Lab-in-a-box for at-home-because-COVID students

[TheUnnamedNewbie]

I work at a university, where part of my job is being the TA of a practical course in the electrical engineering bachelors. This course used to involve teams of 6 to 8 students that work together two afternoons a week while I just hang around in the labroom to help guide them to solutions.

In this lab the students have access to all gear they need - Supplies, scopes, AWG, mechanical tools like screwdrivers and pliers and soldering irons, etc.

With COVID we want to avoid having all students come to the class room at the same time, and instead enable half to work from home. Because it is not feasable for them to take an entire kit of gear home with them each session (simply because all that gear is too heavy to carry on their bike) I was considering some lab-in-a-box type thing like Digitent's Analog Discovery 2.

I was hoping we could share some experiences and insights and ideas on this kind of tool, what others are out there, what are some of the caviats, etc...

[TurboTom]

You may also want to have a look at Analog's dependance, the ADALM2000. Additionally, they offer a component package that includes a reasonable amount of semiconductors, sensors, passives along with a breadboard and related stuff.

This combination should enable the EE student to acquire a lot of hands-on experience at home, at a considerable lower price than the AD2.

[jogri]

What's the main goal of this course? Getting the students familiar with practical lab work? If that's the case, why not let them build a few simple circuits like a analog function generator (USB powered) and an adapter so that they can use the mic input of a soundcard as a simple oscilloscope (software for that is already out there). They could build the soundcard scope in the lab as the first project (so that someone can show them how to solder) and build the rest at home as they now have a "scope" to troubleshoot their circuit. Low cost, and they even get some really usefull kit for starting their own lab at home.

[TheUnnamedNewbie]

What's the main goal of this course? Getting the students familiar with practical lab work? If that's the case, why not let them build a few simple circuits like a analog function generator (USB powered) and an adapter so that they can use the mic input of a soundcard as a simple oscilloscope (software for that is already out there). They could build the soundcard scope in the lab as the first project (so that someone can show them how to solder) and build the rest at home as they now have a "scope" to troubleshoot their circuit. Low cost, and they even get some really usefull kit for starting their own lab at home.

The main goal is learning the process of engineering - turning a (vague) demand into a solution. It's not really to familiarize themselves with practical lab work. The assignment is fixed already (as we want to keep it the same over the years), and I want to change the assignment as little as possible. The current assignment is to 'digitize' the game of Kubb, adding sensors and so on to it. It's a totally pointless project from a 'real-world' point of view, but I like it because it enables them to do a lot of different parts of electrical engineering - digital circuits and embedded programing for reading out accelerometers in the thrown stick, power-management for charging the different blocks, analog circuits for some of the sensors to detect if a block was hit, etc...

So the main goal I have is for them to just be able to do some work at home while they aren't on-campus.

[ResistorRob]

Last time I checked Analog Discovery 2 was sold out everywhere. I think they are having delays in sourcing parts or something.

[alextwin007]

This is a difficult situation.   I graduated a few years ago, and they tried to push on us to use a myDAC, buto one liked them.  That being said I now have an Analog Devices Analog Discovery (1, not 2) and I find that it is much better.  I've tried the JYETech DSO-SHELL and some other cheap digital oscilloscope kits.  They do work in a pinch, but they aren't as nice to use, and I am always wondering if I should trust them.  Xprotolab Plain is also in a similar category.  I find that the UI in all of these are kinda muh, especially when I was spoiled by using Keysight 2000X scopes in school.  The only one I really use now is one that has a battery, just because it's isolated and easy to grab.   

I also found that having some type of multimeter very helpful, many of the cheap oscilloscope can perform this function, but I found that even a 10 buck meter was very helpful.  Also one of the cheap atmegta 328 based LCR like meters is a very useful and comes as a kit.  Unfortunately there aren't many cheap multimeter kits, for that I'd say you'd be better off just buying one assembled, as they are much nicer to use, and far cheaper.

I like the idea of building some tools such as a function gen, and a power supply.  Before I had a nice function gen I used one of the XR2206 kits. A simple LM317 supply can get you quite far, as long as you don't need anything below 1.2V.  For power supplies you'll have to make sure there is enough power for all the tools they might need.  Things like the XR2260 need a 9V supply, I found that having a 9V clip to barrel jack very useful for powering most kit tools.  Batteries can also be a cheap way to get -voltages if you need them for op amps

As someone who recently set up my own personal lab, something else that you need to consider is supplies.  Having a stockpile of resistors, capacitors, and your go to IC (op amp, voltage regulators) are a must.  Assume students may fry a few.  Now if you are giving them labs to do, you can have them pick up goodie bags with all the parts they need.  Also some way to connect them such as bread board.

There is all the necessary tools as well.  Wire, wire strippers, flush cutters, tweezers.

All in all it's quite a lot, but when I was setting things up, the first things I had and everything else I got afterward were more quality of life then absolutely necessary.

analog discovery / other oscilloscopes
Power supply,
multimeter (separate from the oscilloscope)
function gen

wire strippers
flush cutters

Consumables
resistor kit
capacitor kit (ceramic and electrolytic)
Protoboard/ breadboard

Soldering  (this may be better to have students share)
There are some soldering kits that are relatively cheap that have all these components.  They might be a good solution.  You could even get a few and let students check them out (although parts may go missing)

iron (some sort of temp control is a must)
wick
sucker
solder
tweezers
iron holder

I have been setting up my lab from scratch over the pass 3 years so I've put a lot of thought into things like this.  I'd be happy to answer any questions you may have.

