Digital Tension Scale, Part V

posted 09 January 2019,

In my previous post on this subject I described the components I planned to use for my Digital Tension Scale project, and also the design for a box that would mount directly on the S-shaped load cell assembly.

This post describes the ‘final’ (to the extent than anything I do can be considered final) assembly of the completed system into my 3D-printed housing, and the results of some initial battery-powered testing.

As shown in the following photos, the major components (Teensy 3.2 microcontroller, HC-05 Bluetooth Module, HX-711 load cell amp/A-D, and Sparkfun ‘Basic’ LiPo Charger) were mounted on perfboard, which was then in turn attached to the box lid via a set of custom-printed standoffs.  A short piece of ribbon cable connects the Teensy to the LCD display.  The general idea behind this physical layout is to allow easy access to the electronics for troubleshooting, and to allow for battery charging and/or Teensy programming without having to open the box.

3D-printed housing. Note the glow from the Sparkfun charger LED

View of housing showing the access port for supplying USB power and/or programming the Teensy

View with the lid and electronics board removed. The LCD display is face down in its cutout

Exploded view showing all system components

Showing connections from load cell to HX-711

Top view showing how load cell attaches to the housing

Closeup showing load cell lead routing and power/programming port

End view showing charging port


Preliminary Testing Results:

At this point I have everything running on battery power only inside the box, and I have been able to demonstrate remote data capture on my PC using the HC-05 BT link.  The following image shows the data taken from my rowing machine, and a short video demonstrating the setup.

Complete Code:

Here is the complete Teensy 3.2 program as it stands today.  As you can see if you inspect the code, I have the Teensy low-power stuff turned OFF for the moment (that’s the purpose of the ‘#define NO_SNOOZE’ statement.



Future Work:

  • Do some more work to reduce power consumption to extend the battery life.  I got the ‘Snooze’ feature to work on the Teensy, but that only reduces the Teensy’s power consumption; it does nothing directly to reduce the power consumption of the other components.  I tried using a MOSFET to turn the HC-05 BT module on & off, and found this to be impractical, as then the module loses its connection to the remote data collection device.   I have also tried removing power from the LCD module, but that also turned out to be problematic.

Stay tuned,




Digital Tension Scale, Part IV

Posted 25 December 2018

In my copious free time I have been refining the design for a low power battery operated tensionometer.  In my last post on the subject, I had described the components I had planned to use, and in the ensuing weeks I have been working on implementing this design.  There are several challenges in this project:

Bluetooth Link:

There are a huge number of Bluetooth products out there in the Maker-verse, with varying degrees of Arduino support, and widely varying performance characteristics.  To add to the confusion, there is ‘regular’ Bluetooth and the more recent ‘BLE’ (Bluetooth Low Energy) which are completely incompatible with each other.  As I now understand it, BLE is synonymous with Bluetooth 4.0+ (the iPhone 4S was the first smartphone to implement the new standard).  However, the most common product in use in the Arduino world seems to be the venerable HC-05 ‘regular’ Bluetooth module, available from your local grocery store (well, not quite, but from almost everywhere else!) for not much more than a few pennies

I had no previous experience with BT modules, so this part of the project took some time, and was the last major part to be accomplished.  After receiving my HC-05 modules from China, I used this tutorial to get started.  The real challenge for this part of the project wasn’t getting the HC-05 hooked up to the microcontroller – it was sorting through all the layers of BT-related settings on my Win 10 laptop to pair with the HC-05 device and determine which serial port did what.

  • In Windows 10, I used the ‘Bluetooth and other devices settings’ panel (Settings -> Devices -> Bluetooth and other devices) to find and pair to the HC-05.  The device shows up as ‘HC-05’ and the default pairing password is “1234”.
  • When the HC-05 is paired it automatically sets up at least two serial ports that show up in device manager as ‘Standard Serial over Bluetooth’ ports, as shown below.  However, only one of these ports is actually usable for two-way communication, and it isn’t clear to me why, or how to tell which is which;  I had to experiment with each available ‘SSoB’ port to figure out which to use (so far, it seems like the highest-numbered port is the proper one).
  • After the HC-05 is paired and the com ports are set up, then any serial terminal app (I used RealTerm) can be used to communicate between the PC and the microcontroller via the HC-05.
  • On the microcontroller (I used a Teensy 3.2 with multiple hardware serial ports), I wired the HC-05 to Serial1 leaving Serial (Serial0) available for normal communication between the Teensy and my Visual Studio 2017 Community Edition/Visual Micro add-on for Arduino development platform.

