Does anyone know the method of control for this Buck LED driver?

[treez]

Hello,
I cannot tell from the datasheet ,what type of control this MXL10803 Buck LED driver is using....

MLX10803 LED driver datasheet:
https://www.melexis.com/Asset/MLX10803-IC-specification-DownloadLink-4840.aspx

from the block diagram on page 3 it looks like the two comparators set the peak and trough of the inductor current? (and thus the average output current)...thus it would be a hysteretic type control.?
...however, the pin descriptions on page 7 say that these two comparators are concerned with NTC feedback (temperature regulaton).
Also, the pin description page appears to indicate that the ROSC resistor sets the "off-time" (presumably the off time of the FET)...however, how does that link in with the two comparators?
Also, what is the equation that relates off-time and ROSC resistor value?, its not stated.

Does anyone understand this IC as such?

Is it a constant off time converter. If so, whats the maximum source sense resistor trip voltage?...so that we can be sure we don't saturate the inductor.
What makes the fet turn on? (is it the start of a new oscillator cycle?, or the end of some off time interval ?) ....then what makes the FET turn off?

 

[mtwieg]

It's a bizarre part. It's apparently just a fixed (pseudorandom) off time CMC loop with the peak current set by the lower of the two Iref pin voltages. It seems this is not meant for average output feedback, but rather to allow a fixed resistor and a NTC to implement a thermal protection measure. The IC is only really intended to regulate the FET's peak current, which means it will have poor line and load rejection.

 

[treez]

thanks, the datasheet does mention the word "average", but as you say, I cant see it regulating average led current...the vref pin also seems to determine the peak fet source current.

 

[treez]

Please could I also ask about MLX10803 LED Driver IC...on page 9 it says that the leakage current flowing out of the VREF pin is = +20uA to -20uA.....this means it can flow either into , or out of the pin?...surely there is something wrong here?.....these pin leakage currents surely only ever flow either in or out?.........the datasheet implies that it could be either in or out....surely not?