My Glove Remote Design+Progress Thread

Lol hey I’m self taught through trial, error and forums like these.

The wierd oscillation could be several things. When I started testing my sensors it was immediately obvious securing the sensor at more than one point on the hand does some funky stuff. I have my sensors secured behind my knuckles, the rest of the sensor is free to move along your fingers and I get very stable and linear feedback even without adjusting for VRef and with 5% shitty 80’s resistors.

Make sure all your grounds are going to a single point, ground loops can do some wierd stuff.

Put a 10uF electrolytic cap on you VIN, 3.3v and 5v rails and that should clear up most noise being generated by the ESC.

1 Like

damn, I had this exact idea a few months ago and started looking into the process, but I didn’t get much further as I’m a student and haven’t got much money to throw into things that may or may not work.

I love how this is going and once you guarantee that it’s working I might have a crack at my own version, just have to get good with Arduino coding first.

I’m guessing you will send the code through nrf modules? that would help as you might be able to reuse some of @RollingGecko’s code. Oh and I suggest a dead man’s switch just incase you get a cramp or something.

Good luck and I’m definitely watching this.

Edit: I just looked up and you already talked about using a dead man switch

I’m working off of a greatscott video (that’s how i got started on this whole mess of a journey two years ago lmao) as far as my electronics, but the code I’m doing from scratch because it’s fairly simple.

anyway, update time. Here’s a video of the issue i’ve been dealing with:

the problem is basically shown by 1:10. As I said in the video, both sensors are hooked up exactly the same way and are read by essentially the same code.

any ideas on what i’m doing wrong? it’s acting like the resistor i have in series with the sensor just isn’t there or something but my multimeter reads it to be around 100kOhm

1 Like

ok, i’ve narrowed it down to the flex sensor itself. I don’t think any current is getting through it. I don’t know how it took until now to catch this.

are you able to swap flex sensors to see if same result? DBL check wiring <== usually what happens to me is i miss the row on the breadboard im aiming for.

Yeah I swapped them and it seems the flex sensor itself was the problem, and the reason I didn’t catch it earlier was because my wiring was wrong

ok. I think i know what i did wrong. I burned the flex sensor when i soldered it, making its resistance super high. I did it again on the third one but i figured it out and got my fourth (and final!) one soldered correctly. Good thing i ordered more than i needed! now i need to figure out the algorithm for combining both inputs, and i think i’m gonna run it through the EMA thing on this article to smooth it out a little.

Gah I just saw this now. You have to be super fast soldering to the flex sensor pins. A solid alternate to connecting them is by removing the plastic casing from a female jumper wire, 1/16" heatshrink .5" up the wire jacket to the end of the pin. Make two, stick on the sensor then 3/8" heatshrink it together after placing a dab of hot glue in the middle. That joint will not fail you.

Were you able to find a good source of the flex sensors, the cheapest I could find was about $20

@pat.speed I got mine from store (Micro enter) $8.99 “Adafruit Short Flex/Bend Sensor”

Also have “AdaFruit Force-sensitive Resistor”… same store $6.99

2 Likes

yeah these were the only ones i was able to get. There are the short ones (the ones i’m using) and the long ones (the $20 ones which are better)

I looked at your code. What are you using for resistor values on your divider circuit? Are you using D2 as your high side on the divider circuit?

  1. about 100kohm. They’re super ghetto and not at all within tolerances but they work ok
  2. yeah, is there a better way to do that?

You should never use a digital pin as a power source for analog sensors. You are relying on the internal silicon circuits that are used for everything else inside the 328 and this is the best place to get noise on your sensor circuit. It may seem efficient to be able to turn the divider on or off, but at 5V, 120,000 ohms will only draw 0.000041A and thats negligible current draw in this application.

The nanos are 5v, so your positive voltage supply to the divider needs to be tied to the 5V rail (5V pin).

100k is too high. The flex sensors are about 20k - 10k, and if you put them in with 100k resistors you will only be able to read about 20% of the full range on the ADC.

Swapping the 100k for 10k should open up your bandwidth a lot.

2 Likes

i checked my flex sensors with my multimeter and they actually range from about 35 to 70k, but see if i can swap it to something closer. Switching my voltage supply is a good tip though, and it should knock off a few bytes of code for the attiny

I’ve been watching the glove tech coming along from yourself & @InitialDriveGTR things are looking to come together, so congrats to you both.

I had a question regarding fingers gesture being the control method, had either of you considered proximity sensors or are they too unreliable?

You know, the sensor is on the deck mounted on truck holes, the glove has a small cicuit sown in and you reach out to the front truck to accelerate (throttling up) & pull glove back to throttle down, it the toerances are right then the speed stance could favor it?

I am still curious what will happen if you Fall. You want to absorb the landing with your hands and the Board gets wrong or Bad input. Carvons Board just got totally crashed by wrong input in just a couple of seconds… What happens if your hand got hit by a stone or something else (falling fruits, insects,…) and you make a wrong Hand movement because of the pain / Reflex ?

If you make a Low responding Filter you wont Fall with a wrong Hand movement but you loose the ability to respond quickly for obstacles…

In my mind this controll Mode is too risky. Maybe i do not See the better Plan behind it?

Not a bad idea, but if you used that method, at the distance between your hand and the proximity sensor, the sensor would also detect anything and everything with a magnetic field that you pass by…

That’s very true, I was thinking more around a proximity chip system but not sure of the tech

IR maybes? http://qqtrading.com.my/infrared-proximity-sensor-10-80cm-sharp-gp2y0a21yk0f

sorry for the lack of updates, i left NYC (summer break wooo!!!) and i won’t be back for a couple more weeks. I am free to dick around on here though :wink:

I hadn’t considered a proximity sensor, i was going off the flex thing because i wanted to mimic the motion you make when you’re actually using a trigger-style controller. I think the proximity sensor could work, but i’d be worried about having to make a full-body movement to adjust throttle.

1 Like