USB device interfering AM radio

  • Thread starter Thread starter Man-wai Chang
  • Start date Start date
M

Man-wai Chang

I just discovered this. And I also found that the AC to USB power
adaptor does not have a metal ground pin.

So would this AM interference be gone after getting a better AC to USB
power adaptor?

--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.38
^ ^ 20:58:01 up 2 days 32 min 0 users load average: 0.00 0.01 0.05
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa
 
Man-wai Chang said:
I just discovered this. And I also found that the AC to USB power
adaptor does not have a metal ground pin.

So would this AM interference be gone after getting a better AC to USB
power adaptor?

You can't have too many grounds present. Or there would be a ground loop.

If a USB item plugs to the computer, the computer provides the ground path.
If the USB item has its own wall adapter, that wall adapter "floats", so
that the only ground path is from the computer.

safety_ground ====== computer ===== USB Peripheral ---- floating AC adapter

That is how my external 3.5" USB disk drive works. It has a +12V adapter
which plugs into the back. There is no safety ground connection, as there is
no third prong on the AC adapter.

But when the external 3.5" USB disk drive is plugged into the computer,
the ground connection from the computer is being used to establish ground
for the setup.

If the adapter is bad, try another. I think that is where I might start.

I have an Antec power supply, that produces unacceptable interference, so
it is possible for power supplies to have that problem. Examining the
power supply design, it did have a filter on the primary side, but
apparently whatever is wrong with the supply, the emissions are just
too strong for the filter to stop them. That means some component in
the supply is defective. The thing is, it can still power a computer
just fine. But the conducted interference it creates is so strong, it
causes my ADSL modem to lose sync.

Paul
 
Man-wai Chang wrote:
Subject: USB device interfering AM radio

What is the 'USB device' which also has a AC adapter?

Are you talking about an AC device that serves 'purely' as a USB-type
power source like this
http://www.thinkgeek.com/gadgets/travelpower/9124/ Adapter turns any
outlet into a USB outlet

The USB devices I have seen which also have an AC adapter are some
device such as a laptop cooling base with fan which also serves as a USB
hub, so the purpose of the adapter is to provide the powered hub,
optionally unpowered or 'bus powered'.
I just discovered this. And I also found that the AC to USB power
adaptor does not have a metal ground pin.

Do you mean a ground for the AC transformer end? The DC end of a power
adapter is a 2 contact pin including ground. Such AC transformers rarely
have a 3 pronged AC ground end.
So would this AM interference be gone after getting a better AC to USB
power adaptor?

We don't have your complete story yet.

How did you determine that the unknown device is causing the
interference? What if you somehow load the AC adapter's DC end without
the unknown USB DC 'device' being active? Load it with another different
'device'.
 
the supply is defective. The thing is, it can still power a computer
just fine. But the conducted interference it creates is so strong, it
causes my ADSL modem to lose sync.

Never heard of this...

--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.38
^ ^ 15:02:01 up 2 days 18:36 0 users load average: 1.07 1.10 1.12
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa
 
Are you talking about an AC device that serves 'purely' as a USB-type
power source like this
http://www.thinkgeek.com/gadgets/travelpower/9124/ Adapter turns any
outlet into a USB outlet
Do you mean a ground for the AC transformer end? The DC
end of a power adapter is a 2 contact pin including ground.
Such AC transformers rarely have a 3 pronged AC ground end.


Yes. But the one I got is of UK 13Amp standard. The ground pin of the
adaptor is plastic, NOT metal. So the shielding in the USB cable is NOT
working, I suppose.
How did you determine that the unknown device is causing the
interference? What if you somehow load the AC adapter's DC end without
the unknown USB DC 'device' being active? Load it with another different
'device'.

When I unplugged the AC to USB power adaptor, the noise was gone.

The full circuit:

220V AC
---> AC-to-USB power adaptor (UK)
---> 3-feet USB cable
---> 2x100ohm resistors
---> 2 LEDs in parallel connected to the 5V of USB cable

--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.38
^ ^ 15:02:01 up 2 days 18:36 0 users load average: 1.07 1.10 1.12
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa
 
Man-wai Chang said:
Yes. But the one I got is of UK 13Amp standard. The ground pin of the
adaptor is plastic, NOT metal. So the shielding in the USB cable is NOT
working, I suppose.


