Sam Zeloof: Who Needs a Cleanroom When You Have a Garage?

At the end of last year, I was mindlessly looking at YouTube when their "algorithm" decided that I should watch a video made by a guy named Sam Zeloof. He was literally building integrated circuits in his garage using cobbled-together or obsolete equipment. Of course, this was fascinating to me and I ended up watching much of his channel. It was well after midnight I went to bed.

I know that in the early days of semiconductor manufacturing (in the 1960s) that there was no equipment industry and so engineers needed to do everything themselves: create the process, design the layout, build the equipment, run the equipment, and so on. A modern 60,000 wafers per month automated fab this was not! But it seemed somewhat similar to those early days.

I started to dig a little deeper and it turns out that Wired magazine scooped me and published an article on him late in January, just a couple of months ago, This 22-Year-Old Builds Chips in His Parents’ Garage. So it turns out the title to this post is wrong since he doesn't even have his own garage!

Here's a quote from that piece:

The same month, 22-year-old Sam Zeloof announced his own semiconductor milestone. It was achieved alone in his family’s New Jersey garage, about 30 miles from where the first transistor was made at Bell Labs in 1947. With a collection of salvaged and homemade equipment, Zeloof produced a chip with 1,200 transistors. He had sliced up wafers of silicon, patterned them with microscopic designs using ultraviolet light, and dunked them in acid by hand, documenting the process on YouTube and his blog. 

Here's his video about this chip. As he points out in the video, he is putting over a thousand transistors on a single piece of silicon, and Intel's first microprocessor was only two-thousand transistors (actually 2,300), so he is in the same ballpark. Only one node behind the first microprocessor.

Some of the equipment he created himself, and some he purchased on eBay from obsolete fabs. One of his best buys was a $250,000 electron microscope (broken) that he purchased for $1,000 and repaired. It weighs a ton so just moving it was a challenge.

Obviously, since he is in a garage, he doesn't have modern clean-room air cleaning. But he went back to old patents and academic papers from the early days of integrated circuits. They didn't have clean rooms either. Often, they would design the "masks" for manufacture using hobbyist X-acto knives and a red plastic film called Rubylith, a far step from making modern EUV reflective masks.

His first chip, with just a few transistors, was 175um with non-self-aligned metal (aluminum) gates. The second, the thousand transistor one in the video above, is more like 10um with self-aligned polysilicon gates. He worked out how to do that without using silane. In his blog, he says it is explosive, but actually, it is just pyrophoric, it catches fire on contact with air. When I was at VLSI a maintenance worker cut the wrong pipe, silane instead of something boring like nitrogen. Luckily, someone watching was really knowledgeable, ran outside, and flipped the circuit breakers for the silane pumps. So it was just a minor fire.

Here is his semi-serious semi-joking graph from his blog showing how much faster than Moore's Law he is improving. The blue industry dots are the datapoints from Gordon Moore's original paper in Electronics Magazine.

Here's a good summary quote from Sam's blog:

Now we know that it’s possible to make really good transistors with impure chemicals, no cleanroom, and homemade equipment. Of course, yield and process repeatability are diminished. I’ll do more testing to collect data on the statistics and variability of FET properties but it’s looking good!

Videos

Sam has lots of videos. I think that the most interesting ones are the technical ones about individual process steps. Here, for example, his maskless photolithography stepper using a modified DLP projector (the sort of thing used for Powerpoint presentations in conference rooms before we all went on Zoom):

And here's one on his home-made plasma ion etcher:

If you enjoyed these, there are quite a lot more on his channel (link below).

Learn More

Sam's Twitter feed.

Sam's YouTube channel.

Sam's blog.

Sign up for Sunday Brunch, the weekly Breakfast Bytes email.

.