LT3797 SEPIC converter PWM control

[boulnat]

I have been trying to make a sepic converter using an LT3797.
I'm having issues with the PWM control, it seems like I can't dim the LED no mater the frequency or duty cycle I give to the LT3797.
For the sepic design I followed the datasheet, they have one example of a SEPIC.
I m trying to control LEDs;
VIN 24V
VOUT 24V-60V
IOUT 1A

Kind regards

 

[KlausST]

Hi,

I doubt that there is a mistake in the datasheet.
So as long as you dont show your circuit... we won´t be able to help you.
And for sure we need the test conditions like PWM frequency, duty cycle, signal levels... maybe a scope picture is a good idea.


Datasheet tells: "..demands careful attention to board layout and component placement. "
Did you care about this? --> show us your PCB layout.

Klaus

 

[Easy peasy]

perhaps you have wired it as a boost, not a sepic ...? need an accurate "as built" schematic and some clear photographs of the build to answer your question ...

 

[schmitt trigger]

I join the chorus asking to see YOUR schematic.

 

[Easy peasy]

RE: driving leds, once you get up to a high enough voltage to start the string of leds glowing, it is the current that needs to be controlledas the voltage will only rise a very little with increased current/brightness.

 

[boulnat]

So here is the schematic,
I m not in the lab right now as soon as I can hook it up I will l take a picture of the oscilloscope but I can tell you what I tried for the PWM.

120 Hz 5V and 1.4V peak with a DC of 20% 50% 70%.
1KHz 5V and 1.4V peak with a DC of 20% 50% 70%.
50 KHz 5V and 1.4V peak with a DC of 20% 50% 70%.

and for the load I used a bunch of Zener 1N4007 at first because i didn't want to burn my CREE LEDs .

To clarify what I meant by "VOUT 24V-60V". I will configure some output with different number of LED so the output will range from 24V up until 60V.
example: OUT1 = 24V - OUT2 = 45V - OUT3 = 60V.

As I understood I should configure my sepic for 60V and the LT3797 will adjust the voltage depending on the load and keep my 1A constant.

 

[KlausST]

Hi,

120 Hz 5V and 1.4V peak with a DC of 20% 50% 70%.

What does that "5V" and "1.4V peak" mean?

bunch of Zener 1N4007

1N4007 is no zener. You may use 1N4007, but usually connected in opposite direction as zeners. Hopefully you did it correctly.
You know: we can't validate informations that you don't give. ...like PCB layout...

Klaus

 

[boulnat]

What does that "5V" and "1.4V peak" mean?

What I mean by peak is the voltage of the ON state, I tried the different frequencies with different voltages
PWM ON 1.4V - OFF 0V with 120Hz - 1Khz - 50Khz
PWM ON 5V - OFF 0V with 120Hz - 1Khz - 50Khz

1N4007 is no zener. You may use 1N4007, but usually connected in opposite direction as zeners. Hopefully you did it correctly.
You know: we can't validate informations that you don't give. ...like PCB layout...

yes I put the 1N4007 in the opposite direction of a zener, my mistake the 1N4007 is a Diode as you pointed is no zener .

Here is the layout and the pdf is the splitted layouts.
It's a 4 layout board:
L1 : GND
L2: signal (PWM - TGx - Gatex)
L3: power (Vin - Vref - INTVCC) M6=>LED3 - M5=>LED2 - M4=>LED1
L4: GND (SGND (RT, VC1/2/3, SS1/2/3, Sync) - PGND)

 

[KlausST]

Hi,

as far as I can see the green layer of the PDF short circuits all individual areas of the red and pink layer.

Klaus

 

[boulnat]

Hi,

So you mean that some of my Vias are connecting green (GND) to red (GND) and pink (VIN) ?
I think it s a print issue, because I can assure you that on the board there is no connection between the green and the individual area of pink and red.

 

[KlausST]

Hi,
page4 of your PDF looks like it causes short circuits.

Easy to check with an Ohmmeter.

Klaus

 

[treez]

So you can give maximum power to the LEDs, but you just cant dim it?
The dimming is done by the PFETs which should get turned on/off in synchrony with the control signal.
Is the PFET VGS going up and down when you try and dim it?