[Kleinstein]

For such a µC project quite a bit of the work is done without the actual hardware. Much is the upfront planing and also the programming - getting familiar with the µC. Depending on the µC choice, one could do quite a bit with something like a simple µC board (e.g. Arduino,  ARM based board like the STM32 nucleo or similar). The actual lab part using a scope of similar should be a relatively small fraction - but could be more if there are tricky problems / bugs. So the problem may be that nobody really needs the lab for the first half and everybody want's some time near the end, middle, when analog things go wrong.
Much of the work could be done with a µC board and breadboard - though this is likely not the EE standard anymore. It would be somewhat unfair in some groups having access to a private scope and others not. Some low end equipment available for the students may not change that much either. Even if the private scope is old or at the low end - if one knows it well it can be quite some advantage to getting access to a scope one is not familiar with.

For such a project 6-8 students looks like rather large groups. This may lead to splitting the project to that nobody knows all the parts - this may be close to real life work, but the students would learn different things with the tendency to do what they already know best.  So the analog fan would not touch the µC and the more µC affine would stay digital not touching a scope.

With so large groups chances would be that there is quite some private owned instrumentation anyway. So it may not need than many instruments on loan or groups that actually need the lab equipment for the basic tasks. It may be different if the task includes EMI compliance.

[TheUnnamedNewbie]

For such a µC project quite a bit of the work is done without the actual hardware. Much is the upfront planing and also the programming - getting familiar with the µC. Depending on the µC choice, one could do quite a bit with something like a simple µC board (e.g. Arduino,  ARM based board like the STM32 nucleo or similar). The actual lab part using a scope of similar should be a relatively small fraction - but could be more if there are tricky problems / bugs. So the problem may be that nobody really needs the lab for the first half and everybody want's some time near the end, middle, when analog things go wrong.
Much of the work could be done with a µC board and breadboard - though this is likely not the EE standard anymore. It would be somewhat unfair in some groups having access to a private scope and others not. Some low end equipment available for the students may not change that much either. Even if the private scope is old or at the low end - if one knows it well it can be quite some advantage to getting access to a scope one is not familiar with.

For such a project 6-8 students looks like rather large groups. This may lead to splitting the project to that nobody knows all the parts - this may be close to real life work, but the students would learn different things with the tendency to do what they already know best.  So the analog fan would not touch the µC and the more µC affine would stay digital not touching a scope.

With so large groups chances would be that there is quite some private owned instrumentation anyway. So it may not need than many instruments on loan or groups that actually need the lab equipment for the basic tasks. It may be different if the task includes EMI compliance.

Most of the work is still without microprocessors - only the motion-control part is using a uC. And we are aware that not everyone will get to do everything. This assignment does not have the goal teaching them electronics in general - that is what the 4 remaining years of electrical engineering courses are for. The goal of these practical projects (they have one every year except for the last year of the masters, where they have a thesis instead) is precisely learning to divide up a big project, be resourceful in solving problems, and think like engineers.

I have been teaching this course for a few years now, and I have had one student in the past that had anything more than a simple multimeter in their private collection, which is why I can't rely on them having it, and don't really expect anyone to have a lab at home already.

I like the idea of building some tools such as a function gen, and a power supply.  Before I had a nice function gen I used one of the XR2206 kits. A simple LM317 supply can get you quite far, as long as you don't need anything below 1.2V.  For power supplies you'll have to make sure there is enough power for all the tools they might need.  Things like the XR2260 need a 9V supply, I found that having a 9V clip to barrel jack very useful for powering most kit tools.  Batteries can also be a cheap way to get -voltages if you need them for op amps

This might have not been clear enough from the original question I asked, but the assignment is fixed, and as cool as it might sound to change it up, this is an assignment that has been roughly stable over the last 10 years, and it has gone through a lot of iterations to optimize and change it. Going into the why the assignment is what it is remains outside of the scope of this discussion, and would require a thorough discussion of what topics the students are familiar with, how much time they have, etc etc.

Just to be clear here - the students will still come to the university building every other week where they will have access to 'proper' EE lab tools. This lab-in-a-box is just so they can more easily debug their circuits from home on the 'off' weeks

[coppercone2]

how to save boat loads of money in any home kit :

use batteries and LDO. Provide any money saved on PSU to measurement equipment of any kind. It will cost you 10% of the cheapest PSU solution and you don't need to worry about safety, other then a fuse

[Kleinstein]

One can do quite a bit with limited instrumentation (e.g without a scope or a very limited one). However it gets slower and needs experience. So it may be nice challenge, but not so much a real life situation.
I would expect the main item missing for debugging at home would be a scope, but it could also be access to things like resistors and capacitors or other parts.

Quite a lot can be done with a little more thinking and simulations in the office instead of time in the lab. There is a saying : a week in the lab may safe you a hour in the library.

Having lab access every other week could be a little long in between - something like every other days may be more productive. Often if would be just 1 hour in the lab, when you actually need it. So maybe have 1 hours in the evening reserved for those occasional cases when they are normally off that week. Knowing to have a limited lab time may force to plan up front and could even be more productive.