When I unplugged the AC to USB power adaptor, the noise was gone.

The full circuit:

220V AC
---> AC-to-USB power adaptor (UK)
---> 3-feet USB cable
---> 2x100ohm resistors
---> 2 LEDs in parallel connected to the 5V of USB cable

OK, so now there is no ground in the circuit :-)

x---------------\||\--- rectifier--+- --resistor --+
Primary /||/ | |
220VAC Side \||\ cap LED
(Switcher) /||/ | |
x---------------\||\---------------+-- ------------+

This is a rough block diagram, ignoring the details. There is a
transformer between the primary and secondary side, which provides
electrical isolation. There is nothing wrong with your circuit
as such - it's the nature of the adapter that is at fault.

Normally, there would be a ground, if the adapter was connected to a
regular USB peripheral.

x---------------\||\--- rectifier--+----> 5V_in --- USB ---------- Computer
Primary /||/ | Peripheral Side
220VAC Side \||\ cap | |
(Switcher) /||/ | | |
x---------------\||\---------------+-------------------+-------------+
|
Safety ===
Ground =
Computer
Chassis

This really shouldn't make that much difference, except that the top diagram, the
entire secondary can function as an antenna.

If you want to experiment, you can try grounding your LED circuit.
This is for a test.

x---------------\||\--- rectifier--+---resistor --+
Primary /||/ | |
220VAC Side \||\ cap LED
(Switcher) /||/ | |
x---------------\||\---------------+--------------+ <-- touch this to
computer chassis
metal, which is at
safety ground potential.

With that point grounded, listen for a change in the A.M. interference.

A second test you could make, is to place a significant DC load across the
adapter. Some switching circuits, the switching frequency changes with loading.
You might be able to shift the interference frequency away from the A.M. band.
(I wouldn't count on this though. It could also be a constant
frequency circuit for all I know. There are constant frequency
variable pulse width, and fixed pulse width variable frequency
switching power circuits.)

When the noise frequencies are much higher than A.M. band, you can
use a ferrite core as a conducted interference filter. You take
the two DC leads coming from the adapter, and wind both of them around a
ferrite core.

http://www.ce-mag.com/archive/02/11/may.html

But in this case, ferrites aren't as effective at A.M. frequencies, as
they are at say 100-1000MHz. Another device you can use instead, is
a common mode choke. I've never needed to use a common mode choke in
a circuit. I think Figure 3 is the configuration you'd want. I don't
even have any common mode chokes in my parts cabinet.

http://www.butlerwinding.com/store.asp?pid=28349

The noise issue, could also be on the primary side, working its way
back into the wall outlet. There should be filtering elements on
the primary side, to prevent that.

My guess is, the secondary side is functioning as an antenna.

1) Keep the wires to the LED and resistor, extremely short, and see if
the level of interference is decreased. Of course, this affects the
usefulness of the design.
2) Grounding the circuit at one point, may help a tiny bit. (All it
should do, is change the "antenna" characteristics a bit, not eliminate
it entirely.)
3) Using a ferrite bead would help, but you'd need a material that peaks
at a low frequency. Or use several turns of wire around the ferrite.
4) It's possible a common mode choke, would be useful at lower frequencies,
but I've never had to solve a problem like that, so I don't know where
to start. Magnetics have "DC saturation", so there are limits on how
small the choke could be, before the DC from your application saturates
the core. But as for picking number of winds of wire, core material,
inductance level or the like, I haven't a clue.

Paul
 
OK, so now there is no ground in the circuit :-)

x---------------\||\--- rectifier--+- --resistor --+
Primary /||/ | |
220VAC Side \||\ cap LED
(Switcher) /||/ | |
x---------------\||\---------------+-- ------------+

This is a rough block diagram, ignoring the details. There is a
transformer between the primary and secondary side, which provides
electrical isolation. There is nothing wrong with your circuit as such -
it's the nature of the adapter that is at fault.

