small resistor in series with bypass capacitor

In the schematic below, what is the purpose of the R1 and R5 that are in series with the bypass capacitor to the ground.

formatting link

TIA

-Mahen

Reply to
Mahen K
Loading thread data ...

Both the by-pass capacitors - C1 and C5 - that have the series resistor are 33uF parts.

It could be that the capacitors need to be protected against in-rush current when the battery is plugged in, or it could be that the battery lead inductance is high enough that R1 and R5 are required to critically damp the resonance between this inductance and the filter capacitance. This implies an inductance of the order of 100nH which isn't unreasonable.

-- Bill Sloman, Nijmegen

Reply to
Bill Sloman

Probably to add some small ESR (effective series resistance) to the cap to keep the loop stable. Some supply chips especially LDOs require some ESR in their output capacitance and some actually specify particular manufacturers for that purpose.

Mark.

Reply to
markp

None. This person is obviously a victim of one of the stranger bypass religions.

John

Reply to
John Larkin

Specified ESR of a small cheap electrolytic 33uF 10V capacitor would be an ohm. Yes, there are ones better and ones that are worse. Tantalums are gonna have a lower ESR, there are some 33uF 10V tantalums that are less than 0.1 ohms ESR.

Some folks in the past have used tantalums as power bypass caps and had them blow up on power-up. Some have recommended putting small resistors in series to prevent blowing up the tantalums but that's stupid because you put a tantalum in to begin with because low ESR was important. Like "the old lady who swallowed a fly".

Tim.

Reply to
Tim Shoppa

It's stupid unless you put the resistor in series with the input supply and take the output across the cap. You lose a bit in the regulation department, of course.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
 Click to see the full signature
Reply to
Phil Hobbs

Everybody else pretty well stated the obvious its stupid.

Ideally for bypass/ decoupling you want an ideal capacitor these don't exist so you use the next closest thing polymer for bulk and low esl / esr ceramics and/ or arrays.

An example of a low esl cap.

formatting link

The cheaper alternative is 0603 or smaller general purpose MLCC's and standard Aluminums or Tants for bulk.

Reply to
Hammy

It can help to have some ESR in the bulk bypass cap path even if one is not using an LDO which requires this for stability.

If you build a bypass system with only low-ESR capacitors, you can end up with a resonance effect - the capacitance interacts with the inductance of the PC-board traces, creating an LC tank circuit (or a set of them) which can resonate at one or more frequencies. This resonance reduces the effectiveness of the bypassing, and can increase noise and voltage fluctuations at the resonant frequencies involved. The "better" your bypass capacitors (the lower their ESR) the higher the Q of the resonance.

Adding bulk bypass capacitance, with a modest amount of ESR in the cap (or in a resistor in series with it) "swamps" the resonance, reducing the Q, and thus decreasing the noise level and voltage-fluctuation problems. Once again, if your bulk bypass caps are "too good" (e.g. tantalums with low ESR) adding a separate swamping resistor can help matters.

--
Dave Platt                                    AE6EO
Friends of Jade Warrior home page:  http://www.radagast.org/jade-warrior
 Click to see the full signature
Reply to
Dave Platt

Not neccessarily. The series inductance of the 33u happily resonates with the neighboring 0.1u and this is a parallel resonance. The better the 33u is Ohm-wise, the worse the decoupling is at the resonant frequency.

Watch page 6 or 18 of this:

regards, Gerhard

Reply to
Gerhard Hoffmann

Very interesting! Thanks for posting this.

--
Rich Webb     Norfolk, VA
Reply to
Rich Webb

Be on the watch for Larkin to jump in here and tell you that, if you have a power and ground plane, none of what's in that presentation (interesting thing that it is) really applies very much: The plane structure represents a very broad, low reasonance source.

Reply to
Joel Koltner

a

thing

ry

At low frequencies - an if the original circuit designer knew what they were doing, that 0.1 ohm would damp a resonance at around a couple of MHz - the power and ground planes would present a simple capacitance and wouldn't provide any useful damping.

-- Bill Sloman, Nijmegen

Reply to
Bill Sloman

That is used as a snubber, to reduce transient pulses. The cap is what responds to the narrow timed pulse and the R limits the max current.

Since there is a zener there to cover most of the over voltage on that input, conventional zeners may not be fast enough to respond to some transient pulses, so, a snubber in this case helps with the initial pulse.

In automotive circuits, noise from the ignition system would most likely get absorbed in this circuit..

Reply to
Jamie

I think you would be better off inserting a small ferrite bead with a small DCR if you have SRF problems? That is what I've been doing beads are cheap and small. Or at the very least CRC pi filter instead.

To me it makes no sense to place a small R 1 ohm 0.1 ohm. Then you may just be going from SRF problem to noisy rail large spikes generated by the esr every time the gates switch.

Reply to
Hammy

OK, here it is:

A 0.1 uF cap, TDRd along an asymmetric coplanar waveguide (no via parasitics) at various time scales. Note the ringing at about 1 GHz.

ftp://jjlarkin.lmi.net/TDR_0.1uF.JPG

ftp://jjlarkin.lmi.net/TDR_0.1_slow.JPG

And here's the TDR of a 3.3 volt power pour on a board we received just this morning:

ftp://jjlarkin.lmi.net/TDR_3Vplane.JPG

The positive pulse is the 50 ohm hardline from the sampling head to the board. The drop to near zero ohms is indistinguishable from a hard-shorted SMA of similar geometry. The plane looks like an essentially perfect 2.3 nF capacitor. Based on experience with similar boards, loading bypass caps anywhere will make it look like a bigger perfect capacitor.

On a multilayer board with good power planes, bypass cap resonances will be undetectable.

I plan to use the SMA on the plane to check plane noise when the board is running, too.

John

Reply to
John Larkin

Very nice job John, thanks for posting those results!

Reply to
Joel Koltner

Actually, if you look at that last shot, at the little flat spot just as the coax impedance transitions onto the power plane, one could suspect that the current is propagating into the plane as a point-driven sheet transmission line, around 1 ohm maybe, and the little wiggles in the flat region are the various reflections off the sides and corners of the board.

After all the reflections die down, the plane acts like a capacitor, giving the charging ramp to the right.

John

Reply to
John Larkin

It's to move the output esr zero lower to stabalize the loop regards, Bob N9NEO

Reply to
Yzordderrex

Does the shape of the wiggles change if you handle the board? (E.g., run a few fingers along the edges.)

Reply to
Joel Koltner

Our copper doesn't extend to the board edge, so touching the edge has no effect. The 3.3 volt plane is layer 2, so putting my hand on the top adds a little capacitance and changes the assumed reflection pattern a little bit, but not in any obviously causal way.

The SMA-connected TDR doesn't work extremely well trying to TDR what may be a sub-ohm sheet transmission line structure. This is a *very* low impedance structure and I'm guessing that Z drops with frequency as the "splash" spreads out radially from the SMA. Most of the features are in the 3-10 GHz sort of range, way higher than the SRF of bypass caps and vias.

I'll repeat it on the loaded board. I should have the first one this afternoon, just about 24 hours after the bare board was received.

John

Reply to
John Larkin

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.