Adding more leds to an existing LNK-306 design

18 posts / 0 new

For years, we have been producing a product that runs a string of 8-leds in two parallel strings of 4 in series at about 350mA total current. The circuit implements the LNK306 chip in a non-isolated format, as explained in AN-37. Our customer has recently Decided they want to double the amount of leds in the circuit. I figured keeping the output current the same, and doubling the output voltage would be the simplest means to achieve this.I have attached the updated schematic, as I thought it would work. I have only changed the size of the protective zener diode to 36V from 18V, and the voltage ratings of the Cfb capacitor to allow for a higher output voltage. I have retained the values of the Rsense resistor, and the bias network recommended. However, the circuit now goes into the classic .8sec reset condition. What should I be looking to change to correct this? I think I read I should be getting 1.65v at the FB pin, in reference to the Source pin. How long should this stay at 1.65v? Also, how can I possibly measure for the 49uA current limit? I tried to increase the Csense capacitor from 10uF @6.3v to a 22uF, no difference. I have also tried to increase the value of Cout from 1uF to 10uF, without any change. My sense resistor is 5.6-ohms, 1/2W. Any further info would be greatly appreciated, as always.

Hi Richard,
Good day and Thanks for using Power Integrations LinkSwitch TN. Since you doubled the output power, It could be that the 800ms reset is due to output overload. I didn't see your attached schematic on your post, but if you haven't used the PI expert online then it would be helpful for You to use it on your new design. The PI expert will give you an idea what needs to change if you update some parameters on your current design. I tried to used to check what happen if your double the Vout assuming you used Buck-Boost topology. It says in the PI expert that You need to increased the input capacitor and Inductance value. Here is the PI expert online link. You need to create your account so you can use it . When you enter the site, Create New Design and pull the drop down menu to locate the LinkSwitch-TN and select your topology. Please also attached your schematic again and your input and output specification so I can review and recommend solutions.

Thanks for the info. I thought I attached the schematic, but I think I had it in pdf format.So maybe it didn't take it. I was hoping the inductor would stay the same, since it was the same current. I was hoping for only a change to the divider network on the fb pin being an issue. I tried to use the PIexpert download version, but didn't see any provision for this type of circuit.

Reference design102.13 KB

Hi Richard,
Thanks for sharing your schematic. There is a need to increase the inductance to be able to deliver higher output power. Your topology is buck-boost design which is available in PI expert. Could you also give me the Input and output Voltage range requirement? Thanks

My input range is Universal power, so I figure 100-277, my output range is CC 350mA, if possible, with 21.6V - 23.2V range. I was looking for CC topology, but I only selected buck for some reason. Any help would still be appreciated. Thanks. The actual device we are using is the LNK306GN-TL. Is there a device we could increase the operating frequency so that I don't need as large of an inductor?

Hi Richard,
I have attached the PI expert design and the suggested modifications in the schematic for you to verify. Hope to hear from you soon with the results.

Hi! Thanks for the info! I too did that finally this morning using the buck//boost topology, instead of the buck. I'm glad to see we came up with the same results. The only thing I wasn't too sure of is changing the 270-ohm to a 24.68-ohm resistance. I was going by the AN-37 value they show which was 2k and 300 ohm. For some reason I was thinking these were fixed values using the CC LED circuit. I still don't fully understand what's happening with that fb input after reading it several times, other than it is looking for a 1.65V at 49uA input, and that if it isn't good, it goes into the .8s reset cycle. I'm almost afraid to try to measure anything in fear of using the wrong reference point!! I do have an isolation transformer though,at least. I'll let you know how it goes. Thanks again.

Hi Richard
Hope You are doing fine. Circuit location R3 should be 24.68 Kilo Ohms not Ohms. R3 will set the output voltage to 23.2V.

I did all the changes as recommended, and I'm still getting a .8s flasher. In the reference LED design in AN-37 it shows a 2k and 300-ohm for the Rfb and Rbias. It seems like 24k is a big difference in value. Any more clues? I didn't get any warning or anything when I put the values in the PI-Expert calculator, including my inductor resistance, etc. Any help would be appreciated asap! Thanks.

Hi Richard,
Yes you are right it should be close to 300 Ohms. Sorry for the confusion. Can you try increasing R2, say 10 Ohms to verify if the auto-restart is due to overload. Also , have your tried increasing the input capacitor 33uF?

Hello PI-Martok,
I switched back to the 300-ohm. I switched the 5.6-ohm sense resistor to a resistor sub box. The circuit will work okay with a 7-ohm resistor down to 88VAC input, but to get it to work at 6-ohms, I have to increase the input voltage to at least 100VAC. So maybe this clue will help. All I know is I have an anxious customer. Should I have an output cap on this circuit? I also for some reason have a pre-load resistor of 5.1k. This circuit also switches between the flood array shown, and a single-led spot lamp, and I think it was put in to keep the spot from a single flash before it turned on. I have a 100k shown on the schematic, but I believe this is why it was reduced to the 5.1k.

Is there a safe means that I can probe the FB input, to capture what we are seeing? I could use a scope with 2 inputs as a differential input, but what would I reference the pin to? The Source pin?
Just wondering what else I could be looking for to stop this. Thanks again. Please see previous comment.

Hi Richard,
C4 is the output capacitor and it was too small. Please also try to increase to 220uF/50V to reduced the voltage and current ripple. Could be one reason why the device is doing auto-restart since the output is sense for the regulation.

The pre-load resistor R5 prevents the output voltage to go up during no load condition lowering it down will affect your standby power. The recommended value is 30mW resistor power dissipation (R = 24^2/0.03 = 19.2Kohms).

You can try to probe the FB pin to source.


Thanks again for the advice. Much appreciated. Unfortunately, I cannot access our lab at work due to mandatory shut downs. I will definitely try it out!

Best Nude Playmates & Centerfolds, Beautiful galleries daily updates
kira knightley porn top young porn models retro porn xhamster first time home porn janet jakson porn

Sexy photo galleries, daily updated collections

2 school uniform russian girls porn straight bodybuilder porn trick sex porn clips croc revieqw porn robosapien porn

Girls of Desire: All babes in one place, crazy, art

scottish men porn lil candy porn uncensored porn movies boxer porn porn rip cali

Daily updated super sexy photo galleries

is the best free porn site las vegas stars in porn mp4 porn downloeds free mp4 porn vidios lime wire of porn