LT1308B current limit

A quick update to this: I removed the power LED load and with no load the SEPIC circuit immediately goes to 3.2 Vout with 1.22V FB. So the load is the issue. But a power LED only pulling 12mA at 2.56V doesn’t seem like a particularly heavy load given that the chip has 2-3A switches in it. I could see maybe 300mA being an issue but this seems too low. Any suggestions?

The converter does not support any load with 1.2Vin, but the same thing happens when the input is raised to 3.7V.

And, if you remove the load, the output regulates as expected.

I assume when you say all is fine (in the first paragraph), you are talking about a no load condition.

If all the above is right, my guess is your converter is very unstable, has a poor layout, or you are using some wrong component.

Cheers!

DV

I fully agree with you that there's no way I can prove to you over the internet that I'm using the correct circuit, components or layout. :)

Just for argument's sake though (and this is all true I swear it) let's say the circuit is properly laid out within a tight 2cm^2 area following the layout guidelines in the datasheet. All components are exactly the value/tolerances specified in the sample circuit in the datasheet (except for 1 additional capacitor at output). Every ancillary function has been stripped from the board - I used an X-Acto knife to score each trace not directly related to the SEPIC then tested each with a continuity meter to ensure they're dead. There are no active components other than the SEPIC ones and no shorts (or partial shorts) to ground. The schematic matches the datasheet sample circuit exactly.

Tested unloaded running on a NiMh 1.2V cell, the circuit outputs a perfect 3.2V. Loaded with a 5mm white LED it outputs just enough to send 5mA through the LED.

I've even built a 2nd board exactly like the first and confirmed all traces reflowed properly etc and it behaves in exactly the same manner.

My question is: is there any additional logic within the LT1308 that isn't present in the block diagram that would limit output or otherwise affect switching. There is a constant 0.9V at the FB pin, which according to the block diagram, should drive output higher until Vfb reaches 1.22, but in practice it isn't. Is there some kind of "short at output" safety mechanism, or artificial current limit when Vin < Vfb, or reduced switching when Vout > 2.5x Vin or anything along those lines that isn't documented? My circuit looks precisely like the sample except with an LED at the output so I'm struggling to understand either what I'm doing wrong or what it happening inside the black box of the LT1308.

Thanks for your time, I do appreciate any help or answers you can provide.

The FB is low because the output is low. 

The output is low because the converter is current limiting, or your source is current limiting. This of course is my guess, but we need to find out if I'm wrong.

In order to have some visibility here  please provide some hard facts, not that I don't believe you. You do sound like an honest person.

Please post a schematic of you application, with desired Vin range, Vout and max load current. 

Also, provide the following data from an oscilloscope. All waveforms with the same time base.

Do it once without load, when the output is good, and again when the output is low.

1. switch waveform

2. Input voltage

3. VC pin

4. SHDN pin

Cheers!

DV