Physical Layout:

The original idea behind this project was to create a self-contained battery-operated digital weight scale that could display weight values on a local display, but could also stream the data live to a remote recording station like a laptop or smartphone. The ‘self-contained’ part requires that all the electronics be mounted on the S-shaped load cell assembly itself, and to that end I designed a housing that connects to the two bolts that hold the arms of the load cell.  The idea is that all the electronics save the LCD display and the battery will be mounted to the underside of the box lid so that servicing would be easier.  Also, by mounting everything to the lid, I can make cutouts for the charger and Teensy USB connectors for easy charging and reprogramming.  After several iterations in TinkerCad, I came up with the following design

Looking up at the underside of the box lid, showing all modules except the battery and the LCD display

Showing the top of the lid with the mounting bracket for the load cell

Isometric view with transparent box walls. The LCD display module is under the battery.  Note the cutouts for the charging and programming USB-C connections

Module Integration:

I had previously tested each module individually, but hadn’t had all of them working at the same time.  I had tested the HC-05 with an Arduino Mega, and I had tested the load cell with both a Sparkfun Pro Micro and with a Teensy 3.2, and I had tested the Nokia LCD display with a Teensy 3.2, but I hadn’t put everything together.  So I wired everything up on my half-size ASP plugboard and got it all working together with a simple program (included below) that exercised the LCD Display, the load cell, the BT module, and the battery charger, as shown in the following photos

RealTerm Serial Terminal Program showing load cell readouts collected wirelessly via the Bluetooth HC-05 modules

HC-05 Bluetooth, HX-711 Load Cell Amp, Sparkfun Charger, and Teensy Microcontroller modules integrated together. Note disconnected USB cable showing that the circuit is running on battery power. The scale is currently measuring 1.8 liters of water in the suspended plastic bag (note the ‘1.8 Kg’ reading on the LCD display)



The software used for the above integration tests is a reasonably complete sketch for day-to-day use of the digital weight scale.  It displays the measured weight on the LCD display, and also sends it to the USB serial port for display on a directly connected PC, and to the HC-05 Bluetooth module for display/capture via a BT-connected laptop or smartphone.  This program is shown below:

However, this program depends on the proper calibration of the load cell, which I have been doing with a separate sketch (also included below):

What I need to do now is to combine these two programs into a single sketch with ‘operating’ and ‘calibration’ modes.  My calibration program already does this to some degree, as it waits 5 seconds on startup for the operator to send the ‘y’ key via the direct-connect serial port. If the ‘y’ character is detected within this window, then the program starts the calibration sequence; otherwise it starts taking measurements as normal.  This behavior needs to be expanded somewhat in that it should accept a calibration command either through the direct-connect serial port (Serial) or via the BT port (Serial1).

Low Power Operation:

I have already done some experimentation on low-power operation of the Teensy 3.2, using Colin Duffy’s fine ‘Snooze’ library, and have determined that I can easily drop the Teensy’s operating current from around 20-30 mA to about 1-2 mA by putting it to sleep during periods of load cell inactivity.  Assuming I get the full 2500 mA hours out of the battery, then I can expect something like 1000 hours or about 40 days between recharges.  However, more work needs to be done to get the low power mode fully operational.


Stay Tuned!



Digital Tension Scale, Part III

Posted 25 November, 2018

Over the Thanksgiving weekend I had a chance to do some more work on my digital tension scale project.  My wife and I drove to St. Louis to visit our kids and grand-kids, and its a 6-hour drive each way.  I make sure we have an audio book going for my wife, and as a consequence I get 6 hours of (mostly) uninterrupted geek time to work on things like this project.

As I mentioned in Part II, my goal is to construct a battery-operated tensionometer that can be mounted directly on the dual-hook S-shaped tension block, as defined by the features delineated in Part II.  After doing some more web research, I came up with the following possible components for the system:


Apparently, the Nokia 5110 84×48 pixel monochrome LCD display used in prehistoric times as the display in Nokia cellphones has found a second career as a simple, low power display for battery-operated devices like the one I envision.


