Certain synthesizers (Nord Lead series) used strain gauges to bend note pitch (instead of a traditional pitch bend wheel design). This presumably involed spring steel, a Wheatstone bridge, amplification and filter smoothing before or after the ADC. The end result was more nuanced vibrato for synthesisers, akin to violins or a guitar whammy bar which would spring back to a neutral position. Would make a nice project if combined with MIDI.
This is really interesting. I tried to find a video with a teardown of one of these, but not much luck. I did find someone plucking it, so I can hear that it's connected to something metal, and it does look akin to a whammy bar flutter as you mentioned. Thank you very much for mentioning it. -Derek
You did fine. Maybe a little too cautious, but being new, I understand. Worked with them for almost 20 years. In the plastics industry, they have a mold that closes on a pair of 8" - 14" horizontal rails - round. The mold halves close together, by means of a hydraulic ram, or a scissor toggle, Before the hot 700 degree plastic is injected into the mold, there must be an assurance that the mold halves are clamped together tight enough to prevent the injection pressure from opening the mold. Presets for tonnage must be maintained. If the strain is not met, it will not inject, if the strain releases mid injection, the injection stops. The strain gauge feeds an op-amp, and its tonnage set-point is dialed in for each mold. Punch presses use them to detect breakage. When the ram comes down, to stamp a part out, the dies must achieve a preset tonnage value to make all the cuts, and bend all the bends. In the canning industry, the most important finish dimension is the 'score' for the pull tap. If it is to work, the metal along the 'score', has to be thinned to allow it to tear, but thick enough to keep from cracking. The strain gauge, measures this pressure, and then corrects the heat amount in the die, to increase or decrease this cut depth. The worst part in the early years, 1990 to 2000, was the hour of math needed to prefigure the new strain gauge, to the machine. It was an hour with a calculator. We got tired of this, with 2-3 replacements/week, for 600 machines, so we made a spread sheet with the calibration numbers and fixed limits that came with each gauge. Put in 4 numbers, hit enter, and done
I've seen this application used in Large " C " frame punch press. To give indication on frame load. Large CNC mills. For measuring thermal growth. Using a linear growth on gage.
Makes sense on CNC mills - I've always seen that little user interface strain meter bouncing around. They must use a force-torque strain gauge implementation. C frame punch press monitoring makes sense too.. I'm sure there's a tremendous amount of stress in the column. -Derek
One of the more clever ways of integrating a "strain gauge" into a device is the Duet3D Smart Effector which serves as the platform for a Delta 3D printer and has its strain gauge done in as part of the PCB.
Thanks! You can always rely on the Duet3D guys for coming up with something clever. :) But you had to do your bed leveling with the hotend heater off. Because the strain gauge "matrix" acts as an antenna and can throw the measurements off. It took me far longer than it should have to figure that one out. :) And in the real world of loadcells (with their strain gauges bonded to them), the failures that I have seen fall into two categories: 1) A sharp blow, particularly from the side that de-bonds the strain gauge. 2) Moisture. Often moisture will creep up, or "wick" it's way into the loadcell via the cable - even if the loadcell is rated for IP67 or better. So your concerns regarding the proper bonding of a gauge are well founded.
Fun fact: Murphys law was originally coined in response to a miswiring of gauges in a half or full bridge (i forget which). Apparently the tech wired 2 gauges together so that they cancelled each other out, resulting in a sensor that tested perfectly but read absolutely nothing.
Please consider doing an IOT video for something simple like an esp32 and a humidity sensor or something like that which shows the steps to use an online site for communication from phone to device in a home connected to local wifi. Maybe even a small bit on using MIT AI2 companion to create a small app for it to send data thru to the phone for the sensor and send data back to turn on an led. Thanks
This dude is my favorite contributor. I love this particular presentation from Element 14. I can't wait to see more like this that is so educational and entertaining.
