Power Fluctuations - Attributable to Which?

[IsaacT]

I had always thought of the diode as being the only factor in power fluctuations. How hard it was being driven, how much heat building up, etc all causing it to vary wildly. A number of lasers I have built in the past have had a habit of starting high and then dropping quickly.


But it occurs to me that possibly the driver has more to do with the fluctuations than the diode itself. Could it be that heat being dissipated by the chips on the driver could change the amount of current being delivered to the diode such that its power might drop?

Used to I figured unless the driver explicitly needed sinking I wouldn't, but if you can gain significant stability through that action then it suddenly becomes a different ballgame.

Thoughts?
-Isaac

 

[Rifter]

The electronics efficiency will drop as heat increases, I would always heatsink the driver if there is room to do so. Not only for efficiency but life of the driver as well.

 

[Sigurthr]

Most CC drivers will automatically correct for drift from temperature though, but yes it is important to heatsink whenever you can. The issue really comes about when the switching element is near the edge of its safe operating area.

However, to answer your original question about starting off high output and dropping low; this is usually a result of being run on batteries with limited current capacity. Batteries can supply less and less current as their voltage (and thus charge state level) drops. You might get 3A from your LiIon at 4.20V, but at 4.11V you'll only be able to pull 2.8A through it. So while your driver, which is set to 2.9A has to limit current initially, there comes a point where it is now "full open" very shortly into the run and current draw will only drop off further. This is common in high current builds. In low power builds it is more likely that heat or lack of sufficient forward voltage (due to raising dropout voltage and lowering cell voltage) is playing the major role.

 

[Jstr]

there comes a point where it is now "full open" very shortly into the run and current draw will only drop off further. This is common in high current builds.

Isn't that the point of having a buck or boost driver? So that the current stays constant throughout the batteries' discharge. Because the buck/boost driver can alter the forward voltage, right?

That sounds like it might be the case for a linear driver, however, which might be what you are referring to.

 

[Sigurthr]

A Buck driver can compensate for the declining level of possible current draw from the battery(s) but only until the voltage output from the battery(s) drops to the point where the forward voltage and dropout voltage is no longer sufficient to allow full current to flow through the diode,

A Boost driver exaggerates the current to voltage discharge curve of a battery as it increases the require current draw from the battery, thus loading the cell(s) harder causing them to discharge faster.

A Buck fed Boost (or Buck-Boost) can compensate the best for the discharge curve of the battery(s) but again only until the voltage of the cell(s) can no longer sufficiently overcome the input voltage requirement of the Buck stage at the required current level.