Hi-Res Pressure Sensor Matrix with the LattePanda
HTML-код
- Опубликовано: 17 сен 2017
- LattePanda: www.dfrobot.com/product-1498....
Load Cell: bit.ly/2C52wQ9
Patreon: / marcoreps
This project has quite a lot of interesting details to cover, so I am splitting it into 2 parts. In this first part I am explaining the motivation for and the theory behind my pressure sensor matrix mat. After testing various sensor materials, Velostat in particular, I am demonstrating my new favorite single board computer, the LattePanda. Using its included microcontroller, the first 24 x 24 sensor matrix prototype works great. But I suspect that the final 180 x 60 version will come with a whole new set of problems. Definitely something to look forward to :) 
Наука
Just came across your channel about 2 hours ago and am already 10 videos deep. I love your humor and unbelievable wealth of knowledge and interesting projects. Awesome!!!
Really awesome project, I see a ton of potential in many medical fields. Big thumbs up!
good luck, watched this video cause it got suggested and quickly went on a binge of your content. you just got a new sub continue making cool and interesting stuff
Just found the channel because I'm building something similar and saw you used my patch! What a nice surprise! Great stuff!
Excellent idea and implementation! Thoroughly enjoyed watching.
Thanks Marco! You've got brain the size of a planet. Enjoying every video you made.
Wait, wait! @3:07 "... I don't want to bore you with the trivial math..." Please DO bore me, I would love to hear all about that trivial math (which I likely would never understand).
Nah its really super simple: The dynamic range of a sensor is about 0 ... 1 kg. The mat will be recommended for users with a body weight lower than 200 kg and the smallest area they can possibly distribute that is one footprint. The average adult foot size can pessimistically be assumed to be size 9. Using a drawing software and an anatomically correct footprint I measured the footprint area to be roughly 126 cm². Based on that: 200 kg per 126 cm² (126 sensors) is a maximum of 1.6 kg per sensor, so a little bit outside of their dynamic range. So theoretically I would include 6.25 mm gaps between the sensors. But practically there is only 6 mm copper tape so of course I will use 6 mm² sensors and 6 mm gaps
Marco Reps, that's brilliant! What are the chances that copper tape would be available in almost the perfect width? No wonder this was not possible before.
@@reps I see that you calculated the average force on all the feet area, but actually there is way more force on the heel (for instance), where it will be at least two times the 1.6kg/cm². Isn't that an issue for your objective?
Anyway, good job so far!
@@reps
what about using a pseudo pressurized air mattress ? The inside top surface impregnated with reflective dots 1 mm spacing
an array of moveable mirrors on the bottom to direct a laser beam or light pulse which scans the entire ceiling of the inside surface of the top of the mattress at different angles for a certain desired resolution
the same mirrors would be able to direct the reflection of the light pulse back to a receiver
with the known pulse / beam location and the known ' static ' location of each reflective dot - it would be possible to calculate downward deflection of the ceiling by measuring the increased distance between neighboring dots which have been depressed due to the material being distorted
Interesting stuff as always, looking forward to part 2!
Great project. Great Video. Came across your video searching for earthing. You're an inspiration.
0:24 😁 those little details just perfekt work, all the videos are so much fun to watch
Thanks Marco! Really great video.
Awesome project! I would love to see it work out!
so yeah... "good luck!"
Wow man, that is *much* better than the QTCs I tried a year or so ago. I'm going to order some of those velostat pads :)
Loving the ESD crowbar screwdriver!
Early preCOVID Marco Reps - is the best.
Keep up the great videos!
very interesting, the lattapanda looks good too!
Very useful video. I was deciding between using a using velostat and conductive foam, under the impression that the foam might give more dynamic range than a thin layer of plastic. Additionally the foam would be built into the slightly springy application I needed instead of needing to add foam separately. Needless to say, both of those reasonings got crushed in the span of couple minutes, so now I know what I should be going for.
