rotated caps

You _don't_know_? I thought you knew everything ;-) ...Jim Thompson

-- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

google says AVX and samsung, keyword licc capacitor

-Lasse

Almost any large manufacturer does, they call them "wide", "long side", "reversed" and so on. For example, this series is inexpensive:

If you search for SMD capacitors on Digikey, you can select package options that look like this: "0603 (1608 Metric) Wide (Long Side), 0306 (0816 Metric)"

They list AVX, Murata, Taiyo Yuden, and TDK making caps of that sort.

Cheers

Phil Hobbs

Out of curiosity, why/where would you want to use that configuration ?

Less inductance. Specifically, I want to have a bunch of DC blocks in microstrip traces, to pass fast pulses. What we'd been doing is to use several small caps in parallel. For instance, two 0603s side by side on a 60 mil wide trace. I'm thinking I'll get better coupling with a single 0306 spanning the gap; there would be a shorter distance where the signal is traveling in the cap, up above the board, which is an impedance discontinuity. One cap is easier to install than two or three close together.

I'm talking about 35 ps edges; this wouldn't matter for more normal signal speeds.

They might make better bypass caps, too, except that the regular caps seem to work fine for that.

'cause they are "low-ESL"? :)

As John stated, lower ESL (and also ESR). ...Jim Thompson

-- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

Even better are monolayer capacitors. I just have _carefully_ opened a Picosecond Pulse Labs 5828-108 amplifier (22ps risetime):

Sorry, I had only the cell phone camera ready here. The right side of the picture is the input, not the output as i wrote in the picture. On the output side, there is the same structure. They made much ado to get away without conical coils for the output bias (coils and damping resistors).

regards, Gerhard

That is strange. The two 0402s look to be tilted up to rest on the top of the planar cap, which, I would have guessed, would add more inductive loop area than it's worth. But at these speeds, there is art and science.

Here's one of their older bias tees. I guess i didn't _carefully_ open this one.

I think they do use the Piconics gadgets in some of their newer ones.

Since they're wired in parallel with the single-layer cap, I assume that the LC formed by the 0402s is probably low-Z enough not to put a big whoop-de-doo on the transfer function. Pretty slick.

Now _that_ looks like a pain to assemble.

Cheers

Phil Hobbs

Oh, maybe they are in parallel; that makes sense. Does that piece of copper foil hop up from the trace to the top plate of the cap?

When I get some of the rotated caps, I'll make some DC blocks on microstrip and see how they work.

I wonder what that PSPL box costs. It looks like just one MMIC.

That's what it looks like--the left-hand trace goes to the bottom plate of the cap, and the copper foil goes from the top of the cap to the right-hand trace. That puts the 0402s in parallel with it.

If it says PSPL on the box, cheap it ain't. OTOH everything I've ever bought from them (mostly bias tees) worked just like they said it would.

Cheers

Phil Hobbs

The bias tee I posted is just fair, and we bought some of their resistive pickoffs that weren't awfully good. Not everything they do is great.

We did test a board that has the supercheap MiniCircuits surface-mount bias tee, and it's surprisingly good.

That's simple, compared to the tubular filters we built at Microdyne.

If inductance is your concern you can also get there if you hang a 0201 cap in parallel to the usual 0603. Capacitance doesn't matter too much, maybe 1000pF or 0.01uF, whatever you have at hand.

Yes, but there is not much hopping. The single layer cap is about as thick as a sheet of paper.

I wonder what that MMIC might be. The rest should be easy now. BTW, there is some absorber rubber on the top of the enclosure.

Gerhard

Production would lynch me. I hear dangerous grumbling from 0402s.

So did I, until I actually did it. They had zero problems with 0402s. Never did 0201s because the pick-n-place machine wasn't up to them. Where I am now.