Nokia 5110 Monochrome LCD display

Battery & Battery Charger:

The LCD display will operate quite nicely from 3.3V, so as long as I can come up with a 3.3V micro-controller (like the Teensy 3.2) and a small, capable LiPo charger, I should be in business.  For this component, I plan to use the Adafruit PowerBoost 1000C and something like the 3.7V 2500 mAh battery as shown below

Adafruit PowerBoost 1000C single-cell LiPo charger

3.7V 2500 mAh LiPo battery


I don’t think I’ll need the 5V boosted output from the PB1000C and the entire thing may be a bit of overkill for this project, but I had them hanging around from a previous project, so…


For this I plan to use one of Paul Stoffregen’s magical Teensy 3.2’s.  Again this is probably way overkill for the project, but…

Wireless Connection:

This component is the one for which I have the least understanding and confidence.  I currently use a Pololu Wixel for wireless serial comms and programming with my autonomous robot, but I haven’t figured out how to use it with a Teensy, and I thought maybe there were better solutions out there by now anyway.  So, after some more web searching I found that many ‘makers’ are using the HC-05/6 Bluetooth modules for this purpose.  Hopefully with this module I’ll be able to use a Bluetooth connection from the tensionometer to my laptop or even my cellphone to perform calibrations and collect real-time tension data.

Combining all these with the HX-711 load cell amplifier, I came up with the following system schematic.


Stay tuned!



Digital Tension Scale, Part II

Posted 17 November 2018

In a previous post on this subject, I described a digital tension scale arrangement using a load cell incorporated into a 2-hook tension measurement setup, interfaced to a common HX711 ADC board, and a Sparkfun Pro Micro ATMega32U4 microcontroller.

After (finally) getting the setup to work and getting some initial real-time tension measurements on our rowing machine, I decided to see if I could improve the usability of the overall system, with the goal of constructing a fully automatic battery powered tension scale, with the ability to communicate wirelessly to my PC for data acquisition and programming.

Desired Features:

  • Easy calibration:  The system should offer a calibration option when connected to a PC/Smartphone, but otherwise should use the last calibration data (stored in EEPROM) for measurements.
  • Battery operated: This implies a low-power mode to extend battery life if using primary batteries, and/or a charging arrangement if using secondary (rechargeable) cells.
  • Local display:  A low power display (LCD?) for local tension measurement display
  • Wireless capability:  A wireless connection to a PC or smartphone for real-time data acquisition.
  • Small size:  I would like to mount the entire system on the 2-hook tension measurement assembly itself.

Easy Calibration:

The calibration procedure associated with my previous post was a PITA, to say the least, so I decided to attack this problem first.  I modified the software to allow the user to skip calibration entirely or to calibrate the tensionometer automatically using any known weight.  The new software is shown below:

When run on my PC, this produced the following output

Here are a couple of photos showing the calibration process with my current setup

‘Tare Weight’ configuration, used to zero out the contribution from the bucket

Calibration configuration. Bucket now contains 1.8L water (1.8 Kg)

After calibrating, I tested the the system by measuring tension vs time with an elastic strap I am using as part of my rotator cuff surgery rehab, as shown in the following short video clip

The data from this experiment was captured on my PC and plotted in Excel, as shown below

Tension vs time for the orange elastic strap shown in the video

Then, at the request of my physical therapist, I measured the real-time tension for single & double orange straps, and single/double green straps, as shown below

Single and double green strap tension vs time

Single and double orange strap tension vs time


Stay tuned!




Digital Tension Scale

Posted 27 October 2018

I recently underwent rotator cuff repair surgery on my left (dominant) shoulder, and am now starting the rehab process.  My PT person was adamant that I not re-start my normal rowing routine for at least six weeks post-op, due to the possibility that I could re-tear the tendon.  This made me curious as to what the tension really was on my arms when rowing, so I decided to try and build a digital dynamic tension sensor, capable of plotting rowing strap tension in real time.

To start, I had to educate myself on the world of strain gauges and load cells, and what the differences are.  As I came to understand, what I wanted was a load cell configured for tension measurement, with a strain gauge as the active sensing element in the load cell.  So, I started searching for load cells, and was immediately inundated with ‘too much information’.  This deluge is certainly better than the old days where I had to search through paper (really, no internet!) magazines and catalogs, but at least you didn’t have to worry about overload headaches! ;-).