Good luck on your project
Very creative!!!! Good luck.
thanks a lot for this video! you saved me from using foam in my project :D also, i can confirm that using a better heatsink on the lattepanda cpu will greatly improve performance! I just used a bigass heatsink from my old pc, total overkill xD , but the panda now is about 30 degrees cooler at full load (!), which improves performance for especially the heavier stuff like games by a lot.
Very cool! Good luck!
Good luck! Nice project
I wish you luck Marco.
All the best!
This is one of those projects where a jig is a must. It takes a while to make the jig - but a lot quicker to do than making a single one; once you have it!
Nice project
love the shout out for Amanita design
This is super cool! I've wanted to make an force plate for quite some time. How did youtube _just_ recommend this to me now! I have tried to make some stuff with FSRs before, but they are super fragile and as you said, super expensive. Never knew about this velostat stuff. Thank you!
This is extremely helpful with what I am trying to do but I do not get how you collected the data with all those conductive tape. Also did you have velostat in between the conductive tape
Amazing!!
Interesting way to make a workbench. I won't tell the landlord :)
Thanks for sharing 😀👍
cool channel........all sorts of nifty gadgets !
Very interesting project! Isn''t the high speed parallel data acquisition, serialization and possibly even video output a perfect use case for an FPGA though?
@Marco Reps Great video! Could you please provide the circuit diagram and the respective code at 03:27 .This project caught so much of our attention, and we are so much excited to try it out.
Good luck!
Good luck 👍🏻
Good luck!
What are your thoughts on possible miniaturization of this? In the field of continuum robots there is a need for flexible sensors on a tubular surface with a diameter of about 2-10mm.
dude you are good
Yes!
FAB Marco ;)
Good luck. 🙏😇
I was thinking of using velostat and copper tape for a midi keyboard/drum pad with pressure sensitivity/aftertouch, do you think that the range of pressure would translate well enough or it's better to spend on an FSR strip?
In the lab where I used to work they experimented on sandwiching carbon nanotubes into silicone to get sensors for elongation and flexing.
Do you think it would be feasible to create something like a 'pressure-copier', where a charged dielectric sheet leaks charge proportional to applied pressure (maybe through resistive film such as you are using here), then the charge distribution is made visible using a standard photocopier train?
Aha, Universität Bremen.
Hi Marco,
@7:20 do you measure resistor of a cell?
Congratulations for the video! Do you work for which company?
Ingenious! 0:40 may I use this part for my Virtual Reality Sketch Presentation?
hey, u got a writeup for this / any links to some ? would be great to see how to setup the circuit for copper tracks + scan thru them with the microcontroller :)
When I am testing the resistance vs force characteristics, I experience the velostat to be very unstable. It is rather reacting to changes than actual force. At the moment I find it unusable, but I see that you get much more reliable performance. Any idea what I could be doing wrong? I am just using an ohm-meter and velostat sandwich setup + applying force to the sandwich.
Hello there, I am wondering what the name of the program is that you use to see the pressure readings live with the pixels? How do you program that etc?
Hello,
I saw your video regarding the pressure mat. It was really amazing. We are doing similar project but we are not finding a software to show the mapping. In your video that mapping was really good. If you don't mind may i know which software is that?
Did you find out which software he is using?
Hi Marco, does the resistance you measure at 2:10 stay constant over time? To ask differently - if you placed a 1kgf weight on a pad, and logged its resistance over - say - an hour, would the measured value stay reasonably constant? Most FSRs based on piezoresistive carbon-doped polymers show relaxation-behaviour over time; I'm curious about your observations in this aspect. Thanks! Cheers :)
Interesting! Will look into it
What do you use for taking multiple inputs??? Analog switches?? Do you mind helping me a bit, I am working on something like this 4x4 matrix of 4cm x4cm
I just received a box IC's from the 70's. Some of the ESD foam was become so rigid it broke in pieces if I tried to remove an IC. But it turned out there was no reason to remove them because from most IC pins inside the foam were corroded away.
