Pressure-sensor lab went well

The pressure-sensor lab went fairly well today.  We once again borrowed soldering irons from the RF lab (where my co-instructor teaches), but we did not borrow enough board holders, so a number of students had to solder with the boards just resting on the benchtop.  This was not a major problem, but having more board holders would be good.

Many of the students had done an adequate job of doing the design ahead of time, and even those who had not mostly managed to finish within 4 hours (but the lab is supposed to be scheduled for 3 hours).  Most groups managed to demonstrate working PC boards with instrumentation amplifiers amplifying the differential output of the strain-gauge pressure sensors, generally with gains in the range 100 to 300. I have one group coming in on Monday to finish the soldering—they took the more cautious approach of debugging on a breadboard first, and had to do a little redesign to make everything work.  I think that they’ll complete with no problems on Monday, as they had good notes on what their breadboarded circuit was, and had come up with what looked like a feasible layout.

Soldering was much more routine this time than on the first soldering lab.  Students were able to check their own work for cold-soldered joints, and only one group forgot to trim the extra wire length on the back of the board after soldering (they had a short when they put the board down, from two untrimmed leads touching).

Some students called me over to help them debug at one point, but I had to refuse, as they did not have  readable schematics that I could help them debug from.  After I told them why I couldn’t help them, they redrew their schematics, and I then helped them check the design against their soldered board.  It seems that they had soldered everything up correctly according to their schematic, and the DC voltages were right at all the nodes, but the twisted-wire cable to the pressure sensor had the wires scrambled. (Luckily, they had followed my advice of using 4 different colors for the wires, though they had not followed the red=+5v, black=0v convention—several students are still using colors at random.)  When they screwed the wires into the right terminals, their circuit worked.  It seems that schematics were not the only thing that they were sloppy with!

Some of the students used PDF markup tools to add the layout of their extra wires and parts to the prototyping board PDF worksheet.  That seems to have worked well for them, producing neater and more easily checked layouts that the pencil scrawls that I (and many of the students) used.  I hope they tell me in their lab reports what tool they used, so that I can recommend it for the EKG lab and for the pressure-sensor lab next year.  No one did the wire lists that I recommended, but there were relatively few wiring errors (and those were inherited from errors in the schematics, so wiring lists would not have helped).  I think I’ll leave the wiring lists out of the assignment next year.

One thing that surprised me (another moment of culture shock), was that the seniors in bioengineering did not know what a peristaltic pump was.  I was trying to connect what they were doing to something they already knew, only to find out that what I thought was familiar to them was novel.  I demonstrated the basic principle for a couple of them by hooking two ends of the flexible tubing up to the two ports of their differential sensor, and pinching the tube between a pen and the benchtop.  By pulling the tube through the squeezed area, I could get a large pressure difference between the ports.  Since peristaltic pumps are standard lab equipment in many of the labs they are working in, I was surprised that they have never used one or even known of their existence.  I’m now wondering whether I should do the demo demonstrating the principle in class on Monday for everyone.

First day of circuits class went ok

I went into the class with the following to-do list:

1. Introduce teaching staff ✔
2. Go over syllabus, with the main points here:
   • Goal is not to make them EEs, but to bring them to electronics hobbyist levels, so they can design stuff that isn’t too tricky and talk with EEs about tricky designs. ✔
   • Accepted as circuits course for BME, but not as prereq for other EE courses.✔
   • Lab centric—theory taught as needed for design labs. ✔
   • Prelabs to be shown at beginning of Thursday labs. ✘ Oops, I forgot to mention this.
   • Lab demos to be done in Thursday labs. ✘ I forgot to mention lab demos also.
   • Post-lab writeup is major portion of grade (in both courses) due Mondays after labs. ✔ We talked quite a bit about the importance of engineering reports and who the audience for them is (not the professors!).
   • Parts kit to be sold Wednesday. ✔
   • Free on-line books in place of hard-copy textbooks. ✔
   • Rotating partners (prelab separately, postlab together or separate). ✔
   • Academic integrity boilerplate. ✔
   • Disability accomodations boilerplate. ✘ There wasn’t time for this, but I did mention in response to one question about allergies and sensitivities that we could try to work something out. The syllabus does have the needed info about requesting accommodation for disabilities.
3. Demo of pressure sensor ✔ and EKG ✘ with Arduino data logger code.