Normally, there would be a ground, if the adapter was connected to a
regular USB peripheral.

x---------------\||\--- rectifier--+----> 5V_in --- USB
---------- Computer
Primary /||/ | Peripheral
Side
220VAC Side \||\ cap |
|
(Switcher) /||/ | |
|
x---------------\||\---------------+------------------- +-------------+
|
Chassis

This really shouldn't make that much difference, except that the top
diagram, the entire secondary can function as an antenna.

If you want to experiment, you can try grounding your LED circuit. This
is for a test.

x---------------\||\--- rectifier--+---resistor --+
Primary /||/ | |
220VAC Side \||\ cap LED
(Switcher) /||/ | |
x---------------\||\---------------+--------------+ <-- touch
this to
computer
chassis
metal,
which
is at
safety
ground
potential.

With that point grounded, listen for a change in the A.M. interference.

A second test you could make, is to place a significant DC load across
the adapter. Some switching circuits, the switching frequency changes
with loading. You might be able to shift the interference frequency away
from the A.M. band. (I wouldn't count on this though. It could also be a
constant frequency circuit for all I know. There are constant frequency
variable pulse width, and fixed pulse width variable frequency switching
power circuits.)

When the noise frequencies are much higher than A.M. band, you can use a
ferrite core as a conducted interference filter. You take the two DC
leads coming from the adapter, and wind both of them around a ferrite
core.

http://www.ce-mag.com/archive/02/11/may.html

But in this case, ferrites aren't as effective at A.M. frequencies, as
they are at say 100-1000MHz. Another device you can use instead, is a
common mode choke. I've never needed to use a common mode choke in a
circuit. I think Figure 3 is the configuration you'd want. I don't even
have any common mode chokes in my parts cabinet.

http://www.butlerwinding.com/store.asp?pid=28349

The noise issue, could also be on the primary side, working its way back
into the wall outlet. There should be filtering elements on the primary
side, to prevent that.

My guess is, the secondary side is functioning as an antenna.

1) Keep the wires to the LED and resistor, extremely short, and see if
the level of interference is decreased. Of course, this affects the
usefulness of the design.
2) Grounding the circuit at one point, may help a tiny bit. (All it
should do, is change the "antenna" characteristics a bit, not
eliminate it entirely.)
3) Using a ferrite bead would help, but you'd need a material that peaks
at a low frequency. Or use several turns of wire around the ferrite.
4) It's possible a common mode choke, would be useful at lower
frequencies,
but I've never had to solve a problem like that, so I don't know
where to start. Magnetics have "DC saturation", so there are limits
on how small the choke could be, before the DC from your application
saturates the core. But as for picking number of winds of wire, core
material, inductance level or the like, I haven't a clue.

Paul

A snap-on ferrite may help.

http://www.fair-rite.com/cgibin/catalog.pgm#select:freq1

Winding both power leads many times through the ferrite will work as a
common mode choke.
 
david said:
A snap-on ferrite may help.

http://www.fair-rite.com/cgibin/catalog.pgm#select:freq1

Winding both power leads many times through the ferrite will work as a
common mode choke.

They make it hard to bookmark one of their links. This is an example
of one of their suppression cores. You'd wrap the two DC power leads
through the hole in the core and around the ferrite. Multiple turns
helps. What bothers me about this, is the performance isn't that good
at 1 MHz. (You'd try to get this as close as possible to the body of
the adapter, so that any power leads away from the body of the adapter
are filtered. The further away from the adapter you install this, the
less good it will do, on the output side.)

http://www.fair-rite.com/cgibin/catalog.pgm?THEONEPART=2631102002&SEARCHAGAIN=N#select:onepart

At work, they bought a few "kits", which is a sampler with all kinds
of forms of ferrite in it. Which is another way to buy them (if you
have bags of money to burn). With a "kit", you can try the different
types, and see how effective they are.

Paul
 
http://www.fair-rite.com/cgibin/catalog.pgm?THEONEPART=2631102002&SEARCHAGAIN=N#select:onepart
At work, they bought a few "kits", which is a sampler with all kinds
of forms of ferrite in it. Which is another way to buy them (if you
have bags of money to burn). With a "kit", you can try the different
types, and see how effective they are.

