Thermocouple mux

Easier said than done!

Nah, you're just grumpy because it's the end of November and you're in Toronto. We emigre Canucks get two Thanksgivings a year, so the holidays run from the second Monday in October to whenever New Year's wears off. ;)

Cheers

Phil Hobbs

A BFT25 is a far better picoamp diode than a PAD-1.

Yep. If you dig wayyy back in sed(*), I think you'll find I predicted that, you suspected it, then you quickly tested, confirmed, and put it to use.

But, that's a diode-connected transistor. All my transistor-connected diodes(**) used two diodes, and had gains less than one.

(*) 5/2/09 "PAD5", 6/12/09 "Collector-base diode..." (**) inventions of my youth.

is

Yeah, I bet a lot of us tried that one. Two 1N60* diodes don't have much gain.

• dating myself, but the equipment was old enough to be scrapped even then.

Spehro Pefhany a écrit :

When I was a kid (11 or 12) when I read how a transistor was done I had to try a transistor-connected-diodes-pair...

Do I need to say how that worked? :-)

If I were that same kid nowadays I guess I'd be even more confused. MMBT3906 vs BAV70: same package, same multimeter test result but not same working...

snipped-for-privacy@yahoo.com a écrit :

The OA70s germanium transistor didn't either.

Seems to be a pattern there...

Hey, I did that too!

Me too! I thought there must be some trick; so I used zener as C-B junction.

VLV

Well, not much actual use. I designed a killer FTMS preamp in May of

Back-to-back LEDs would have gain. We were just before our time.

You must be _very_ lonely, if you have to date yorself. ;-)

But I've always kinda been partial to calling myself up on the phone and asking myself out, you know? Oh yeah, you call yourself up too huh? Yeah, well one thing about it, your always around. Yeah I know, yeah you ask yourself out, you know, some class joint somewhere. The Burrito King or something, you know. Well I ain't cheap you know. Take yourself out for a couple of drinks maybe. Then there'd be some provocative conversation on the way home. Park in front of the house you know. Oh yeah, you smoothly put a little nice music on, maybe you put on like uh, you know, like shopping music, something thats not too interruptive you know and then uh slide over real nice and say 'Oh I think you have something in your eye'.

-- Tom Waits

(Burrito King in LA is pretty good, IMHO).

is

i(c) = 5 fA, Vbe=0. Nice. I bet i(cbo) is even lower.

And still are.

Probably, but it's getting hard to measure.

Capacitance and series resistance are also much lower than a PAD-1, which is actually a jfet with the source and drain shorted. The jfet channel resistance sucks.

Il 26/11/2012 21:58, Spehro Pefhany ha scritto:

Thank you for your clarification! Marco

First question, how often do you want to switch versus the full settling time of the meter? BTW nothing is all that bad with magnetically latching to-99 relays for this application.

?-)

Il 28/11/2012 10:26, josephkk ha scritto:

I need to acquire both tcs every second. What do you mean about the settling time? The electrical settling time should be negligible, and the thermal one shouldn't matter.

I'm going to use the solution with an AD773 to easily read both tcs and an ambient temperature sensor for cold-junction compensation.

Marco

Just out of interest I googled the AD773.

AD says its discontinued, may not matter if your're doing a one off - just thought I'd pass it on.