Capacitive Reactance Circuit

[Fish4Fun]

Sooooo....it has been close to a month since the $0.99 LED driver from China race began.....has anyone received their package yet? If I had known everyone was going to rush off and order them I would have taken a bit more time selecting one I have actually ordered//tested...but I assume most of them in this power range come from the same design.....Any way....just wanted to see what the progress was....

Fish

 

[KrisBlueNZ]

I haven't received mine yet. It's overdue. They said if I don't receive it within a week, to let them know and they'll credit me back. I would rather they tried again :-/

 

[chopnhack]

Sooooo....it has been close to a month since the $0.99 LED driver from China race began.....has anyone received their package yet? If I had known everyone was going to rush off and order them I would have taken a bit more time selecting one I have actually ordered//tested...but I assume most of them in this power range come from the same design.....Any way....just wanted to see what the progress was....

Fish

I received correspondence from them on 11/13 asking if I had received it. I told them I had not and am still waiting on its delivery. They are asking me to wait 5 more days and then they would issue a credit for it.

As a friend told me, these are typically shipped via the least expensive method - tied to a tire and thrown in the ocean

 

[Fish4Fun]

Rats, sorry to hear that. There are numerous vendors with whom I have had really good luck with, I wish I had taken the time to select one of their offerings....I only linked that particular one because it came up 'first' in the search window....i didn't really expect anyone to order it....If you two PM me your mailing addresses I will send you one....I have plenty on hand and certainly wouldn't mind. I generally only order things from China that use the "e-packet" shipping option....I have received some items as quickly as 5 business days....other things not so fast, lol....But the e-packet service costs a little more, so on a single item like these it generally adds ~$0.50 to $1.00 to the total price....not a big deal on say 20 of them, but on a single piece it can make a lot of difference in price....Anyway, PM with your addies if you want me to send you one.....

Fish

 

[chopnhack]

Rats, sorry to hear that. There are numerous vendors with whom I have had really good luck with, I wish I had taken the time to select one of their offerings....I only linked that particular one because it came up 'first' in the search window....i didn't really expect anyone to order it....If you two PM me your mailing addresses I will send you one....I have plenty on hand and certainly wouldn't mind. I generally only order things from China that use the "e-packet" shipping option....I have received some items as quickly as 5 business days....other things not so fast, lol....But the e-packet service costs a little more, so on a single item like these it generally adds ~$0.50 to $1.00 to the total price....not a big deal on say 20 of them, but on a single piece it can make a lot of difference in price....Anyway, PM with your addies if you want me to send you one.....

Fish

5 days! Wow, I will have to look for that service next time I need something, thanks for the info and the offer! I will give them a few days as this has happened to me in the past and the item literally shows up within a day or two.

 

[Fish4Fun]

chopnhack,

Yea, if you are in the US then e-packet is the way to go for anything you want in the near future.....typical delivery is ~7 working days from when it is actually mailed....and e-packets come with tracking info....I would say that 5 days is the 'exception' not what you should generally expect....But only very rarely do I have an e-packet delivery take 10 working days or more (again, from when the item is ACTUALLY SHIPPED....if it takes them a week to mail it, then obviously the arrival date is pushed back....) I order lots of bits and pieces from China and am constantly amazed by both the price and the speed of delivery....quality is occasionally a little lacking....but if you steer clear of the 'deals too good to be true'....and stick with vendors with 1000+ transactions then I find I am frequently pleasantly surprised....

Fish

 

[chopnhack]

My tire from China came in with the LED driver tied on

It is a constant current driver! See the pics. The driver is marked BP9021
The cap is 4.7μF, the MB6F in front of it is a bridge rectifier, there are a few small resistors in front of the IC and a diode to the right of the transformer. I found it interesting how they too messed up the outlet ports of the wires from the little plastic box - they were mirrored LOL - glad to see I am not the only one it happens to! I found the air gap between line and neutral of the incoming power to be an interesting method of isolation. I wonder if I could implement it on my other circuit that needs good isolation.

There is no means for adjustment that I can see so I take the range of voltages/current on the packaging to be subject to incoming levels. I guess to some extent also the load. I wish I had a way of measuring the output to show you if it is clean or dirty. If I use this I will most likely need to drop the output current with a resistor.

 

[KrisBlueNZ]

Mine arrived a few days ago and looks exactly the same as yours. The IC is also a BP9021. Thanks for the link to the data sheet. There's no English version available (yet) but here's what Google Translate made of the Chinese main summary text.

BP9021 is a high precision PSR constant current LED driver chips. Chip in discontinuous inductor current mode for 85~265VAC full range of input voltage, power 3W following anti-isolated flyback LED constant current power supply.
BP9021 chip integrated 650V power switch, using the primary anti-feeder mode, without secondary feedback circuitry, and without compensation circuit, only requires minimal external components to achieve excellent constant current characteristic.
Collection with a patented drive and current detection methods, the chip operating current pole low, no transformer auxiliary winding detection and power, further reducing external components, greatly saving system cost and size.
BP9021 chip inside with a high-precision current sampling circuit, while mining with the patented constant current control technology to achieve high-precision LED constant stream output and excellent line voltage regulation. By adjusting the external resistor LED open-circuit voltage can be precisely controlled.
BP9021 has multiple protection features, including LED open / short circuit protection care, CS resistance short circuit protection, undervoltage protection, chip overheating tune Festival.
BP9021 using SOP-8 package.

