Synchronous Switchers

I described that experiment already. If I touch the probe ground lead and the tip to the top of one of my output SMB connectors, I see tons of switcher trash.

My plan is to replace the buck switchers with LTM8023's, plus a lot of extra LC thingies coming and going in the power zone. Once I have RF customers (which I seem to have now) there's no such thing as quiet enough.

The rotating-machinery guys, our intended customers, don't need 120 dB s/n ratios. This arb has the ability to *add* wideband noise to any of the channels, as additive, AM, FM, PM, or PWM noise, and they like that.

z IF

My spectrum analyzer begs to differ.

Larkin == White Trash ...Jim Thompson

induced

grounded SMB

of high

currents.

place,

rotate

Maybe, maybe not. Ever heard of circular / elliptical polarization? There are equivalents in 3-space as well. You can expect to find minima, but not nulls.

Oh, and i expect the spectrum to be similar to d/dx(sin(x)/x) perhaps with a tad of sin(x)/x greatly attenuated.

?-)

make a

copper ground

end of the

is the

higher

the highest

On another subthread, the different harmonics will have different B field geometries, no nulls can be expected, just different spectrums seen.

test

not.

Interesting coming from someone who is never quantitative in this ng.

?-)

voltage

sitting

copper

run, and

switcher

fields fall

to the

That's

Of course not, the problem lies in the shoot thru currents. What do those paths look like? Especially paths to 1000 pF bypass caps.

?-)

the

Mhz IF

Then it would depend some on the particular receiver, many older receivers had a tuned RF before the mixer, and it would take many (about 40) consecutive harmonics to wipe out the 20 MHz wide band. Just one at 10.5 could wipe the whole band by walking through the mixer. What is the relative strengths of the 176th through 216th harmonics versus the 21st harmonic?

?-)

induced

SMB

high

Wrong on both guesses. Try again some other time.

fall

the

sitting

and

switcher

fall

the

Everyone knows that is the magic value a bypass cap should be.

(I'm up to 2.2uF now, used to be 1uF but they got cheaper).

Or is it like all resistors are 10k unless proven otherwise? Although I seem to recall you saying 4k7s are prettier.

sitting

copper

and

switcher

fall

the

That's

Absolutely. It's 330 nF.

Just be careful you don't make your LDOs oscillate. If you have, say,

Alas, we don't use axial color-coded resistors any more. I haven't designed one into a product in years. We use all 1% surface-mount stuff now, and lots of packs.

?? You only help hobbyists?? What in the world is unethical about helping people trying to earn a living? Or is "unethical" a synonym for "doesn't help me any?"

You aren't making much sense today, Tim.

Cheers

Phil Hobbs

He's bluffing about the fix, too.

Yes, it's not like John hasn't given away a lot of stuff too (help, ideas). Carping from the usual suspects aside.

Is it OK for hobbiests to accept advice from professionals? Or for professionals to give it?

I regard myself as a "hobbiest" in some areas, and like to think I am a "professional" in others. I had better pick my questions carefully.

Anyone who tries to assist John "White Trash" Larkin out of his asininity has got to be a bit of a fool. ...Jim Thompson

e:

Mhz IF

.5

The near-field FM-band harmonics' amplitude roughly matched mid-power local FM signals. Since the customer lives in a crowded area with distant stations, the problem seemed pretty clear.

voltage

sitting

copper

and

switcher

fall

the

That's

Most of the time I use local RC decoupling to small circuit sections (rather than a powerplane, say).

Yes thats true. I can still sort of see the colours though. When I first started with SMT I foolishly bought some of the "mini-melf" cylindrical SMT resistors mainly because they *did* have colour codes. Stupid things shoot from the end of your tweezers, then *roll* away.