Eight pressure-sensitive Velostat/Linqstat pads for a velocity-sensitive MIDI controller

Рет қаралды 11,827

3 жыл бұрын

Here I've made eight more Velostat pads similar to the single one shown in the previous video ( • Using pressure-sensiti... ) and test out their properties as well as show the inside of the previous prototype pad. These new pads have a smaller range of resistances (typically 6K-9K instead of the 4K-15K measured from the previous pads) but do seem to react much more reliably.
Unfortunately, I don't think my microcontroller is up to the task of handling all eight pads with its single ADC. The controller feels somewhat laggy and imprecise (not a good feeling for a musical instrument!) and can miss brief taps. Dropping the number of pads being handled greatly improves the performance, so I think proper performance is reliant on a high sampling rate. When only a single pad is being tested the pad feels very responsive and doesn't miss brief taps.
I still think the Velostat could be used as a cheap way to make velocity-sensitive pads for DIY musical instruments/MIDI controllers, but you'd need to keep your ADC performance in mind.

Пікірлер: 32

@deltaray3 2 жыл бұрын

Great idea using the foam. Looks so much like production drum pads. I just started playing with velostat and am surprised how sensitive it is. I was able to get it to register a slight change in the reading when I physically dropped a resistor (less than a gram) on it to give it some weight.

@murrrr8288 3 жыл бұрын

Cool idea! I might try it later as well if I can get velostat :)

@kapilso 2 ай бұрын

Super work .

@synthgal1090 Жыл бұрын

re: calibration, it may be a good idea to use a small weight placed on each pad instead of your hand, to rule out any differences in pressure you make.

@petersords6435 8 ай бұрын

I think the reason the resistance is lower than expected(when unpressed) is becasue the width (length of the two side conducting bands) is longer - meaniing you can model is as more resistors going across the bands in parellel and hense loweing the resitance. For a bigger dynamic range - you may want to try reducing the width (side with the bands) and increasing the length (non-conductive dimension inbetween the conductive bands)

@AndrewKroll 3 жыл бұрын

Nice idea. I've done similar using the foam pieces from through hole ICs. Works the same way, basically free, but not durable.

@benryves 3 жыл бұрын

Thank you, that's a good idea to experiment with that too!

@AndrewKroll 3 жыл бұрын

@@benryves Eventually I have to see if 3D printed conductive filament does this too.

@mallappa1234 Жыл бұрын

About the resistance decreasing when you increased size of the material. That is what is expected. Wider material conducts more electricity (means less resistance).

@Gin-toki Жыл бұрын

Funnily enough, if the material is kept as a square, it wont change it's resistance with size, only if the shape changes to a rectangle will it affect the resistance. Check out sheet resistance, or ohm per square. Quite interresting. The change in resistance that @benryves experiences, is most likely due to difference in the shape of the copper foil that contacts witht he velostat film, thus changing the resistance.

@mallappa1234 Жыл бұрын

@@Gin-toki Thank you. I did not know that. Makes sense now.

@AlenHR 2 жыл бұрын

Thanks, very informative. I just got a couple of velostat sheets. Looking for ideas on how to use them. Was thinking about some foldable pocket pad array with android connectivity. Or some knee straps with integrated pads.

@AlenHR 2 жыл бұрын

You are using atmega16 with an internal clock? Maybe you could double that sampling rate using an external crystal. If I remember correctly, atmega16 is limited to 8Mhz using an internal oscillator. But it can go up to 16Mhz with an external crystal.

@benryves 2 жыл бұрын

@@AlenHR Thank you, for you kind comments and tips. :) I'm using an external 16MHz resonator in the video, the problem isn't so much the CPU frequency but the ADC clock frequency (recommended to be between 50kHz and 200kHz, with a conversion taking 13 ADC cycles). That would give up to ~15kHz sample rate which is great for a single pad, but switching between eight of them gives under 2kHz each and I found the pads became too unresponsive at this speed to be useful. According to the datasheet you can clock the ADC above 200kHz if you don't mind a lower-resolution conversion, but I didn't get around to experimenting with this too much to see if low resolution but higher sample rates would produce more reliable results.

@GSXRRizla Ай бұрын

I have zero knowledge of electronics but found your video interesting.... Out of curiosity is it possible to make the pads slightly larger, say 10cm by 10cm. I'm thinking how this could be used for less able people to stimulate certain reactions. Any response appreciated

@benryves Ай бұрын

I'm not sure how well they'd scale up to larger sizes - you might have a smaller change in resistance so you might need to press harder or have a more sensitive circuit to measure them; alternatively you could try dividing the larger pad into a few smaller pads underneath. Could certainly be interesting to experiment with, at the time I was playing around with this I was trying to replicate the small pads on MIDI controllers so never tried anything bigger than what you see here.