But the ferrite still needs a grounded shield to work, doesn't it?

--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.38
^ ^ 18:35:01 up 3 days 22:09 0 users load average: 0.00 0.01 0.05
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa
 
Man-wai Chang said:
But the ferrite still needs a grounded shield to work, doesn't it?

Ferrite is resistive to AC current flow. I would expect, if there was a
powerful RF signal in the wires, the ferrite would get warm.

In terms of magnetic fields in the ferrite, they're contained in the
toroidal shape. There shouldn't be a lot of flux leakage.

Using a ferrite bead, ring, or donut, would be an attempt to prevent
RF current from getting into the long (antenna) wire. By placing the
ferrite next to the adapter body, the intention would be to stop the
antenna from having a radiating area. Only the adapter body itself would
be able to radiate the A.M. signal.

Notice that some computer cables come with a "ferrite blob" on the cable,
for the same reason, so that the computer cable will not function as an
antenna.

http://ecx.images-amazon.com/images/I/41i3O4jU1eL._SL500_AA300_.jpg

Paul
 
Ferrite is resistive to AC current flow. I would expect, if there was a
powerful RF signal in the wires, the ferrite would get warm.

The USB cable should carry only DC, not AC.
Notice that some computer cables come with a "ferrite blob"
on the cable, for the same reason, so that the computer cable will
not function as an antenna.

In fact the USB cable I am using has 2 ferrite blobs.


--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.38
^ ^ 23:52:01 up 4 days 3:26 0 users load average: 0.04 0.04 0.05
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa
 
Man-wai Chang said:
The USB cable should carry only DC, not AC.


In fact the USB cable I am using has 2 ferrite blobs.

Radio frequency signals come in various amplitudes.

The adapter you purchased, should meet something like FCC part 15, if
it is designed to be exported to the USA. Our DOC in Canada would
have a similar EMI specification. That should bound, the amount of
A.M. signal that can come out of the thing.

Some of the techniques I'm suggesting, are for "small" leaks, many dB
down. You can't reasonably expect to stop a large signal problem, where
there was say, a 5V amplitude A.M. signal on the cable (like if the
adapter was actually oscillating, instead of just leaking a bit
of signal from the primary side switching being coupled into the
output).

In this example, a product is being tested, using a spectrum analyzer.
The red line would be the FCC class B limit. Products are supposed to be
tested by the manufacturer - failure to test properly, or to stay
under the red line, could result in product being stopped at the
border to a country, when being imported. All it takes is a complaint
to authorities, that the product is radiating. Usually, the complaint
comes from a competitor, who seeks to put them out of business :-)

http://www.low-powerdesign.com/picts/mcculley_fig4_455x371.png

The use of ferrite beads is a "compliance aid", it isn't a complete
solution.

Paul
 
The adapter you purchased, should meet something like FCC part 15, if
it is designed to be exported to the USA. Our DOC in Canada would
have a similar EMI specification. That should bound, the amount of
A.M. signal that can come out of the thing.

It's something like this picture (but a HK$10 one):

http://cgi.ebay.com.hk/AC-to-USB-Power-Adapter-Charger-for-iPod-iPhone-2G-3GS-/180510504738

Note the plastic ground pin. I don't think there was a similar adaptor
out there that connected the shield mesh of the USB cable to the AC
ground....

--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.38
^ ^ 19:31:01 up 4 days 23:05 0 users load average: 0.00 0.02 0.05
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa
 
Man-wai Chang said:
It's something like this picture (but a HK$10 one):

http://cgi.ebay.com.hk/AC-to-USB-Power-Adapter-Charger-for-iPod-iPhone-2G-3GS-/180510504738


Note the plastic ground pin. I don't think there was a similar adaptor
out there that connected the shield mesh of the USB cable to the AC
ground....

The USB cable shield shouldn't be connected to the "active" pins on the AC.

The third prong, the safety ground, is used in some designs as a place
to dump ripple current or the like. But as I was explaining previously,
when it comes to computer peripherals, you want the computer to
provide a ground path, through the computer chassis safety ground
connection. The peripheral power supply in that case, should be
floating. You don't really want two DC ground paths, because the
grounds could be at slightly different potentials.