The EKG board, which I checked over the weekend, refused to work for me in my tests before class, so I waved it around but didn’t demo it. I did demo the pressure sensor board, which worked fine. I used the v1.0.0b4 version of the Data Logger, (available from http://bitbucket.org/abe_k/arduino-data-logger/downloads under “tags”, which is the latest named version. The version currently installed in the labs is v1.0.0b2, I believe, and I have to decide whether to update before Thursday’s lab or not. The b2 version should work well enough, I think, but if students switch Arduinos between different versions without uploading again, they might have problems.

I’ve put an improved version of the parts list up on the class web site and told the lab-support staff about it, so that they can start figuring out how much the student lab fee will have to be next year.  This year, students are paying $30 for the Arduino and $65.50 for parts and tools, but I doubt that the lab fee will be less than $150 next year.

The board designs for the three PC boards used in the course have also been released (with an appropriate Creative Commons license—non-commercial, since I’ve not paid the Eagle license fee needed for commercial designs).

Finally, I’ve tentatively lined up a biologist to talk about excitable cells and action potentials towards the end of the quarter, before the EKG lab.

In other news, my son told me yesterday that his Christmas present to me (which he had told me he would get me later), was the Data Logger code. He couldn’t have given me a nicer present: it was something I wanted, that he put a lot of thought and effort into, that he made himself, that I couldn’t have bought in a store, and that would have been a lot of trouble to do for myself. OK, so we’re both geeks, seeing a software package for teaching a course as a good present, but it really was.

Hysteresis board

Now that we’re using a 74HC14 Schmitt trigger in the capacitive touch sensor for the hysteresis oscillator, that lab can be the first soldering project, in addition to learning about hysteresis.

I tried laying out a very compact PC board for the students to solder (still requiring them to do some design—they’ll have to breadboard their design first to get appropriate R and C values). I came up with one very compact design that could get 4 copies into the 50mm×50mm limit of the $1 boards from ITEAD, making the boards only 25¢ each.

Compact layout to get 4 hysteresis oscillator boards out of one 50mm×50mm board. The gutters are pretty narrow, though, and I’m not sure I’m skillful enough with the board shears to cut that accurately.  The yellow “airwires” are Eagle telling me that the Gnd and +5V wires are not connected between the different copies.

It seemed a little silly to try to squeeze the price down to 25¢, when the other parts cost 90¢: 59¢ for the screw terminals, 28¢ for the Schmitt trigger chip, 1¢ for the resistor, and 2¢ for the capacitor. With this layout it is also a little tricky for the students to properly wire the unused inputs high.

Given the high risk of ruining the boards trying to cute them with the board shears, I decided to redesign for a 50¢ board.

Much looser layout, having only two copies on the 50mm x 50mm board. This version makes it easier for the students to see how things are connected, and has lots of room for the board shears to make the cut.

The lab would now require that the students measure the thresholds of the Schmitt trigger, breadboard the hysteresis oscillator, make a touch pad out of foil and packing tape, measure the frequency of the oscillation to estimate the touch pad capacitance, adjust the parameters of the Arduino program to match the frequencies of their oscillator, solder up the board, and demonstrate it working to control an LED. I think that is plenty for a 3-hour lab.

When I set up the web pages for the course, I’ll try to make sure I put the Eagle design files (.brd and .sch) for each board the students use on the web, so that future instructors can easily order more copies of the board, even if my laptop gets run over by a beer truck.  That will also make it easier for instructors at other schools to try to duplicate the course.

New PC board design for pressure sensor

I designed a new breakout board for a pressure sensor today, for the MPX2053DP sensor. It took me a while to get the unibody outline drawn in an Eagle library so that I could place the part.

I could get 10 copies of the board from ITEAD for the same price as the instrumentation amp prototyping boards ($9.90 for 10, plus shipping), since they are 1.25″×1.1″, or 3.175cm×2.794cm, which is under the 5cm×5cm breakpoint for that price.  I could get 3 copies from OSHPark for $6.88 (including shipping).  If I later need a larger number of boards, I can get 100  boards smaller than 5cm×5cm for $75 or 200 for $120 from ITEAD, but I don’t think I’ll need that many this year, since we’ll either make a dozen pressure sensors for the lab, or have each student solder up their own (for which 30 boards would be plenty this year).

