Analog Squarer {Y=X^2} 1 Quardrant

both inputs tied together.

smoke on overload conditions, so tracking the I^2*T constant of FET and also need a temperature term which is strait forward. The FET current changes 10:1 so the "time on" changes 1:100.

No surprise. Analog Devices seems to be a changed company... management by marketing :-( 'Tis a shame. ...Jim Thompson

third, both inputs tied together.

smoke on overload conditions, so tracking the I^2*T constant of FET and also need a temperature term which is strait forward. The FET current changes 10:1 so the "time on" changes 1:100.

In my _custom_ world I just roll my own, big round bipolar structures to minimize RB and also maximize Vbe matching.

Kind of a fun place... "design" your own transistors.

In the old days, with only one layer of metal I did all kinds of crazy things to make it all fit... "L"-shaped transistors, elongated devices for ease of crossover, you name it, I've done it ;-) ...Jim Thompson

More tricks up your sleeves than Earl Schibe, with a full roll of masking tape... ;-)

hird, both inputs tied together.

ing smoke on overload conditions, so tracking the I^2*T constant of FET and also need a temperature term which is strait forward. The FET current chan ges 10:1 so the "time on" changes 1:100.

Ok JT, You have won a bottle of Oban for your winning squarer circuit usin g the MAT14. It works great over my input range of 100uA to 1.0mA. Advanced Linear Devices has low threshold (Vgs= 0V) matched FETS (ALD1106). Do yo u think they would track just as well?

Highest regards, Harry

third, both inputs tied together.

smoke on overload conditions, so tracking the I^2*T constant of FET and also need a temperature term which is strait forward. The FET current changes 10:1 so the "time on" changes 1:100.

the MAT14. It works great over my input range of 100uA to 1.0mA. Advanced Linear Devices has low threshold (Vgs= 0V) matched FETS (ALD1106). Do you think they would track just as well?

Hey, great, Harry! Glad it helped you!

The trick doesn't work with MOS devices. Bipolar devices have a true (mostly, the degradations are minute :-) logarithmic relationship between voltage and current:

Vbe = kT/q*LOGe(I/Is)

An MOS device connected as a diode has a _mess_ of interrelationships, more like 3rd order with skews thrown in as well.

I've seen bandgaps made using "strong/weak" MOS devices to make a "flat" voltage, but they're mostly crap. ...Jim Thompson

third, both inputs tied together.

smoke on overload conditions, so tracking the I^2*T constant of FET and also need a temperature term which is strait forward. The FET current changes 10:1 so the "time on" changes 1:100.

How do I claim the prize? Sounds marvelous! ...Jim Thompson

There is a bottle waiting for you at Costco. I will refund by PayPal.

Cheers,

Harry

JT, We need a full report on your taste for Oban whiskey and while you are at Costco, buy the Belgian Chocolate Cups from Kirkland. I deem them deli-sh!

Cheers,

Harry