I have a two prong adapter here, which powers my external USB hard drive.
(The two prongs, means it is the floating kind.) It is connected to the
wall right now (which means the switcher is currently operating), and
the USB hard drive is switched off and the USB cable is unplugged. And yet,
there aren't any emissions coming from the adapter or the long wire on
its output. So it is possible to find adapters that don't upset the radio.

Paul
 
I have a two prong adapter here, which powers my external USB hard drive.
(The two prongs, means it is the floating kind.) It is connected to the
wall right now (which means the switcher is currently operating), and
the USB hard drive is switched off and the USB cable is unplugged.

Turn it ON, place it near a AM radio and see if there would be noise...
And yet, there aren't any emissions coming from the adapter or the
long wire on its output. So it is possible to find adapters that
don't upset the radio.

Already trying to find one...

--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.38
^ ^ 15:39:01 up 5 days 19:13 0 users load average: 1.19 1.10 1.06
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa
 
The USB cable shield shouldn't be connected to the "active" pins on the AC.

Of course!
... you want the computer to provide a ground path, through the
computer chassis safety ground connection. The peripheral power
supply in that case, should be floating. You don't really want two
DC ground paths, because the grounds could be at slightly different
potentials.

How could I re-create the ground path of a power supply without a real
one? I supposed once I connected the mesh shield of the USB cable to the
safety ground, the AM noise would be absorbed!

--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.38
^ ^ 15:44:01 up 5 days 19:18 0 users load average: 1.14 1.09 1.05
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa
 
Man-wai Chang said:
Of course!


How could I re-create the ground path of a power supply without a real
one? I supposed once I connected the mesh shield of the USB cable to the
safety ground, the AM noise would be absorbed!

OK, I dug out a portable AM radio, and I can pick up a signal
from my adapter, at about a foot or two away. The interference
is coming from the body of the adapter, and not from the
wire extending away from it. As soon as my portable AM
radio moves away from the body of the adapter, the signal
level drops.

The signal is not completely gone, until I move to the next
room. So the signal is still detectable at perhaps 12-15 feet.
That is how far away I have to move, so I can pick up distant
stations (so-called "skip" reception). A local radio station
has enough signal level to blot out the adapter noise.

*******

Making the ground connection, may make a slight difference to
the antenna effect, but won't stop it entirely. I tried the ground
connection on mine, and it made no difference.

I do have a regular radio in the room (line powered) as well, but
it's always tuned to FM (88-108MHz) rather than AM (0.53 to 1.710MHz) band.

Waving the portable AM receiver around the computer room, there are other
sources of audible noise as well. So the adapter is not the
only noise source. If I hold the AM receiver right next to the
computer, the front end of the AM receiver saturates (and
quiets down as a result). So there is garbage coming from the
computer too.

This is partially one of the reasons, that my analog TV antenna (NTSC)
is located on the other side of the house, and a 75 ohm coax cable
brings the signal into the room I'm sitting in. That way, any digital
noise in the room, doesn't get into the broadcast TV signal. I don't
use cable TV here, and just go for broadcast stations. I've never been
a cable customer and I'm proud of that :-)

I guess I don't consider that adapter to be a significant interference
source, since as soon as I leave the room, the AM receiver can pick up
regular signals again. The adapter would not be upsetting my next door
neighbor's AM reception.

A ferrite bead wouldn't fix my adapter, since the signal comes from the
adapter body itself. I can move the AM radio along the 12V wire from the
adapter, and there doesn't seem to be a signal on it.

The signal I'm getting, seems to peak around 0.60 MHz (low end).

So maybe your adapter isn't abnormal after all.

*******

If this still bothers you, there are other alternatives.

This is how you build a "quiet adapter". This is how people
did hobby power supplies, before all those small switching adapters
came out. You can tell you've got a "transformer" based adapter,
based on the weight. The newer switching adapters are all
lightweight. A transformer based power supply is heavy by
comparison.

http://images.myfilehost.us/images/rgh1232594112x.JPG

If you held your AM radio next to that, there would be no
additional noise, compared to holding the AM radio next to
an AC line cord.