Anyway, I found this item ‘Degraw 40Kg Tension Load Cell and HX711 Combo Pack Kit‘, as shown in the screenshot below

Amazon catalog item for Degraw load cell

This looked perfect for my intended use, as I could hook one end onto my rowing machine strap, and connect some sort of handle to the other end.  Now all I had to do was figure out how to hook the thing up and get it to work.  Fortunately Degraw also provided a sketch of the hookup using an Arduino Uno, so that part was pretty easy.

Degraw-provided hookup diagram

After a bit more research, I found a nice HX711 library by bogde and some example programs, and got the whole thing to work using an Arduino Mega 2560.  Once I got a program running with some preliminary (but believable) results, I started thinking about how I was going to manage the physical aspects of hooking this assemblage to the rowing machine and recording dynamic tension.  I couldn’t really just let the HX711 board hang by the strain gauge wires while connected with jumpers to the Mega board, as the #28 strain gauge leads would surely break.  So, I came up with the idea of somehow attaching the HX711 board and a small microcontroller to the load cell assembly, and then connecting the whole thing to my laptop with a USB cable.  Hopefully the USB cable would be long enough to allow full extension of the rowing machine strap so I could collect full rowing cycle data.

After some digging around in my parts cabinets, I came up with two candidates for the ‘small microcontroller’ part of the plan; a 3.3V Teensy 3.2, and a 5V/16MHz Sparkfun Pro Micro.  I tried the Pro Micro at first, and almost immediately went down the rabbit hole (my term for getting lost in some technical wonderland without a clue how to get back) trying to figure out how to program the device – a challenge due to the way it handles com ports through the USB connector (it actually implements two different ones, depending on whether the boot loader or the user firmware is running – yowie!).  After climbing my way out of the rabbit hole, I decided to try the Teensy 3.2 instead, as I familiar with it from several other projects.  With the Teensy, I got a test program running and started taking data with known weights attached to the load cell. The way I did this was to suspend a plastic bucket from the load cell, and poured water into the bucket one liter (1Kg) at a time while recording data.  This was successful because I got good data, but unsuccessful because the data didn’t make much sense, as shown in the plot below

Results of pouring 1L (1Kg) water at a time into bucket suspended from load cell, using a 3.3VTeensy 3.2

As can be seen, the data was anything but the stairstep function I was expecting to see.  At this point I wasn’t sure if I had a hardware problem or a software problem, or something else entirely, so I sent an email to Degraw Product support with the above plot attached, asking if they had any insight into the problem.  Amazingly, they replied almost immediately, and offered to send me another load cell unit gratis so I could eliminate their hardware as the cause of the problem.  Although I was quite pleased with their offer of support, I thought maybe the 3.3V supply of the Teensy 3.2 might be causing the non-linearity (the HX711 advertises 2.7-5V operation but the lower voltage might be causing output linearity problems).  So, I tried again with the Sparkfun Pro Micro, and this time I managed to make the programming magic work. Then when I did the same test as above with the 5V Pro Micro instead of the 3.3V Teensy, I got the plot shown below.

Tension vs time plot created by pouring 1L (1Kg) of water at a time into bucket suspended from load cell, using Sparkfun 5V Pro Micro

So, now that I had the software and microcontroller problems solved, I started working on the mounting issue.  After a few minutes in TinkerCad and some quality time with my PowerSpec 3D PRO 3D printer, I had a mounting platform that clipped onto the two vertical rods in the ‘S-shaped’ tension load cell, as shown in the images below.

Reverse side of assembly, showing mounting plate clips attached to load cell vertical members

Sparkfun Pro Micro and HX711 board mounted on load cell

After getting all this set up, it was time to take some real data. Since I was still in the ‘no rowing’ zone after my surgery, I enlisted my lovely wife to do the honors while I recorded the data.  We have an Avari magnetic rowing machine, which thankfully doesn’t make much noise.  I recorded a total of 12 rowing cycles on two different ‘wave’ settings (I’m still not sure what the different ‘wave’ settings mean) at the lowest tension level, with the results shown below

As shown in the above plot, the peak tension reading was around 18Kg (about 40 lbs).   I’ve included a short video of the test below.