Fantastic commentary, lol. How might you go about making a pair of pressure sensor socks using flexible piezo-conductive fabric and conductive flexible thread? It would be great to see an in depth look at the programming for this.
I was thinking of the Lattepanda but decided to go with the UdooX86.
Cool!
what about using a pseudo pressurized air mattress ? The inside top surface impregnated with reflective dots 1 mm spacing
an array of moveable mirrors on the bottom to direct a laser beam or light pulse which scans the entire ceiling of the inside surface of the top of the mattress at different angles for a certain desired resolution
the same mirrors would be able to direct the reflection of the light pulse back to a receiver
with the known pulse / beam location and the known ' static ' location of each reflective dot - it would be possible to calculate downward deflection of the ceiling by measuring the increased distance between neighboring dots which have been depressed due to the material being distorted
Ha ha, good vid. I made a 3x3 seat cushion using the esd foam and some conductive fabric to test the exact same concept. Is yours to reduce bed sores?
I am not much involved in the application, but I have seen a certain physiotherapy 'gamification' made in unity. Hope I can get some video material of that in the end
Do you mind if I use your design to this in a project of mine? I've been looking for a way to implement a reliable pressure sensor matrix for a while and this seems like the perfect solution to my problem. Also, are there any issues with increasing the resolution of the matrix by making the square pads smaller and closer together?
In the resolution I have tried the velostat resistance in X and Y direction (length and width) were neglectably high, but when you really want to go to town, you might have to cut the velostat into small sections too. Just a guess ...
Very cool. Why use that SOC specifically? Did you "HAVE" to for some reason you can not say? Pretty much any Micro with a good ADC and 5GPIO pins would do. You could use it to acquire data and then dump it to a higher powered device of choice?
Well they sent me that LattePanda for free to be featured in a video and I thought this was quite a good opportunity.
Fair point :)
Wow. At $90 including the windows 10 licence these thing are a steal.
Am i see here the way to invent again the already using capacitive touch register surfaces?
Considering such surfaces, i had the idea of improving current drawing boards to send the writing input into computer recording using just sound of marker and microphones on x & y axises. Super.
alex ale capacitive usually tells you only where the touch is. This also tells you about the force. One could use capacitive with a flexible dielectric material to get a similar rwsult
I am not able to obtain stable readings from the sensor.
Our setup: We have connected the velostat (1 cm^2) in series with a 220 ohm resistance and
used a voltage divider to get the readings. However, when we apply a constant weight the
readings are not stable (changing every time we measure and the readings are also continuously varying).
Can you give me the best way to calibrate the sensor to get constant reading .
I've seen conductive filaments for 3D printers. They sort of suck at conducting, but apparently they are also pressure sensitive. Using a 3D printer would improve resolution and reduce the necessary build labor.
Hir sir, may I know what is multiplexer module that you used in this project? I'm still confuse whether to use i2c or analog multiplexer
Wouldn't the copper oxidize at some point and change resistance? Back when I tried something like that I electoplated the copper, but that's obviously not gonna work on that scale. Maybe they make some other metalized tape. Nickel would be a could candidate, I think.
where did you get the velostat with the glue backing?
how did you avoid the problem with the current taking a second route through 3 other points
Just curious about aging? or is this not of interst or concern?
Yes, I am curious too about the Velostat aging. Didn't find any data yet. But the adhesive layers are for car wrapping, so they should last a while
Im more thinking about oxidation of the copper etc. also i have noticed that sometimes Soft Buttons age quit fast. But never looked into why
Yes, the copper tape is a good point! It seems to be lightly coated in something, that's why I was wearing rubber gloves for the assembly. Touching it with bare hands leaves oxidized finger prints behind which was also pretty cool honestly imgur.com/a/zAWjI
Could be a oil coating from the rolling mill, oh yes it is quite good to get a finger print image ;). Maybe it is Cr2O3, SiO2 but i really don't know.I used also some copper tape for transformer manufacturing and i also noticed that, I also had once wich where not coated. Im looking forward to see your progress, awesome work!!
