# constant current generator

Joined Feb 27, 2008
104
Hi all,
I have developed a constant current provider using HV9921 for LED lighting. The output current of HV9921 is limited max to 20mA. At 20 VDC to 75 VDC output current is 15mA constantly. After 75 VDC output current starts increasing around 115VDC input voltage resistor blasted. my current requirement is 20mA constantly from 85 VDC to 220 VDC. can anybody tel me what is the reason for this?

#### Attachments

• 5 KB Views: 26

#### Wendy

Joined Mar 24, 2008
21,929
So where is the data sheet for a HV9921?

#### SgtWookie

Joined Jul 17, 2007
22,210
R4 doesn't really serve a purpose other than to waste power.
I don't see a reason for D5, either.
C2 seems quite large. 0.1uF should be OK.
L1 seems very small. Increasing it will decrease the ripple and also decrease the switching frequency of the HV9921.

You should have a 0.1uF cap between the junction of D2/D4 and -ve.

Joined Feb 27, 2008
104
Approximately how much should L1 be increased?

Joined Feb 27, 2008
104
I have increased L1 to 32mH. now from 20 to 58 VDC input the output current is 20mA after 58 VDC starts increasing... can you tel me how much should be the inductor value?

#### SgtWookie

Joined Jul 17, 2007
22,210
Offhand, I'd say for a single LED, 10mH or higher. If you're going to use several LEDs in series, increase the inductance by another 5 to 10 mH per LED, depending on the Vf.

The larger L1 is, the lower your ripple will be; but your parasitic capacitance increases.

The datasheet goes into a fair bit of detail on that.

Joined Feb 27, 2008
104
If i use an optocoupler(MCT2E) in the place of LED then how much is the value of inductor?

Joined Feb 27, 2008
104
as per formula given in the datasheet the calculated inductor value is 7.8mH. I tried with 7.8mH then also output current starts increasing after 58VDC. what is the problem?

Last edited:

#### SgtWookie

Joined Jul 17, 2007
22,210
Are you following the guidelines in the Layout Considerations section?

Is your inductor a high-Q type? Are you winding it yourself, or are you using a purchased version?

If your inductor is going into saturation, you will wind up with increasing current at higher voltages.

Joined Feb 27, 2008
104
Am using general purpose PCB to conduct this experiment. I have wounded Inductor myself.My inductor is not saturating, current is kept on increasing beyond 20mA for increase in input voltage

#### SgtWookie

Joined Jul 17, 2007
22,210
Am using general purpose PCB to conduct this experiment.
There's a problem.
Why don't you try the layout in the datasheet?
I have wounded Inductor myself.
What are you using for a core?
What are the dimensions of your inductor (core type, core diameter, length, number of turns, etc.)
My inductor is not saturating
How do you know this? Do you have an oscilloscope?

current is kept on increasing beyond 20mA for increase in input voltage.
So you still have a problem.

Try starting out by using the suggested layout in the datasheet.

#### italo

Joined Nov 20, 2005
205
Most LED are with a breakdown voltage 5-10 volts. So if you connect a 75 v of LEDS some may leak more then others. If this is the scenario then you may find out that poof one after the other will fail. there is a solution add a very large resistor across each LED all of the same value so the leakage will be the same for all. the value well it depends on the LED doesn't

#### SgtWookie

Joined Jul 17, 2007
22,210
Most LED are with a breakdown voltage 5-10 volts.
Are you thinking Vf or Vr?
Must be Vr, because most common LEDs have a Vf of somewhere between 1.3v (infrared) to around 4v (blue and white).

Vr is generally somewhere between 5v to 7v.
So if you connect a 75 v of LEDS some may leak more then others.
Did you even look at his schematic, or look at the datasheet?

He is powering ONE LED.

If this is the scenario then you may find out that poof one after the other will fail. there is a solution add a very large resistor across each LED all of the same value so the leakage will be the same for all. the value well it depends on the LED doesn't
You have no idea what you are talking about.

#### thatoneguy

Joined Feb 19, 2009
6,359
From the notes, a design running < 85V is slightly different than one designed for higher.

What is the purpose of the 1N4001 diode after L1 in your schematic?

If the current increases, the inductor is going into saturation. What is the core and number of turns, or, if you have an LCRZ meter, what is the value and Q of L1?

#### SgtWookie

Joined Jul 17, 2007
22,210
From the notes, a design running < 85V is slightly different than one designed for higher.
I think that reference was for AC vs DC; AC input requires rectification and filtering - or perhaps I'm missing something?

What is the purpose of the 1N4001 diode after L1 in your schematic?
You meant D5, a 1N4007, right? I don't see any condition where it might become forward biased either.

If the current increases, the inductor is going into saturation.
That's my train of thought as well. Additionally, in the datasheet under the layout considerations, numerous cautions are made along with specific recommendations. That's why I suggested above that he start with the recommended layout.
What is the core and number of turns, or, if you have an LCRZ meter, what is the value and Q of L1?
I have a feeling that pradeeba is using an inappropriate core, and/or may not be isolating the core from the windings by using tape or a bobbin, and/or may be using so many turns that the self-capacitance of the inductor has risen quite high.

Ronald Dekker's "Flyback Converters for Dummies" page is a good resource for projects like this:
http://www.dos4ever.com/flyback/flyback.html

Joined Feb 27, 2008
104
The larger L1 is, the lower your ripple will be; but your parasitic capacitance increases.
For the value of 2.1mH the parasitic capacitance is in the order of nF. if i increase the inductor value then it will be in the order of pF. what do you suggest?

Joined Feb 27, 2008
104
Hi,
can anybody help still am not getting constant current output. i have changed the inductor value to 33mH. still same result............... am using axial leaded inductor now,