IR Modulation Processing Algorithm Development – Part IV

Posted 07 June 2017

I seem to have grabbed a tiger by the tail in my continued collaboration with my old friend and mentor John Jenkins on this project to extract the estimated magnitude of a square-wave modulated IR signal in the presence of ‘flooding’ from ambient light sources.  The whole thing started out innocently enough when John was at our house a few weeks ago and I showed him my spiffy wall-following autonomous robot Wall-E2.  I mentioned at the time that I was having some trouble getting Wall-E2 to successfully home in on an IR beam to mate with a charging station, and he suggested that a square-wave modulated signal might do the trick, as it would allow me (and Wall-E2) to discriminate between the IR beam and the other interfering signals.

I should have known right then that I was in trouble, as John had that look in his eyes – the one that says “Hmm, that’s an interesting problem…..”.  I have seen that look any number of times over the 40 years or so that I have known him, and it always results in me tearing up my previous work (and my previous assumptions) and starting over again.  The only saving graces in all this are a) It’s a lot of fun when it happens, b) I’m retired now so I don’t care how long it takes or how much work is involved, and c) I’m a masochist at heart! ;-).

So, here we are.  John and I have been exchanging emails over the last week or so, discussing the ‘best’ way to solve this problem.  One of the first things that happened was John started blabbering about ‘Degenerate N-path Bandpass Filters’, and I had no idea what he was talking about (I hate it when that happens!).  Of course I couldn’t tell John that I was completely ignorant, so I made the appropriate noises and raced to educate myself before he discovered my ignorance.  I found a neat video that explained the technique in a way that even I could understand. The video focused on implementing the filter technique in CMOS hardware, but the general technique is applicable to the Arduino world as well, as long as the modulation frequency is low enough to accommodate the lower clock speeds.

John came up with a brilliant graphical illustration of the algorithm used for implementing a N-path band-pass filter.  Even more amazing, he did it in Excel, so the whole damned thing is live – wow!   As shown in the ‘dead’ version below, there are two ‘channels’ – an in-phase (I) and quadrature (Q) ‘channel’.  Both channels start with a group of samples spanning exactly 1/4 cycle that are summed together to form a ‘sample group’.  In the I channel, each group is given a sign in the sequence ‘+’, ‘+’, ‘-‘, ‘-‘, and in the Q channel this same group is given a sign in the sequence ‘-‘, ‘+’, ‘+’, ‘-‘.  The four groups in each channel are summed to form the I & Q ‘cycle sums’, and each such sum is added to a N-element circular buffer (one each for the I & Q channels).  The running sum of all elements in each circular buffer are the band-pass filtered I & Q components of the input signal.  The final result is formed by adding the absolute values of the I & Q running sums.

Page 1 of the above document shows the sign sequence for the in-phase and quadrature ‘channels’.  Note that the input to both channels is the same ‘sum-of-5-samples’ group, but the sign changes in a different sequence for the I & Q ‘channels’.  Each 1/4 cycle of the input signal is treated in the same fashion.  The horizontal time scale is specific for the planned 500Hz modulation signal.

Page 2 depicts the real heart of the algorithm, as it shows how each ‘sample sum’ (1/4-cycle sum-of-samples) feeds into the final circular summing buffer.  Four ‘sample sum’ groups (comprising one signal cycle) are summed into a ‘cycle sum’ for both the I & Q channels, and each such ‘cycle sum’ is loaded into the corresponding I or Q circular summing buffer.  The latest ‘cycle sum’ element replaces the oldest element, and therefore the  circular summing buffer for each channel represents a N-element  band-pass filter, where N is the length of the circular summing buffer.  This diagram has a lot of information in it, so it can take some time to get comfortable with it (it did for me, anyway).

Some other random details:

  • This diagram is set up for a 64-element circular summing buffer, but shorter or longer is OK too.  The band-pass filter bandwidth is inversely proportional to the buffer length.
  • The times noted in the labels at the top are for a presumed 500Hz modulation, with a period of 2000uSec.
  • The notation m = n/5 comes from the fact that (in this particular implementation) there are 20 samples/cycle, which means that each 1/4 cycle group of contains 5 samples.  For 64 elements, each comprising 1 entire cycle of samples, there are 64X20 = 1280 samples. So in this implementation, m/4 = 64 ==> m = 256 = n/5 ==> n = 1280 total samples represented in the circular buffer at any one time.