I may have a look at the output ripple at some stage. I would like to run a performance test but I don't have a variac here.

It's not unusual for PCB designers to cut out part of the PCB to improve isolation by increasing the creepage distance. In this case they have used a simple straight cut in from the edge, but if the PCB maker supports it, you can also have cutouts that are fully internal and of various shapes, cut by a routing bit. The routing bit path can be specified by tracks on a mechanical layer. Ask the PCB manufacturer whether they can do internal routing and the diameter of their routing bit.

 

[chopnhack]

I will certainly ask about that! Thanks.

Care to fathom a guess to how much current is output when run by 120v?

 

[KrisBlueNZ]

I think the eBay page said 300 mA ±10% over the full input voltage range.

 

[chopnhack]

I got around to plugging this thing into mains today and boy, it buzzes like a hornet's nest!! It is certainly not suitable for its intended application (bedroom devotional light). The device stays lit until the LED burns out/supply fails whichever is first.

With all this talk of cell phone chargers, I realized I still had one from an old samsung. It is rated for 5v upto 0.7A. Open circuit shows 5.04V - kudos to Samsung!
It's UL listed to boot, so we know its isolated and safe - its also audible quiet. I don't know if it's electrically as well.

I would like to use this supply to power 1 LED bulb that I can dim if its too bright.

1. Should I work out a constant current supply circuit from this charger?
   
2. Since the adapter is already limited/filtered and regulated to 5v should I just install a potentiometer to power the LED?

Or is this charger simply too wasteful to use constantly plugged in as it will be?

 

[chopnhack]

Edit:

Ignore my drivel above..

Simple circuit = cell phone charger + 500Ω pot = constant current supply

If it's simple, I will make it needlessly complex

The only question that remains in my mind is how efficient is this supply? Am I losing a lot of power in the adapter or is this simply a matter of how many mA the LED is consuming plus the reciprocal of the switcher efficiency x power consumed?

I am consuming 15mA at 3V drop across the LED that would be 0.045W power being used by the LED.
Assuming the switcher is 90% efficient, 1/90%= 1.111 so total power consumed would be ~0.05w.

Have I got this right or am I completely off base?

 

[KrisBlueNZ]

A switching power supply's efficiency is normally specified at or near full load. It will be lower at lower output currents. This is partly because the SMPS has a fixed power overhead for its internal operation, which becomes a greater proportion of the total power consumption as the output power is reduced.

Also you need to calculate from the total output power, which is 15 mA at 5V, not 3V. If you only calculate based on the power dissipated in the LED, you are ignoring the power lost in the potentiometer.

So I guess your input power would be somewhere between 0.1W and 0.2W.

 

[chopnhack]

A switching power supply's efficiency is normally specified at or near full load. It will be lower at lower output currents. This is partly because the SMPS has a fixed power overhead for its internal operation, which becomes a greater proportion of the total power consumption as the output power is reduced.

Also you need to calculate from the total output power, which is 15 mA at 5V, not 3V. If you only calculate based on the power dissipated in the LED, you are ignoring the power lost in the potentiometer.

So I guess your input power would be somewhere between 0.1W and 0.2W.

Thanks Kris, that makes more sense. I didn't realize until you said that. So, power consumed by the switcher would not be seen when I used my meter, because it was happening on the other side of the transformer! Makes sense.

Worked out well until I fiddled with the potentiometer to adjust the brightness and it started getting dimmer, so I thought I was going in the right direction, but when I tried to make it brighter it didn't response. I think it's the wiring, I had to make some questionable connections and jam them inside the old socket receptable
I will take it apart and put everything on a strip board and make the proper connections.

Thanks!

 

[KrisBlueNZ]

You might have turned it too far and damaged the LED with overcurrent. You need a resistor in series with the potentiometer to limit the current to the LED's maximum rated current. For example if it's rated for up to 50 mA and its forward voltage is 3V, that means the resistor must drop 2V at 50 mA so it must be 40Ω so you would use 39Ω. That will ensure that even when the potentiometer is fully clockwise, the LED current will not exceed its rated maximum.

 

[chopnhack]

You might have turned it too far and damaged the LED with overcurrent. You need a resistor in series with the potentiometer to limit the current to the LED's maximum rated current. For example if it's rated for up to 50 mA and its forward voltage is 3V, that means the resistor must drop 2V at 50 mA so it must be 40Ω so you would use 39Ω. That will ensure that even when the potentiometer is fully clockwise, the LED current will not exceed its rated maximum.

Great idea! Thanks I will tear it apart tomorrow and fix.

 

[chopnhack]

Ended up using a 180Ω resistor, limited current to 17.5mA. Works great, thanks Krisl! Thanks all!