Making the video was partly self-serving, as I wanted to make a torque sensor for a robot. So this is a good application. If you look up force-torque sensors, usually they're bolted between the robot "hand" and the wrist. If the hand bumps into something, the torque spikes and the robot drops the servo power. Internal to the sensor, is a web of thinner metal that has some degree of compliance (not much), but enough to register a resistance change in the strain gauges. From that you can determine if the robot crashed, thereby preventing damage to expensive transmission components. -Derek
Maybe it's the Algorithm, but only a couple of days ago while searching for an old email I stumbled upon a 2016 link I sent myself from a work account regarding "G-putty", which is a mix of Silly Putty and Graphene. It is a good conductor of electricity. However, when the material is subject to even a tiny strain or impact, its electrical resistance increases sharply - before relaxing to its original value as the material “self-heals”. A sensor of G-putty when mounted on the neck and chest of a subject could measure breathing, pulse and blood pressure. The sensor was even able to detect the footsteps of a small spider. I probably triggered the algorithm when I searched for more info, as it sounded like (no pun intended) a candidate for a better acoustic guitar pickup, which is on my to-do list as I dislike the ceramic ones. I found "strain gauge" mentioned in my search hits, but I wonder if the dynamic nature of the output would limit its use for static strain measurement.
Strain gauges are used all the time in syringe pumps to measure the force on the plunger of the syringe, the force being proportional to the pressure in the syringe (though friction can vary) They are calibrated with a force gauge. Some have their own pressure sensor that interfaces to the patient tubing, these also have to be calibrated. But some use only low side motor current monitoring which is hard to calibrate and not very good (in my opinion). I think the ones that have pressure sensor, it is a MEMS micro electronic measuring system, its like a strain gauge but is produced by semiconductor lithography, I think the bending of the sensor changes the capacitance as elements are moved closer of further from each other, the change of capacitance can be measured, I think they might be laser trimmed for accuracy ?
Piezo crystals would work with a constantly changing dynamic load. As soon as the load settles, deflected or not, the output signal also settles. For static loads like scales, you need a resistive element. -Derek
Certain synthesizers (Nord Lead series) used strain gauges to bend note pitch (instead of a traditional pitch bend wheel design). This presumably involed spring steel, a Wheatstone bridge, amplification and filter smoothing before or after the ADC. The end result was more nuanced vibrato for synthesisers, akin to violins or a guitar whammy bar which would spring back to a neutral position. Would make a nice project if combined with MIDI.
This is really interesting. I tried to find a video with a teardown of one of these, but not much luck. I did find someone plucking it, so I can hear that it's connected to something metal, and it does look akin to a whammy bar flutter as you mentioned. Thank you very much for mentioning it. -Derek
7:10 Thank you for the included metric measurement, appreciated.
Of course! (Didn't hurt that it was already printed on the surface ;) -Derek
You did fine. Maybe a little too cautious, but being new, I understand. Worked with them for almost 20 years. In the plastics industry, they have a mold that closes on a pair of 8" - 14" horizontal rails - round. The mold halves close together, by means of a hydraulic ram, or a scissor toggle, Before the hot 700 degree plastic is injected into the mold, there must be an assurance that the mold halves are clamped together tight enough to prevent the injection pressure from opening the mold. Presets for tonnage must be maintained. If the strain is not met, it will not inject, if the strain releases mid injection, the injection stops. The strain gauge feeds an op-amp, and its tonnage set-point is dialed in for each mold.
Punch presses use them to detect breakage. When the ram comes down, to stamp a part out, the dies must achieve a preset tonnage value to make all the cuts, and bend all the bends. In the canning industry, the most important finish dimension is the 'score' for the pull tap. If it is to work, the metal along the 'score', has to be thinned to allow it to tear, but thick enough to keep from cracking. The strain gauge, measures this pressure, and then corrects the heat amount in the die, to increase or decrease this cut depth.
The worst part in the early years, 1990 to 2000, was the hour of math needed to prefigure the new strain gauge, to the machine. It was an hour with a calculator. We got tired of this, with 2-3 replacements/week, for 600 machines, so we made a spread sheet with the calibration numbers and fixed limits that came with each gauge. Put in 4 numbers, hit enter, and done
I've seen this application used in Large " C " frame punch press. To give indication on frame load.
Large CNC mills. For measuring thermal growth. Using a linear growth on gage.
Makes sense on CNC mills - I've always seen that little user interface strain meter bouncing around. They must use a force-torque strain gauge implementation. C frame punch press monitoring makes sense too.. I'm sure there's a tremendous amount of stress in the column. -Derek
I remember using these at the wind tunnels at Langley NASA back in 1981. Another thing to consider is putting thermal compensation on one leg.