Page 3 shows the contents of one of the two circular buffers, after each complete ‘rotation’ of the buffer.  After 64 complete cycles the buffer is full, as shown in the ’00’ column.  64 cycles later, the buffer contents are as shown in the ’01’ column, and so on.  The text below shows the procedure for start and running the buffer.  Note in this text that “next cycle value” and “current cycle value” are the same thing, and the “input pointer” and “output pointer” variables are incremented MOD N, (N = 64 here).

Pages 4-12 show the input signal, the I/Q channel values, and the I & Q circular buffer running sums for different phase offsets between the transmitted and received signals (transmit & receive frequencies are the same in all cases).  Page 4, for instance, shows the situation for the receiver perfectly in phase with the transmitted signal.  If we look at the I & Q summed values at integral cycle boundaries (-1, 0, +1, etc) we see that the I signal is at 128, and the Q signal is at 0, giving abs(I) + abs(Q) = 128 + 0 = 128.  If the same calculation is performed for all the other phase relationships (i.e. pages 5-12) at the same points, the answer will always be the same, i.e. 128.  This shows that the band-pass filter implementation works as intended, even without any phase-locking requirement.  This treatment assumes that the Tx & Rx clocks are identical, so that the 1/4-cycle sample groups span exactly 1/4 cycle.  Any difference between the two clocks will show up as ripple on the results, proportional to the difference between the two frequencies.  This error term should not be significant for the typical single-board-computer crystal-controlled clocks.

I sent this post off to John for approval, and got the following email back, with some additional clarifying comments:

  • n-path filter is degenerate (my term) because it only uses two paths — I and Q — vice many as shown in video (very good video btw,learn a lot from it too)
  • possible topics for more detail related to last paragraph (but beware that perfect is enemy of good enough or something like that):
    • output ripple
      • frequency ==> difference between tx and rx freqs should be very small
        • only spec found was +/-50ppm, so 100ppm delta worst-case w/o aging, etc
        • 100ppm at 500Hz ==> 0.05Hz
        • believe ripple will be twice difference in freq because of this being a fullwave synchronous rectifier (not verified), so 0.1Hz
      • magnitude ==> variations due to sample(s) from one 1/4-cycle group getting into an adjacent 1/4-cycle group as tx and rx phases slide past each other at <<0.1cycle/sec
        • non-50/50 duty cycle is most likely cause
        • with 5-samples per 1/4-cycle group normalized ripple could vary from 1.0 to 0.8;
          • not a problem for the relative outputs of two sensors for steering
          • could be important in cases where absolute magnitude is needed because bpf (eg, 10Hz) would not average signal for long enough to average out ripple this slow ((eg, 1 cycle every 10sec)
          • increasing # of samples per 1/4-cycle group will reduce this effect
    • no antialias filter is used so # of samples per 1/4-cycle group needs to be significantly higher than nyquist requirement for 2/bw to avoid issues (unless clock rate is servo’d to correct freq by monitoring and setting Q value to zero)

More to come – stay tuned!






How to update the ACP Pyramid

The first thing to understand about updating the Pyramid is that you don’t have to unless there is something wrong with your current Pyramid version (i.e. there is a
a bug or problem) that has been addressed in the latest version available on the ACP website).  As each new version is placed on the ACP website, a ‘change log’ entry is also uploaded, so you can tell what changed from your current version to the one that is on the website. If none of the listed changes affect your practice, then there is no real need to update.

the second thing to understand about updates is that, even if you download the latest version from the ACP site, there is no need to immediately update all your client Pyramid files – it is completely reasonable to update client Pyramids to the latest version only when you next deal with a particular client file, and this is done by simply importing the client’s Pyramid data into the new (blank) version using the ‘Import from File…’ button on the Inventory page (don’t forget to use ‘Save As…’ to save the updated client file back to the client folder while leaving the new version file unchanged).  Here’s a video tutorial showing the update procedure

OSU/STEM Outreach Pulse Detector Project, Part II

Posted 22 October, 2016

In my last post on this subject, I discussed a modification to the OSU STEM Outreach’s Pulse Detector schematic to eliminate the dual 9V battery supply in favor of a single one, taking advantage of the LM358 op-amp’s single supply operation capability.  However, before actually recommending that the new arrangement be adopted, I wanted to make sure that it would work properly with the OXO ‘Soft-Clip’ finger clip module instead of the one I created for testing in my lab.  It should work, but as a long-time engineer I have been bitten more than once by the difference between should and would! ;-).

