up to 35V input.

Rutger

[Simon]

it's possible, I'm not sure of how a transformer copes with different voltage ranges. But I'm sure you can get similar power transformers of a similar size with different in/out voltages so all you need to do is change the transformer for different areas, maybe keep it in 2-3 main ranges

[deephaven]

This might be a silly suggestion but could I not just use a linear transformer that is rated from 220V to 20V 60hz, a rectifier etc and then use a MC7812 fixed voltage chip? In the 220V line voltage case it works with a higher voltage drop across the 7812 and in a 110V supply it should have enough input voltage. My application doesn't need very much power so I don't think a need a heat sink for the 7812.
According the datasheet the 7812 can take up to 35V input.

Rutger

Unless you did what Simon suggested, what you're suggesting isn't practical. If you are designing it for 0.5A, and you are using a linear regulator, you will nead a heatsink even if it's for a single supply voltage.

If you want to do it yourself, I would suggest a mains transformer with a tapped 110/220 volt primary and use (for example) a National Semiconductor "Simple Switcher" IC which probably won't need a heat sink if your layout is carefully designed. Their website has a good interactive utility where it will design you a suitable circuit and give you heat analysis etc.

[Zero999]

Why are those XP Power all-in-one power supplies so expensive when compared to AC adapters?

They're better much better quality: lower ripple, higher efficiency, better power factor, hight MTBF, wider temperature range and how many cheap AC adaptors come with a three year warranty?

[Rutger]

Unless you did what Simon suggested, what you're suggesting isn't practical. If you are designing it for 0.5A, and you are using a linear regulator, you will nead a heatsink even if it's for a single supply voltage.

If you want to do it yourself, I would suggest a mains transformer with a tapped 110/220 volt primary and use (for example) a National Semiconductor "Simple Switcher" IC which probably won't need a heat sink if your layout is carefully designed. Their website has a good interactive utility where it will design you a suitable circuit and give you heat analysis etc.

Thanks deephaven for the tip, the National Semiconductor IC & website is much more helpfull than the powerint website, the powerint have software you can download but the solution uses exotic part(s) that are hard or impossible to get. I like the National Semiconductor much better and their web interface to design a circuit is really cool.

Rutger

[saturation]

Have you thought about just reclaiming a lot of discarded cellphone chargers and use those PSUs?  Many are in the trash, I have buckets of them in reserve for miscellaneous projects, and most are free [ I use a lot of trash electronics as a parts source.] 

They are switchers, compact, small and can deliver your rated amperage.  However, most are from 100Vac-240Vac.  FYI, by power quality definitions, AC can only deviate +/- 10% before line power is defined as unreliable, in fact most AC power line regulators will shut down if AC drops below 100V or exceed 260V if your official line voltage is 110V to 240V.

I though I did, but here is my specific spec:

Input: 85V - 260V AC
Output: 12V DC, 0.5 A

Rutger

[Rutger]

Thanks saturation, no I didn't think of that, that is a great idea!

I can live with the 100V-240V restriction, but I would like to get a lot of the same ones, do you know of a source for these units like a recycle place here in the US?

Rutger

[Zero999]

Here's the kind of thing I'm talking about.

Phihong PSM11R-120-R
Input 90V to 264VAC
Output 12V 840mA
Comes with a selection of plugs.

Link to RS:
http://uk.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=0234193#header


You can get similar mains adaptors for more (higher quality) or much cheaper (ebay or buying direct form China).

[saturation]

I know a number of recycling locations for drop off, but I don't know if you can take some from their collection boxes.  Often, your friends or family will have bunches of these, often ending in the trash.  If you put a notice out in your school or work, you could have a lifetime's worth is a few days.

One source I use are no-name Chinese adapters, I get them for $1-2, shipping included.  You can cannibalize the electronics and embellish it, rather than building a small switcher from scratch.  Many of these units are screwed together rather than glued, making it easier to gut.  Here's just an example:

http://cgi.ebay.com/New-AC-DC-Wall-Adapter-Charger-Plug-110-240V-t-12V-9989-/270583093649?pt=Other_MP3_Player_Accessories&hash=item3f00025991

I often can find them still cheaper than this, just make sure the Vin and Vout and amps are what you need.

Thanks saturation, no I didn't think of that, that is a great idea!

I can live with the 100V-240V restriction, but I would like to get a lot of the same ones, do you know of a source for these units like a recycle place here in the US?

