Supply-cap ESR

[wiss]

In an old style linear power-supply:
Will increased ESR in supply-caps increase the total power consumed by a device?
Assuming that the (regulated) current required does not increase, think rectified to 10 V and feed through a 7805 for example.

I could probably spend some time calculating this but it's probably faster and more accurate to ask her

[T3sl4co1l]

A little, due to increased ripple.  The minimum peak voltage (dropout) also drops, which eventually leads to ripple passing through the regulator.  Not to mention capacitance usually falls at the same time ESR rises, making things that much worse.

I think there's a small amount of runaway, in that, as a cap dries out, its ESR rises, making it hotter, making it dry out faster...

Tim

[wiss]

Mmm, pretty much what I guessed,  ofcourse the cap will run hotter and the regulator cooler but no major total change...

[lewis]

A little, due to increased ripple.

Increased ripple actually reduces power dissipation in a linear regulator because the RMS voltage across the regulator is lower. This might compensate for any power loss in the capacitor due to its increased ESR, someone better at maths than me will be able to tell you!

[wiss]

But then the charging-pulses will be larger generating more heat in the transformer. But how much? 

[T3sl4co1l]

Mmm, pretty much what I guessed,  ofcourse the cap will run hotter and the regulator cooler but no major total change...

No, I think total will go up slightly.  The peaks are much higher, and the valleys lower; the transformer's RMS current is higher than the DC load current, so there's more loss there.  (The RMS/DC ratio will be lower, but always above 1, unless active PFC is used.)

Total dissipation may drop if capacitance rises, causing ripple to rise even further and making dropout a bigger issue.

Tim