pretty bad data sheet

Hey, that's the same company that has design kits with parts numbers such as... GRM18-KIT--------E, with helpful notes in the catalog that, yes, the dashes are a part of the kit's SKU, and need to be included in ordering. (E.g., page 5 of

.) One has to wonder who came up with such a brilliant idea...

There oughta be a law that any string of four or more digits (like, say, my checking account number) should be delimited somehow. The telephone company figured this out long ago.

Hey, we could add dashes every 4 characters in the above, turning

GRM18-KIT--------E

into

GRM18-KIT----------E

John

How about refills?

--

    Boris Mohar

They are better at caps than resistors.

Vishay owns many of the SMD resistor manufacturing houses, and they and many of their other brands are the right maker(s) to source from.

The trimpot is cool, though. It's about 0.1" square, and has surprising high frequency performance.

John

I have yet to see a Japanese company that produces good datasheets. Try reading some of the LCD datasheets... Impossible to understand and even when you follow the datasheet, the circuit still does not work

-- Bill Naylor

Electronic Kits for Education and Fun

?%C2%A0&sNHinnm=PVA2A104A01R00&sNhin_key=PVA2A104A01R00&sLang=en&sParam=PVA2

I'd rather avoid surprises, when it comes to passive component selection, specification or sourcing. Cool won't enter into it.

A 100K resistor may tend to be a bit of a disappointment, at RF.

RL

How dull and grown-up all that sounds.

The 100 ohm one rocks.

John

?%C2%A0&sNHinnm=PVA2A104A01R00&sNhin_key=PVA2A104A01R00&sLang=en&sParam=PVA2

Trimpots? Euwww.

I think the last time I spec'd a trimpot into a design was, wait a minute ... ... oh, yes, in the mid 80's. On my masters project where uC servo-ing would have pushed me over budget. But it didn't have to go into production.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

?%C2%A0&sNHinnm=PVA2A104A01R00&sNhin_key=PVA2A104A01R00&sLang=en&sParam=PVA2

Accurately gain trimming a DC-to-1GHz gain path is a perpetual problem around here. Soldering selected 0603 resistors is a nuisance.

Just now, I'm doing a 16-channel ADC board with differential inputs. I want 1 PPM CMRR (120 dB) from DC to maybe 500 KHz. It's going to have set-once CMRR trimpots.

John

e

gn_kit_catal...

Reminds me of a program manager I once worked with who came up with a "configurator" system by which a prospective customer could order a custom display by selecting from a set of letters in each of several positions in the part number. Some of the "configurable" displays thus described would have been impossible to build, others would have had no useful function.

?%C2%A0&sNHinnm=PVA2A104A01R00&sNhin_key=PVA2A104A01R00&sLang=en&sParam=PVA2

I try to do everything with servoed FETs, and PIN diodes for the RF stuff. So far I could get away sans trimpots, it's all done in software. Although the last time I said in a design review "... and that'll just be software" some stuff was thrown at me across the table. The launch pad was the hand of the SW manager.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

?%C2%A0&sNHinnm=PVA2A104A01R00&sNhin_key=PVA2A104A01R00&sLang=en&sParam=PVA2

Fets get nonlinear at small voltages, have bad tc's, and need complicated drivers, dacs and shifters maybe. PINs don't work at DC.

2 GHz is easy. DC-to-2 GHz ain't.

John

?%C2%A0&sNHinnm=PVA2A104A01R00&sNhin_key=PVA2A104A01R00&sLang=en&sParam=PVA2

Non-linearity can be an issue but nowadays 2GHz doesn't necessarily pose a problem for a FET. DC doesn't either :-)

A real cool thing to have would be one of those active laser trim stations from the good old hybrid days. Bzzzzzzzzzt ... open window to let the stench out ... done.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

?%C2%A0&sNHinnm=PVA2A104A01R00&sNhin_key=PVA2A104A01R00&sLang=en&sParam=PVA2

I've tested some NEC NE2509 phemts as grounded variable resistors, with gate voltage adjusting the resistance. They get down to maybe 6 ohms at zero gate voltage. They're good from DC to several GHz. I even characterized the required gate drive vs drain resistance and temperature.

It works pretty well up to about +-20 mV swing. Much more and things get nonlinear. I didn't try any drain-gate feedback tricks, like people do with jfets. It would probably oscillate.

Cruddy lower gain mesfets would probably allow more undistorted voltage swing.

But those little Murata trimpots are so cute! About the size of an

0805 resistor.

John