Loved that! 👍
Wow this is so cool
Thanks! -Derek
One of the more clever ways of integrating a "strain gauge" into a device is the Duet3D Smart Effector which serves as the platform for a Delta 3D printer and has its strain gauge done in as part of the PCB.
This is brilliant. I'm not a big 3D printer user, so I had no idea this existed. I'm going to dig into this more. Thank you for sharing!!!! -Derek
Thanks! You can always rely on the Duet3D guys for coming up with something clever. :) But you had to do your bed leveling with the hotend heater off. Because the strain gauge "matrix" acts as an antenna and can throw the measurements off. It took me far longer than it should have to figure that one out. :)
And in the real world of loadcells (with their strain gauges bonded to them), the failures that I have seen fall into two categories:
1) A sharp blow, particularly from the side that de-bonds the strain gauge.
2) Moisture. Often moisture will creep up, or "wick" it's way into the loadcell via the cable - even if the loadcell is rated for IP67 or better.
So your concerns regarding the proper bonding of a gauge are well founded.
so cool! :D
it is like a tiny microphone in modern wireless earbuds! :DDDDD
Fun fact: Murphys law was originally coined in response to a miswiring of gauges in a half or full bridge (i forget which). Apparently the tech wired 2 gauges together so that they cancelled each other out, resulting in a sensor that tested perfectly but read absolutely nothing.
Please consider doing an IOT video for something simple like an esp32 and a humidity sensor or something like that which shows the steps to use an online site for communication from phone to device in a home connected to local wifi. Maybe even a small bit on using MIT AI2 companion to create a small app for it to send data thru to the phone for the sensor and send data back to turn on an led. Thanks
resistivity not conductivity
This dude is my favorite contributor. I love this particular presentation from Element 14. I can't wait to see more like this that is so educational and entertaining.
Thank you Jacob. Nice to get this kind of feedback! -Derek
More DC to Daylight please! What would be a practical application in something like this?
The force readout i made in a previous episode is one.
Making the video was partly self-serving, as I wanted to make a torque sensor for a robot. So this is a good application. If you look up force-torque sensors, usually they're bolted between the robot "hand" and the wrist. If the hand bumps into something, the torque spikes and the robot drops the servo power. Internal to the sensor, is a web of thinner metal that has some degree of compliance (not much), but enough to register a resistance change in the strain gauges. From that you can determine if the robot crashed, thereby preventing damage to expensive transmission components. -Derek
Neat video 😊
Thank you! -Derek
Keep up it ❤
Much appreciated. Thank you! -Derek
Maybe it's the Algorithm, but only a couple of days ago while searching for an old email I stumbled upon a 2016 link I sent myself from a work account regarding "G-putty", which is a mix of Silly Putty and Graphene. It is a good conductor of electricity. However, when the material is subject to even a tiny strain or impact, its electrical resistance increases sharply - before relaxing to its original value as the material “self-heals”. A sensor of G-putty when mounted on the neck and chest of a subject could measure breathing, pulse and blood pressure. The sensor was even able to detect the footsteps of a small spider.
I probably triggered the algorithm when I searched for more info, as it sounded like (no pun intended) a candidate for a better acoustic guitar pickup, which is on my to-do list as I dislike the ceramic ones.
I found "strain gauge" mentioned in my search hits, but I wonder if the dynamic nature of the output would limit its use for static strain measurement.
Strain gauges are used all the time in syringe pumps to measure the force on the plunger of the syringe, the force being proportional to the pressure in the syringe (though friction can vary) They are calibrated with a force gauge. Some have their own pressure sensor that interfaces to the patient tubing, these also have to be calibrated. But some use only low side motor current monitoring which is hard to calibrate and not very good (in my opinion). I think the ones that have pressure sensor, it is a MEMS micro electronic measuring system, its like a strain gauge but is produced by semiconductor lithography, I think the bending of the sensor changes the capacitance as elements are moved closer of further from each other, the change of capacitance can be measured, I think they might be laser trimmed for accuracy ?
Nice demonstration. 👍
Thanks! -Derek
Than you use a HX711 digital amplifier and an Arduino with a display to see ounces (of grams) 🙂
HX711 would also work and I believe it’s 24 bits so plenty of resolution. -Derek
Can’t we just use piezo crystal ? As a strain gauge ?
Piezo crystals would work with a constantly changing dynamic load. As soon as the load settles, deflected or not, the output signal also settles. For static loads like scales, you need a resistive element. -Derek