So, Prof Anderson was gracious enough to bring a spare OXO clip to our next Outreach session, and since then I have had the opportunity to test this clip with my new circuit, as shown in the following photo.

OXO 'Soft-Clip' finger clip with new circuit in background

OXO ‘Soft-Clip’ finger clip with new circuit in background

After fiddling with the tension screw a bit, I got the Soft-Clip tension set properly for my finger, and lo-and-behold, the pulse detector circuit worked out quite nicely, as shown in the following short video clip (as you watch the clip, note that the IR beam from the IR LED is visible as a blue-white glow).

At this time I also made one other minor change.  In the original OSU circuit, the DC blocking capacitor (C2 in the schematic) was a 0.1uF, but in my circuit this value was changed to 0.01uF, because that’s all I had on hand.  In the meantime, however, I got some 0.1uF’s from Mouser, and so my final circuit as shown above incorporates a 0.1uF vice the original 0.01.

The next step in this project is to transfer the detector circuit from plugboard to a more permanent version on perfboard.  This will allow me to demonstrate the new circuit to Prof Anderson and the rest of the OSU STEM Outreach team, to lend credence to the idea of modifying this project’s documentation to eliminate the now-unneeded second supply.  Stay tuned!

Posted 10/24/2016

So, tonight I had the time to finish transferring the pulse detector circuit from my old trusty plugboard to a more permanent medium – i.e. perfboard.  The idea here is to provide the OSU STEM Outreach team with a working pulse detector circuit running from a single 9V battery, as a working example of my recommended modifications to the circuit being used presently.  The image below shows the perfboard arrangement, and there is also a short video clip of the new pulse detector circuit in action.

The astute observer might notice there are only three wires going from the OXO ‘Soft-Clip’ assembly to the pulse detector circuit.  While I was wiring up the perfboard version, I realized I could eliminate one wire by simply moving the 100-Ohm IR LED current-limiting resistor from the anode (positive) side of the diode to the cathode (negative) end. This allowed me to connect the IR LED anode to the PhotoDiode cathode, and both to the +9V lead.  Implementing this change for all of the current stock of OXO clips may be more than the OSU STEM Outreach crew wants to take on, but I thought I would mention it ;-).  The finished schematic (with the transposed 100-Ohm resistor) is shown below

Final pulse detector schematic.  Note change to IR LED current limit resistor, and DC blocking capacitor

Final pulse detector schematic. Note change to IR LED current limit resistor, and DC blocking capacitor


Perfboard version of the single-supply pulse detector circuit

Perfboard version of the single-supply pulse detector circuit

The next (and hopefully final) step in all this is to 3D print a small box for the perfboard circuit.  Because I can’t help myself, I have decided to try printing a transparent box – not a trivial undertaking for 3D printing.  Stay tuned!

Posted 10/25/2016

As mentioned above, 3D printing a transparent box is a non-trivial undertaking, at least with the current hobbyist FDM (Fused Deposition Modeling) materials and techniques.  After a number of iterations, the best I could do without a lot of post-processing was a semi-transparent (but still pretty neat) container for the perfboard version of the ‘new-and-improved’ OSU Pulse Detector circuit.

As I have noted previously, the current 3D printing technology makes it easier, faster, and cheaper to go through a number of test cases (literally in this ‘case’!) on the way to a final product, rather than trying to design a final product all in one crack.  Each test takes about 15-30 minutes and just a few cents’ worth of material, and quite often the iterative process illuminates a design problem or opportunity that wasn’t obvious (or even considered) at the start.  In this case, it became apparent after about the 4th iteration that the perfboard should be mounted to the ‘lid’ via printed-on standoffs rather than to the ‘bottom’.

The TinkerCad model for the box and the lid is shown below, as well as the finished product and some of the precursor test boxes.  And, as usual, a short video of the final product.

Pulse Detector Box and lid. Note printed-on standoffs

Pulse Detector Box and lid. Note printed-on standoffs

OSU Pulse Detector Box and precursors

OSU Pulse Detector Box and precursors




OSU/STEM Outreach Pulse Detector Project


Posted 10/16/2016

A week or so ago I participated in an OSU STEM Outreach program that showed high-school students how to build a working pulse detector circuit, using a commonly available op-amp and an IR LED/Photodetector pair.  After an initial presentation, the students were given step-by-step instructions for building the circuit on a small plugboard, and I helped when students ran into trouble.  By the end of the one-hour session, most students were successful, and were able to see the output LED illuminate in time with their pulse – cool!

