1G resistor

50fF up from 6fF in your measurement sounds like a big 'step' in the right direction. Certainly no harm in it.

Me neither--my motto is "Why/never wonder when you can measure."

But we can objectively calculate just removing the FR4 between the 1G resistor and a ground plane beneath is at least e_r(FR4) = 4.8x better just from the flat-plate capacitor approximation.

Removing the groundplane is even better, forcing a longer path for the remaining electric field.

Sure. The only certain way for that, though, is to Dremel a working divider proto first.

(P.S. I rather liked the idea of soldering the 2512 1G on its edge--that's almost as cryptic as your opto-cap. :-)

(P.P.S. Fat dummy caps could be soldered on board as spacers. The 1G could then be soldered air-borne, on top!!! :-). / / / / / .-.-------------.-. /-. | | | |/-/| '-'-------------'-'-/|/ .-.// .-./|/ | |/ | | / '-' '-'' Cheers, James Arthur

There are surface-mount test points (little loops, actually) that would serve, as well.

It'd be easier, though, if one could find a few-kV rated socket (fuse holder?) and use a glass-encapsulated high-ohms resistor. Compact wiring and high voltage are a disquieting pair.

OK, you come here and explain that to our production manager.

ps - He's bigger than you are.

Seriously, I checked a few boards to estimate the air gap under parts caused by the solder. I see some parts maybe 5 mils up, and other parts with no visible gap.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Ceramic caps (and electrolytics) seem to be self-balancing. They leak nonlinearly at higher voltages. Series film caps can blow up.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

-Lasse

OOH! That '-5' option with the little J-lead board standoffs looks ideal! I was thinking of old-school hardware, like PTFE inserts with hand wiring. If only these were stocked in the 1G sizes (my next prototype is unlikely to wait for special-order parts).