@benrish1095 10 ай бұрын

Nice work! Your pads look really good wrt cost and tactile feel. I have been experimenting with 72MHz STM32 micro which has 3 ADCs (up to 24 inputs). The peripheral bus runs at 84MHz and supports DMA transfer. I have it working as an oscilloscope driving a Nokia 5110 phone display. It gives pretty detailed piezo sensor graphs. It should work fine with 16 pads. One thing I noticed is that on current Akai gear the pads are externally multi-plexed. I am wondering if that is necessary these days now that low cost microcontrollers offer so much configurable I/O. I have some questions about your setup. Are your pads externally multiplexed? Is it stable mechanically and wrt ambient temperature? Have you noticed any cross talk (false triggers) between the pads? Also, I see below you bought an LPD8. What do you think of it as a midi controller? Would it be possible to hack into the pads or replace the micro without destroying it? Thanks, Ben

@benryves 10 ай бұрын

Hi, sorry for the delay in getting back to you! I'm using the ADMUX built into the microcontroller rather than any external multiplexer (one of the performance limitations is only being able to sample one pad at a time). I never got around to saving the calibration data for pads so don't know how stable it would be in the long term. Never had any issues with cross-talk, though! The LPD8 is fine... once you insert some plastic shims between each rubber pad and the conductive membrane (I cut mine from a plastic sandwich wrapper). By default you need to hit the pads pretty hard to register, I guess they erred on the side of caution to avoid false triggers. Quite a few videos on KZfaq show similar mods (e.g. putting electrical tape in as a shim) so it's not specific to my unit. Not sure about hackability, but as the device speaks class-compliant USB-MIDI I'd be tempted to just use a USB host microcontroller rather than start cutting PCB traces.

@blueangel333333 3 жыл бұрын

I was just trying to look for how to do a sensitive finger-drum-8-pads without a piezo ..... I don't find anything 'cheap' and effective (exept a force sensor resistor for 8€)

@excrubulent Ай бұрын

Okay two things, number one I absolutely plan to find a way to hook up a set of ADCs to a standard arduino to make a large number of responsive pads. I've been researching how to make velocity sensitive MIDI pads and this is one of the only useful results I've found. Secondly, what is that microcontroller? Is it the IC from an arduino or something else? Is there any good information on where to get these and how to do it? I'm looking at making some custom PCBs for my projects and having ICs instead of daughter boards would make them much cleaner.

@benryves Ай бұрын

Thank you! The microcontroller here is an ATmega324P which is related to the chips found in some Arduino boards. The Arduino system tends to involve a bootloader preloaded onto that chip, a USB to serial interface to load a program onto the chip via the bootloader and a bunch of libraries to ease development in C, whereas here I'm just using the plain chip with no bootloader and no USB interface (I program it via the ICSP pins and an USBtinyISP) and I don't use the Arduino libraries, I write the code in plain C using avr-gcc as a compiler (the AVR datasheets are very clear on how to access on-chip peripherals like the ADC via their registers with plenty of sample code). In this case as I wanted to be a USB MIDI device I used the V-USB library which implements USB in software, if you search the web for "V-USB MIDI expression pedal with capacitive sensor" you should be able to find a good sample project that gives you an idea of how a USB-MIDI device using a cheaper AVR might look.

@excrubulent Ай бұрын

@@benryves Thanks so much! This is great info, I can imagine actually making some viable professional-seeming products this way. And I've found the github project by that name, I think I can go from here. Thanks again :)

@shivamacharya8837 2 жыл бұрын

Hey I wanted to know a little about the calibration of resistance, could we get in touch?

@TheConwenk 2 жыл бұрын

Hi there! I’m a guitar maker out of Woodstock NY and I was looking to purchase this exact load out for a prototype I’m working on. Is there a good way to reach you so that we could talk about you creating a kit for my specific needs?

@benryves 2 жыл бұрын

Thank you very much for your interest! Unfortunately I was not able to get a setup that I was properly happy with, I don't think the microcontroller I was using was able to sample eight pads quickly enough to provide good results (a single pad would work pretty well and be quite sensitive and responsive once calibrated, but having eight made it quite unresponsive). In the end I just bought myself an LPD8 and didn't continue the project, sorry.

@mrdrgonzo Жыл бұрын

What microcontrollers did you test on? Ideally I’d like something with just 4 maximum but like to know the limits, cheers :-)

@bambusbambusowski3028 2 жыл бұрын

Hi, Is it possible to use velostat to measure weight?

@benryves 2 жыл бұрын

I think so, though I'm not sure how accurate it would be!

@reshephgani9798 2 жыл бұрын

Does phone battery has velostat paper inside???

@benryves 2 жыл бұрын

I don't think so, but I would not recommend dismantling them to find out as they can react violently if accidentally shorted. Velostat is used for anti-static packaging, so if you've got any black plastic bags that had electronic chips in they might be usable.

@reshephgani9798 2 жыл бұрын

@@benryves ohh Thanks for helping :)