As I helped out with the class, I was a bit shocked to see that the circuit being build by the students required two 9V batteries wired in series to create a +/-9V supply for the op-amp. At the time I just assumed somebody forgot to spec the op-amp to be one with a common-mode range including ground, and the extra battery was the ‘field-expedient fix’ for the problem. When I asked Prof. Betty Lise Anderson about this, she said that a single supply had been tried at one point, but ‘didn’t work out’ for unspecified reasons.  This piqued my interest, so decided I would investigate the problem a bit further in my home lab.

As it turned out, the op-amp being used in the students’ circuits was the ubiquitous dual-LM358 in an 8-pin DIP, and this unit does indeed have a common-mode range including ground, so the single-supply idea should have worked.  Here’s the original OSU circuit (transcribed into DipTrace’s schematic capture format)

Original OSU Pulse Detector circuit.  Note the dual 9V suppies

Original OSU Pulse Detector circuit. Note the dual 9V suppies

And here’s my final single-supply detector circuit

Final single-supply Pulse Detector schematic

Final single-supply Pulse Detector schematic

Comparing the two, the only real difference is the addition of the 470K resistor from the inverting input to ground.  In the original circuit, the inverting input was tied directly to ground, while the non-inverting input had a 470K to ground.  This can be a problem, as the input bias current for the LM358 can be on the order of 100 X 10-9 (100 pico-Amp), which means that the DC voltage at the non-inverting input due to bias currents could be as much as 50mV or so.  Since this is on the same order of magnitude as the photo-diode signal at this point, there is a real chance the op-amp would never toggle the output.  The dual-supply setup eliminates this problem, but at the cost of a second battery.

The final circuit, as laid out on my plug-board, is shown below.

Pulse Detector circuit with my 3D-printed finger socket

Pulse Detector circuit with my 3D-printed finger socket

Final Pulse Detector circuit

Final Pulse Detector circuit

And, because I can, here’s a short movie showing the pulse detector in action ;-).

Speaker Amplifier Project, Part V

Posted 10 September 2016

Since my last post on this subject, I have actually gone through two revs of the PCB. The first set was fine electrically (and great quality AFAICT, but I didn’t get the physical border just right, and I couldn’t file it down enough to fit without breaking a PCB run.  So, I went back to DipTrace, redid the board outline (which also involved some rejiggering of component placement and a bit of manual net editing) and sent off another order to Bay Area Circuits.  After returning from a week at a duplicate bridge tournament in Atlanta with my wife, the new rev boards were waiting for me – cool!  The image below shows the new board as installed on the Adafruit 20W amplifier board.

Rev2 PCB installed on Adafruit amplifier

Rev2 PCB installed on Adafruit amplifier

After ensuring that the PCB fit was OK, I populated it, checked all the net connections and resistor values and gave it the old smoke test – and of course it failed – ugh!  I quickly found that the problem was that common-mode range of the the MC748 dual op-amp I was using for the second system doesn’t include ground – oops!  A quick trip to the local Microcenter and the purchase of a wildly overpriced NTE928 dual op-amp solved that problem, but because I was I had soldered the op-amp package directly onto the board (I was sure this was all going to work, after all), I was now faced with a messy removal and cleanup job.  After getting the old op-amp off and the PCB cleaned up, I installed the new one  (this time, suitably chastised, I installed a socket, and then installed the op-amp).  The following image shows the finished PCB connected to the amp board, next to the hand-wired original.

Completed PCB connected to Adafruit amplifier

Completed PCB connected to Adafruit amplifier

Bottom of PCB and hand-wired original (note messy area around op-amp)

Bottom of PCB and hand-wired original (note messy area around op-amp)

The last thing to do was to get the PCB actually mounted onto the amplifier board, and the assembly into its custom-printed housing, as shown in the following images.

hand-wired and Rev2 PCB mounted to the Adafruit amp boards

hand-wired and Rev2 PCB mounted to the Adafruit amp boards

hand-wired and rev2 PCB mounted in custom-printed enclosures.

hand-wired and rev2 PCB mounted in custom-printed enclosures.

Now all I need is a ‘speaker’ presentation so I can find out if all this work was really worthwhile (actually, I’ve already had a great time with this project, so having it actually work for the students will just be icing on the cake!)