Podcast: Play in new window | Download
Subscribe: Apple Podcasts | RSS
Today’s episode is sponsored by Keysight, who recently started Keysight University. We hear from Daniel Bogdanoff and Jit Lim about USB, including the specs on the newly announced USB 4.0. To check out the USB 4.0 Pitfalls course and many more classes about test and measurement, check out Keysight University.
Welcome, Dr Brock LaMeres!
- Brock came from industry. He was working at HP and took classes at night to work on his Masters and PhD. The classes in his program had few traditional students. The profs were working engineers.
- After graduating, he stuck around to teach classes and loved it. He wanted to make it part of career.
- Using (modern) CAD tools in school
- Application was tied to the tools
- Chris went to a conference of EE Dept Heads (ECEDHA) and was put off by their unwillingness to try new tools.
- Teaching with captive tools in virtual machines?
- To get the context of the material, you have to show the application.
- Brock likes to show the historical aspect of engineering, or the “Rapid historical evolution of technology”. As an example: the battles between Tesla and Edison (War of the currents)
- Another example is how Intel (“INTegrated ELectronics” made up the name, if you didn’t know) hasn’t really been around that long, historically (<50 years).
- Brock teaches every level at MSU:
- Beginners: EE 101
- Middle of curriculum: Logic circuits and design
- Advanced: Embedded systems
- Grad level: Digital design, including real world problems like signal reflections (Editor’s snarky note: so analog too!)
- MSU has been dealing with COVID-19, like any other higher ed institution. But Brock was already teaching classes online.
- He has written books and paired it with youtube videos, so it wasn’t a huge transition for a lecture class.
- However, there was a range of experience across the campus, especially the profs who weren’t super digital savvy.
- Brock was interested in distance learning 10 years ago, especially addressing problems with the labs.
- He got a grant from the NSF to research how students could access equipment on campus and perform the lab somewhere else. This was before the availability of the low cost instruments of today.
- Chris asked how they measuring engagement / frustration
- It’s a wide range: Students that need full engagement is not 100%. In fact, there’s a percentage that don’t want interaction at all. “They just want to crank”
- Breaking the class into different groups based on learning patterns means more time to engage the students who are struggling.
- Engineering is more about internalizing
- The thing that is missing is self-regulation: “If you’re stuck on something for 30 minutes…then stop”
- They do a large number of low stakes homework assignments to track progress. It helps let them know when students get left behind and he’s not a fan of exams.
- Vocational education vs traditional education.
- Senior design class is a must to be ABET accredited. Students learn that the circuit design is not the problem. It’s the “everything else”
- Brock started working with NASA, continuing on his work designing FPGAs with HP. NASA wants reconfigurable hardware to accelerate computation, but also want to reprogram on the fly. Reconfigurable hardware can also flush out faults in space.
- They came up with a digital platform: Rad PC
- Processor voting on the output. “Voting” = 3 circuits produce outputs: if one crashes, the two that are the same will validate the broken one. This results in triple modular redundancy.
- They have gone up to space a bunch of ways!
- Put on sounding rockets
- Put on the ISS (!!!)
- Two cube sats
- Rad Sat U
- Rad Sat G
- Internally, the hardware has a bunch of monitors all over the place, including error correcting code (ECC).
- Rad hard parts are expensive and hard to get, so they’re trying to use commercial FPGAs.
- Chris asked about the overhead “cost” of having monitoring internally on FPGAs.
- Looking at MTTF
- The mission itself is driving a lot of the requirements
- Other ventures like Planet Labs (past guest), SpaceX (past guests) also are using commercial hardware with smaller mission horizons.
- Any missions requires an analysis of how to burn up.
- They’re currently working on a project to go to the moon! Part of the Artemis Program.
- The computer will use the landers and rovers radios for talking back. Brock’s computer will connect to the Lander and the Lander will take care of all the data (over some satellite uplink).
- Their Cube Sats are amateur band using SATNOGS for base stations.
- New book/course that’s coming out, which is an Intro to Embedded Systems course reboot at MSU.
- Took a different approach, a different way of learning
- Wanted every student to come to class with the micro attached to their laptops, so a $10 dev board/platform was the goal.
- Brock is passionate about reducing costs for students, especially because of the public school charter.
- Previous target was Freescale processor, but the dev boards will still $100.
- The lecture for this class leads right into the lab, hosted in a room with more infrastructure.
- Wrote the book/recorded videos.
- A new book about Embedded Systems is coming out this week!
- Breaking down material into smaller chunks
- Some students are faster than others students
- Chris asked if the students coming out of the class, but Brock says it’s tricky to figure out “are they better”.
- Regular employers for MSU students? Bozeman has a strong optics community, as well as large scale employers like Boeing in Seattle.
- Brock was named a Boeing Professor, which means he got a grant to work on the program.
- Check out more about Brock:
Bill Stacey says
I am not a fan of vocational learning.
Understanding what tools are actually doing is a foundation that changes little as tools evolve. Education is fundamentals, training is tool use.
HOWEVER, if tool or technology vendors publish training content in a structured and template compliant form, (like a software library is structured) an educator can link to the tech vendor content and then only the interface changes need to flow out both ways into the educational content and the tech vendor tool dependent content. This way the educator only references the template compliant package from the tech vendor and as tools change even radically the references from the educational material don’t break.
There is a need for an independent body to sit between these groups maintaining templates and promoting standard approaches to skill acquisition. Crazy idea perhaps but it could solve the problem of the overhead involved in Professors doing the tech vendors job. Also different vendors are forced to present their offering in an industry standard framework so it can be compared, evaluated, and made more accessible.
In summary, its in the interest of tech vendors to actively understand and support the role and skills of educators.
Tech vendors focus mostly on the benefits of their product. They want you to buy their product. Benefits are often high end expert level capabilities.
Whats needed is easy skill acquisition and low complexity. So many tools evolve into massively complex and intimidating fortresses. Little effort is made to support skill acquisition. I often avoid expensive tools because the effort barrier to achieve basis competence is just too high. Most of what we do is actually pretty simple.
When I started writing embedded code everything was easy to understand. Now IDEs are super capable but getting to the blinking LED or ‘hello-world’ point in some cases can take weeks not minutes. Novices get turned off and even worse never get to understand the fundamentals under the hood.
The AmpHour is a forum like no other. I dont know how you can monetise what you are doing but it is valuable to me and our profession.
For example I had a half hour discussion with a consultant today regarding a serious issue raised on TheAmpHour. (Professional Indemnity Insurance, no actuarial data to assess risk and provide meaningful insurance rates and binding valid coverage).
Thanks Chris for all you are doing,
You never mentioned … what was the convenient low power $10 USB-connected microcontroller dev board you selected?
Frank Bolton says
A bit of digging shows it’s an MSP430FR2355 launchpad