Circuit Ideas?

[mypasswordis]

I'm taking a Microelectronics lab course and we need to pick a particular not overly complicated circuit we think is interesting that is in the literature to build and test. If necessary, we are allowed to make use of the VLSI fabrication laboratory but have to design the layout ourselves. Any cool ideas people have for this project?

Edit: Forgot to say that unless we provide the transistors/opamps/etc ourselves, we are using bog standard parts. 2N390x BJTs and ancient 4007 MOSFETs. Opamps are some newer version of the 1458.

Edited February 1, 2011 by mypasswordis

[kevin gilmore]

HR2?

design is on headwize

[mypasswordis]

No, I need to pick something I think is interesting. I was thinking on doing some tests on a cascoded differential amp with current source loads (hm, wonder where that idea came from) or an RF power amp or maybe something completely different, just trying to get some ideas up in the air first.

[mypasswordis]

Something like this might be cool, since I want to try making an IC.

IC cascode crystal oscillator.PDF

[Craig Sawyers]

How about something based on a Gilbert Cell? This is a truly wonderful circuit that can be used for everything from four-quadrant multipliers, phase detectors, ultra-high bandwidth amplifiers etc.

Developed by Barrie Gilbert while working for Tektronix, and the basis of may of Tek's custom high bandwidth hybrid and integrated circuit amplifiers. Still active and working now for Analog Devices in his 70's (AD set up a lab in Beaverton just so they could poach Gilbert from Tek).

http://en.wikipedia.org/wiki/Gilbert_cell

[Craig Sawyers]

Here's an excellent article about Barrie Gilbert http://www.ieee.org/portal/cms_docs_societies/sscs/PrintEditions/200710.pdf

[mypasswordis]

Ooh that definitely looks interesting. Seems like parts of the IEEE site are down right now so I'll have to wait a bit to read the article.

[mypasswordis]

So I wanted to build an updated IC Gilbert mixer but apparently we don't have scope that measure at high enough frequencies (20-80GHz range) so I'm trying to build the original circuit. I'm wondering about the opamp circuit though, any ideas? Should I just leave it out and use an external purpose-built opamp or layout my own opamp circuit into the IC? Going to be doing the layout in MAGIC.