Paul
 
So maybe your adapter isn't abnormal after all.

Thank you for the time doing the experiment! I learnt something as well.
This is how you build a "quiet adapter". This is how people
did hobby power supplies, before all those small switching adapters
came out. You can tell you've got a "transformer" based adapter,
based on the weight. The newer switching adapters are all
lightweight. A transformer based power supply is heavy by
comparison.

http://images.myfilehost.us/images/rgh1232594112x.JPG

Studying...

--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.38
^ ^ 21:17:01 up 6 days 51 min 0 users load average: 1.07 1.09 1.13
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa
 
http://images.myfilehost.us/images/rgh1232594112x.JPG

Where is the ground pin in the diagram??? I can't find it.

--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.38
^ ^ 21:22:01 up 6 days 56 min 0 users load average: 1.19 1.17 1.15
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa
 
Man-wai Chang said:
Where is the ground pin in the diagram??? I can't find it.

The supply has a floating secondary. The transformer provides
isolation from the line side, making it safe.

*******

When you connect that supply to a computer, ground would go
to the rail at the bottom of the picture. The computer would
establish the ground reference, if you were powering a peripheral.

If all you're doing, is operating your LED circuit, no ground is needed.
Just stick your resistor and LED combination across the output on the right.

The three terminal regulators, are available for many different voltages.
These are some examples of ones in my parts bin.

7805 +5V Good for USB devices, a USB 2.5" hard drive enclosure
7810 +10V Good for running remote burgler alarm circuits needing 9V.
7812 +12V Good for powering portable computer fans, for cooling.
I have one of these in the computer room right now.

7905 -5V Useful for putting back a signal on pin 18 of an ATX supply,
for computers that insist on a -5V rail.
7912 -12V In combination with a +12V regulator, used for powering
operational amplifier circuits. Or for powering RS232
level shifter chips (GD75232 type).

The transformer, bridge rectifier, and filter cap connection details change
a bit, for the negative regulators. And if you need further connection
diagrams, the three terminal regulator manufacturers provide data sheets.

(Start at the back of this document and work back to the beginning.)

http://www.fairchildsemi.com/ds/LM/LM7805.pdf

(Example of a three terminal regulator, used for making negative voltages
with respect to ground. The incoming voltage to the regulator, must also
be negative, so care must be taken in connecting the transformer, bridge
rectifier, and filter cap, to this thing.)

http://www.national.com/ds/LM/LM7905.pdf

Your LED circuit doesn't really need a regulator. You could do just
a transformer, bridge rectifier (or even a single diode), and your
resistor and LED circuit. A transformer and bridge rectifier produce
a 120Hz pulsating DC voltage, which would still be suitable for making
a light source from a LED.

1) Simplest circuit. A $0.10 LED has a PIV (Peak inverse voltage) rating
of 5V. The transformer would have to be specially selected
in this case, to meet that. I don't think I've seen any
3V AC transformers.

x---------------\||\------------------- resistor --+
/||/ | +
220VAC \||\ LED
/||/ | -
x---------------\||\-------------------------------+

2) If you have any concerns about PIV rating on the LED, you add a 1N4001
thru 1N4004 type diode in series. For example, you could purchase a
6.3V filament transformer, rated to run from your supply voltage there.
This gives a pulsating 60Hz DC signal, across the LED. The "band" on the
1N4001 in the diagram, would be on the right hand side.

|\|
x---------------\||\---------| |------ resistor ---+
/||/ |/| | +
220VAC \||\ 1N4001 LED
/||/ | -
x---------------\||\-------------------------------+

3) If you use a bridge rectifier (four lead device, two leads labeled
with "~" symbol for transformer input, other two are "+" and +-"),
then the DC fed to the LED pulsates at 120Hz.