Rutger

[ShiftPlusOne]

Another noob question... what's wrong with just a simple transformer -> full wave rectifier -> capacitor-> voltage regulator circuit?

[Zero999]

Another noob question... what's wrong with just a simple transformer -> full wave rectifier -> capacitor-> voltage regulator circuit?

Apart from being heavier and less efficient than an SMPS, the voltage needs to be manually set by flipping a switch or configuring jumpers. An SMPS can run from a wide range of voltages and give a constant output voltage.

[ShiftPlusOne]

Another noob question... what's wrong with just a simple transformer -> full wave rectifier -> capacitor-> voltage regulator circuit?

Apart from being heavier and less efficient than an SMPS, the voltage needs to be manually set by flipping a switch or configuring jumpers. An SMPS can run from a wide range of voltages and give a constant output voltage.

Ok, thanks. Looks like I have some learning to do.

[saturation]

Linear supplies can be adjusted with external controls, e.g. using an adjustable linear regulator, they still make lab grade linear supplies because they do have cleaner output.

But the real reason linear PSU are used less is because they are inefficient, and thus, waste power in many ways.  This adds up, and there is a worldwide movement to reduce the quiescent drain of consumer PSUs.  SMPS are noisy but efficient, and many devices are designed to suppress or can live with the noise.  Linear supply losses occur from the small quiescent current leaked by the regulator, heating from various sense resistors and eddy currents in transformers etc., so even when not used, linear supplies draw a higher quiescent current when unused, but powered ON, so current flows in the primary coils.

Another noob question... what's wrong with just a simple transformer -> full wave rectifier -> capacitor-> voltage regulator circuit?

Apart from being heavier and less efficient than an SMPS, the voltage needs to be manually set by flipping a switch or configuring jumpers. An SMPS can run from a wide range of voltages and give a constant output voltage.

[ShiftPlusOne]

But for a uni project or diy bench equipment that's not really a problem though, right?

Also, I have a question about safety when working with 240V. At which point does it become 'safe' on the circuit. Simply going through a transformer will still leave a potentially high current source, right? So is it only after the regulator that it's safe? In general, what sort of precautions should I follow in terms of design? I've been staying away from 240V, but I'd like to start work on something that doesn't require a wall wort.

I hope I am not hijacking the thread >.>

[saturation]

Yes, a small linear is easier to build for uni use and that its still sold to labs because its reduced noise, relative to SMPS, is hard to beat.  

If you design power supplies, you'll be exposed to both high and low voltage, at the primary and secondary side of the transformer.  

To simplify things, you need to pay a great deal of precaution when voltage exceed 30Vdc or AC peak.  The definition varies, but the low end is debatable, some feel its as low as 6Vrms in humid conditions.  Some DMM will flash a warning sign when the voltage being measured exceeds the 'low voltage' area.

http://en.wikipedia.org/wiki/Low_voltage

http://en.wikipedia.org/wiki/Electrical_shock

But for a uni project or diy bench equipment that's not really a problem though, right?

Also, I have a question about safety when working with 240V. At which point does it become 'safe' on the circuit. Simply going through a transformer will still leave a potentially high current source, right? So is it only after the regulator that it's safe? In general, what sort of precautions should I follow in terms of design? I've been staying away from 240V, but I'd like to start work on something that doesn't require a wall wort.

I hope I am not hijacking the thread >.>

[ShiftPlusOne]

Yes, a small linear is easier to build for uni use and that its still sold to labs because its reduced noise, relative to SMPS, is hard to beat. 

If you design power supplies, you'll be exposed to both high and low voltage, at the primary and secondary side of the transformer. 

To simplify things, you need to pay a great deal of precaution when voltage exceed 30Vdc or AC peak to peak.  The definition varies, but the low end is debatable, some feel its as low as 6Vrms in humid conditions.  Some DMM will flash a warning sign when the voltage being measured exceeds the 'low voltage' area.

http://en.wikipedia.org/wiki/Low_voltage

http://en.wikipedia.org/wiki/Electrical_shock

Correct me if I am wrong here...
Based on the Wikipedia article, if you have a 240V source attached to a 1/20 transformer, the LV side of the transformer is actually safe to touch since the current would be about 6-12mA which is enough to feel, but not enough to kill. So, what you really have to worry about is just the live mains wires and the possibility of shorting them to something you might touch, but anything after the voltage has been stepped down is actually safe?