Now the whole internet has your finger prints!
I wanna see you play with muscle sensors, then build a exo suit with LattePanda. The German Iron War Machine. How is your fission knowledge, we might need a good power source...
i know im asking randomly but does anybody know a trick to get back into an instagram account?
I was dumb lost the account password. I would appreciate any tips you can offer me
@Nathan Ivan Instablaster ;)
@Manuel Kyle I really appreciate your reply. I found the site on google and I'm in the hacking process now.
I see it takes quite some time so I will reply here later when my account password hopefully is recovered.
@Manuel Kyle It worked and I finally got access to my account again. I'm so happy:D
Thanks so much, you saved my ass!
@Nathan Ivan Glad I could help xD
Hi, Do you have 24x24 Pressure Sensor compatible to Arudion, that I can buy from you for my small experiment?
Can someone please let me know where I can buy the self adhesive foil used to sandwich velostat
7:31 how did you do the plotting ? And Is it possible to plot it in Arduino !!
Where can I find more details and schematics on this project.
how do you know the output voltages from every column and row?
Do you have the schematics for PCB you built?
How much power does it pull @7:21 ?
Can you share the circuit of the test matrix?
Machinarium, classic :)
how to measure weight with the same concept?
Any ways?
I want that screwdriver, I can only see the wowstick on Gearbest.
Video about that one coming up next
Hurry up then, I got my wallet out. It looks better than the Wowstick :)
Why not use HDMI and have a converter interface?
Good project. Your employer ok with you sharing this?
+
Marco Reps
Thanks for another great video about an interesting project.
I have a question though, how does Your employer view the fact that You "publish" things that You develop that at least seemingly pertains to part of Your work ?
((I apologise if the question is impertinent))
Best regards
Maybe some practical sense here is, like measuring the weight of fatso on the chair - to change some parameters, but people can be very mad that device measuring their weight. About medical use - maybe the "smart" bed to register the body moving when dreaming, but such data not really useful.
Can we get circuit diagram for this
3M velostat how to purchase give me a link
The video stops working at 7:23. RUclips just gives an error. Very strange, I've never seen this happen before.
thank god, i though my gpu had started giving up the ghost
Well shoot, I have heard multiple complaints about this now ... but what to do? It seems to work well in some browsers and some devices?
Interesting, it fails in Chrome on my MBP but works fine in Safari. Also works fine in the RUclips app. Maybe just leave it for now and hope RUclips fix the problem in the future. There is not much left of the video when it fails; you just mention that there will be a part 2 to follow.
Fails watching with a Chromecast too
I have added a blurred area to the wall between 0:07 and 0:11 making YT re-render the video (I think), maybe that'll fix it
5:53 Rate my RGB setup
I need to make this project almost exactly. Would you mind sharing code and more details. I would gladly trade you a finished product or even add you to the project if your interested.
Finished project in my newest video, code in that videos description. Also have 2 leftover PCBs ...
Would a vinyl cutter made that matrix easier to make?
How would I contact you in a non public message
My e-mail is in the 'About' tab on my channel, hidden behind a capture I think
Yes, with that you could make a central layer with very precise sensor-sized cutouts, so you don't have to bother with manual alignment of copper strips and release liners.
We did a balancing game with the same principle. But we stacked multiple layers on top to get a greater range.
tim.gremalm.se/interactive-led-balancing-board/
4:00 Latte Panda has big thermal issues, even with coper heatsinks, mine loose the whole filesystem after some days, running only a CNC software doing nothing ! I'm pretty unhappy and do not recommend it, except if you using a water cooling or with a 2 fan (one on bottom and one on top)!!!!!!
Just use the door... xD
Hi marco! Do you received my email (from ermanh...)? Thank you!
Gud luk
Marry me Marco!
Good luck. Oder wie wir Deutschen sagen, viel Glück.😉😀