Teo Bergkvist

Reflections on a Test Beam

Thu, 13 Aug 2026 00:00:00 +0000

I spent a week at a test beam at CERN, here are my reflections on the experience.

First of all, if you are not working in particle physics, you might wonder what a test beam is.

Imagine a one-week hackathon, but instead of getting some credits for AI usage and being forced to use some startup's shitty SaaS prototype, you get a particle beam from the Proton Synchrotron. If that doesn't sound cool enough, you also get an experimental area enclosed by thick concrete blocks to protect everyone outside from the radiation from the beam, as well as a control room with measurement equipment connected to both the Proton Synchrotron beam data and experimental area equipment.

Inside such a control room, I spent most of the past week, solving hard problems together with incredible engineers and physicists.

With the larger goal of testing the new inner tracker for the high-luminosity upgrade of CMS, we also tested a lot of other subsystems, and got some real physics data. Throughout the week, every 20 seconds two beeps sounded in the control room, signalling a spill: the beam sending particles through our test setup. I still hear the beeps in my head.

Some reflections:

No matter how prepared you are before, your stuff will break. And if it doesn't break, somebody else's stuff will break, breaking your stuff.
Debugging and fixing things under extreme time pressure is hard, but also very fun.
Human communication is inefficient and prone to errors. This might be one of the largest bottlenecks during a test beam.
Logbooks are good. Hacking at 2am feeling like a boss will not feel good the morning after unless you documented what you did.
Keeping a good mood, even during bad times, might be the most important skill.

I'll finish this post with some poetry.

Just as computers run to the beat of the clock, humans at test beams run to the periodic double-beeps of beam spills.

Got any long cables?

Thu, 09 Jan 2025 00:00:00 +0000

Today I wrote my first article on medium. It is about using a NanoVNA (a portable type of Vector Network Analyzer) and Python to measure the length of a cable. It is a fun project with the goal of getting more people to use NanoVNAs and the Python module pynanovna.

You can read the article here.

PICC Group Winning 1st Place!

Thu, 08 Aug 2024 00:00:00 +0000

We ended up winning 1st place in the IEEE Student Design Contest 2024 at the AP-S URSI conference in Florence! It has been amazing leading the team during the past year. Now it is time for new adventures!

Artificial Neural Networks trained by a Genetic Algorithm to play... Pong?!

Tue, 28 May 2024 00:00:00 +0000

Last night, I sat up working on my final project for a computational biology course. The project is an attempt to train artificial neural networks to play pong, by letting them compete with each other in a genetic algorithm. The results were surprisingly good, and the project is on GitHub for anyone interested: https://github.com/tbergkvist/ANN-GA-pong

PICC Progress

Sat, 10 Feb 2024 00:00:00 +0000

We have started working with the hardware of the PICC setup now. Yesterday we also sent out requests for sponsorships and grants to a number of companies and foundations. Let's hope we can raise enough to go to Florence!

The PICC Group

Mon, 29 Jan 2024 00:00:00 +0000

I am happy to announce that I will, along with 3 fellow students, participate as a finalist in the IEEE AP-S Student Design Contest. We call ourselves the PICC group (Polarization and Impedance Controlled Car) and we will present an educational and fancy way to control an RC car with on-body antennas. We will compete against 5 other universities in Florence, Italy, in July. https://2024.apsursi.org/sdc.php