+---------+
| |
x---------------\||\---------| ~ + |------ resistor ---+
/||/ | | | +
220VAC \||\ | | LED
/||/ | | | -
x---------------\||\---------| ~ - |-------------------+
| |
+---------+

There is an example of a bridge rectifier here. This is the one
I used in a recent project. It's huge, so you can't lose it :-)
This is part of my new audio amp, power supply section. This is
much bigger than needed, but was purchased off the rack at my
electronics store, because it was actually cheaper than some
smaller rated ones.

http://www.nteinc.com/specs/53000to53099/pdf/nte53006_10.pdf

4) Adding an electrolytic capacitor, removed most of the pulsations
and smooths the voltage. But if the flicker is at 120Hz, I don't
really think this is necessary. Maybe 100uF 50V would be a safe choice.
Being careful, to connect (+) on the cap, to (+) on the bridge rectifier.
All the pluses, go together. The high working rating on the voltage
of the capacitor, is so you can use the capacitor for other projects
later.

+---------+
| |
x---------------\||\---------| ~ + |----+--- resistor ---+
/||/ | | | + | +
220VAC \||\ | | Cap LED
/||/ | | | - | -
x---------------\||\---------| ~ - |----+----------------+
| |
+---------+

As an example of purchasing a transformer, there are some examples here.
Since I'm in Canada, I shop for 115 or 120V transformers.

http://www.circuittest.com/Data/CTCAT-15.pdf

For example, take 640-1203. It is a 120VAC input, 12 volt center tapped output.
In your case, all you want is half the output. That would be plenty to play
with. The nice thing about a center tapped transformer, is it is easier to
tell which side is the primary, and which is the secondary.

black red +--------+
x---------------/||/---------| ~ + |---- ...
\||\ | |
/||/ white | |
120VAC \||\---------| ~ - |---- ...
/||/ +--------+
black \||\ red
x---------------/||/---------/ NC

The output of the thing in the previous diagram, looks like this (120Hz).

http://www.play-hookey.com/ac_theory/images/ps_sine_fw_rect.gif

( http://www.play-hookey.com/ac_theory/ps_filters.html )

*******

You're probably wondering "what voltage will I get" ?

That is partially answered here. See the diagram in the
"Root mean square amplitude" section here.

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

When you buy the transformer, it is rated by (3) or RMS. We have
a 12V transformer, and we take half the output (red and white leads)
or 6V. When rectified by the bridge rectifier, 6V * SQRT(2) = 6 * 1.414
gives 8.48V peak. The diodes have a small drop across them as well,
which becomes more apparent when a little bit of loading is present.
You'd adjust your LED/resistor combination, for the new, higher voltage.

If you add a filter capacitor, plus a 7805, then the pulsations would be
completely eliminated, and your original arithmetic for selecting the
resistor for the LED, would still be valid.

*******

A word on safety. When you build an AC powered home project, add a fuse to
it. Such as a cartridge fuse. If you make a mistake, there may be fewer
consequences if a fuse is present. The fuse is placed close to the line side.
For example, if you're clumsy with insulating things, you definitely want
a fuse here. Where the fuse is connected on the left hand side, you
definitely want that well covered, so no accidents can happen.
You can still be electrocuted by this thing, is you're really
really stupid (like dropping the whole circuit into your bathtub
while the tub is full of water, and you're in it). But that is
true of so many other electronic items that run from AC.

black red +--------+
x----- fuse -----------------/||/---------| ~ + |---- ...
\||\ | |
/||/ white | |
120VAC \||\---------| ~ - |---- ...
/||/ +--------+
black \||\ red
x----------------------------/||/---------/ NC

I have a fuse in my new audio amplifier project, for protection.
I haven't managed to blow it, yet :-) For the small circuit
you're building, perhaps a 1 amp fuse would be sufficient.
If you make the fuse value too small, it might blow on inrush
current.

(Cartridge fuse holder)
http://www.radioshack.com/product/index.jsp?productId=2102784

This is the dangerous kind of fuse holder I used in my audio amplifier.
To make this safe, I cover the wiring area with a plastic plate. The
plastic cover is intended to prevent me from dropping my screwdriver onto
the contacts. My implementation is *not* waterproof. There is
still danger, if I spill a beverage in the box.

(Chassis type fuse holder - dangerous!)
http://www.radioshack.com/product/index.jsp?productId=2062257

Have fun,
Paul
 
Back
Top