Voltage doubler with ICL7660S

Started by yeeshkul, August 20, 2020, 07:08:57 AM

Previous topic - Next topic

yeeshkul

I am about to use the doubler on the picture below to pump 9V to  18V.
I see 10uF caps are used.

Would bigger caps - like 100-220uF improve the function somehow?





Mark Hammer

Your use of the word "somehow" is open-ended enough that you have no particular expectation.

One of the caveats of charge pumps is that they rely on an internal clock, which can provide audible whine in some circumstances.  Connecting pins 1 and 8 doubles the clock frequency, but that doesn't mean the clock couldn't be audible in some way.  So, like yourself, I'm curious about whether use of other cap values can offer some improvement in that area.  I'll qualify this by stating that I don't actually know if:
a) this is possible
b) the charge pump itself is actually the optimum place to try and reduce whine
c) any change to cap values has to be the same across all such caps (e.g., if tripling voltage, do all 3 caps have to be the same value?)

Like yourself, I eagerly await replies from others.

yeeshkul

#2
Mark, i have been using 100uF/50V between the diodes and 220uF/35V at the output. First i used 56ohm/220u low pass filter at the output, but then i removed the resistor since it wasn't whining at all and the resistor was lowering and softening the output a bit.

I am asking merely because i am courious, the above stated has been working just fine for me. Star grounding helps also.

antonis

IMHO, bigger caps improve (to some extend) load drive current regulation/capability..
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

Mark Hammer

So are the 10uf values we generally see in drawings a recommended minimum value and no sort of absolute requisite?

antonis

I presume so, Mark..
(they are a golden mean between affortable voltage ripple & maximun current..)
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

duck_arse

the Intersil ICL7660 datasheet [3072.4] includes the note:

QuoteNOTE: For large values of C OSC (>1000pF) the values of C1 and C2 should be increased to 100μF.

suggesting only lower pumping frequencies need bigger caps. it also mentions increased cap size benefits can be lost due to increased ESR of the large caps. so low ESR 10uF's might be the best bets.
" Hence the duck effect. "

antonis

Can't see any reason for facing it like something different of "conventional" voltage doubler charge pump..
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

PRR

A couple-hundred uFd might be good for 50/60Hz AC.

This thing runs, what, 10KHz? 20kHz?? So a couple hundred times *smaller* would be to-scale. One or two uFd.

At heavy load, ESR matters. Old e-caps in small uFd could have high ESR. That's why they spec 10uFd (which don't cost more than 2uFd).
  • SUPPORTER

Rob Strand

#9
There's a graph in the data sheet of output droop vs cap size.

Low ESR caps improve most aspects.   See appendix A for ripple,

https://www.maxlinear.com/appnote/ani-19_selectingchargepumpcaps_072406_d.pdf
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.