I also found a source for a cheap capacitor collection that we may be able to use for the student kits.  It has 10 each of 25 different values from 1pF to 0.1µF.  I would have preferred slightly larger values (say, 47pF to 4.7µF), but at $4.80 for 250 capacitors it seems like a pretty good deal.  Unfortunately, I can’t tell if they are 0.1″ or 0.2″ spacing on the leads, which would make a difference for my protoboard design.  If I order the pressure sensor breakout boards from ITEAD, I could order capacitors at the same time.

 

 

Thinking about PC boards and parts kits for circuits lab

On my way to work today, I was thinking a bit about redesigning the PC boards for the circuit lab.  I have 2 boards designed so far: an instrumentation amp protoboard and pressure-sensor breakout board.

I have to redesign the pressure-sensor board for two reasons: 1) it is mis-wired, so the screw terminals are labeled wrong (see Pressure sensor miswired), and 2) I’m going to change to a pressure sensor with a built-in barbed port, so that hoses can be directly connected (see Rethinking the pressure sensor lab). The new choice of pressure sensor has a different mounting

We’ll need to decide whether to make the pressure sensors lab equipment (which we’d have to check out and keep track of) or put them in the student parts kit for the course.  If I put them in the part kit, then the students can solder the breakout boards themselves, but it adds the cost of the sensor, the board, nuts and screws, a capacitor, and the screw terminals (about $13) to the parts kit.

The instrumentation amp protoboard is functional as it is (I wired up the instrumentation amp for the pressure sensor on it), but I’m not real happy with the design.  The core is ok, but I think now that the space for a barrel jack for a wall-wart is wasted space. Wall wart power is too low-quality for analog work, and I don’t want to put a regulator on the board. The students have a fine bench supply (Agilent E3631A), so we might as well use it.

I need to have a 4-pin connector for connecting to the pressure sensor (3 of which would be used for the EKG), 2 pins for connecting to the power supply, and 3 or 4 for connecting to the Arduino (Gnd, analog out, and Aref at least). I think that I’ll want to have more Vdd and GND points on the board, as the routing for those was trickier than I would have liked. It might be good to have a well-separated pin for connecting an oscilloscope ground. I’d like to add spaces for connecting up 1 or 2 transistors and another dual-op amp chip.  These changes would almost certainly increase the size of the board, raising the per board price from $1.40 to $2.60 (unless I buy many).  I need to think about what circuits we’ll have the students solder (versus building on a breadboard).  Currently, I’m leaning toward having them solder the EKG, the pressure-sensor amp, and the capacitive touch sensor.  The simple op-amp audio amp should be breadboarded, but I’m not sure about the variant with an output transistor for more power.  This means that each lab group would need 3 or 4 boards for the quarter, which adds another $5–10 for boards.  Screw terminals for connecting power and such to those boards adds another $6. Breadboarding is certainly cheaper, especially since the op amps and other parts could be reused from one project to the next.

I’m also leaning toward getting each student a large collection of resistors (like this collection of 10 each of 112 values for $12.90) and a smaller collection of capacitors.  I’ve not been able to find a cheap assortment of capacitors in different values (other than surplus assortments of random values, which is not of much use to us), so we may have to pick a small number of useful values, and buy the parts separately.

It looks like the parts and boards for the student kit including everything will come to far more than the $43 lab fee currently being charged to students in the usual circuits lab (which uses far fewer parts, and those mostly very cheap ones). If we require Arduinos as well, we’ll certainly far exceed that price. I don’t know if we’ll even be allowed to charge a lab fee.  The page about the current fees says “The fees shown below have been established using prior course history and have been approved by the Dean of Engineering and reviewed by the UCSC Student Fees Committee. The UCSC Miscellaneous Fees Advisory Committee recommended adoption of these fees and final approval was made by the Chancellor.”  I doubt we’ll have time to figure out all the parts we need in order to set a fee in time for a committee to meet and approve fees.  We may have to do direct sales of kits and parts to students, if that is permitted. It is going to be difficult even to get a parts list together in time to buy the parts, much less to figure out the prices, set the fee appropriately, and get approval for the fee.