The EZ-Robot EZB / Iotiny controllers PWM are compatible with EZ-Robot servos (1-180) and specific us range (not advertised) but I measured before and is specific for the EZ-Robot servos not the common 500-2500 us with middle point @ 1500us.

ARC's EZB' PWM protocol is limited to 180 degrees.

Two options:

1. Custom firmware (Arduino)
2. EZB + servo board.

I found this specs: https://nenovtechnic.eu/electronics/td-8135mg-waterproof-35kg-180-degree-large-torque-digital-coreless-servo.html

It's a 180 degrees using the standard range 500-1500-2500 us.

thanks PTP I guess I will look at arduino. I appreciate the help.

Hey DJ remember when we used to get points for answering questions and we could buy stuff.  How about points towards subscription fees ;)

if you want to do a quick test:

Pick an arduino firmware and change the attach method: https://www.arduino.cc/reference/en/libraries/servo/attach/

.attach(... , 500, 2500)

Then send 1, 90, 180 to test the limits of your servo.

Those 270 degree winch style servos use the same pwm range, in my experience. The degree number and range of the hardware are not actually related. We ,arduino, etc just added an abstract value to the pwm range so it appears a standard servo has 180 degrees

I really like that you’re pushing the limits of what a robot can do. It’s tough with trying to find mechanical pieces that work together - wish there were more robust kits for this sort or thing. Made of light metal or even strong plastic. But mostly I think it’s the tradition servo that needs to be changed. We need some sort of diy joint building kit that reduces the strain on servos. So the torque doesn’t need to be as high with because if the high fulcrum.

So many robot companies have spent hundreds of millions of dollars and bankrupted. Meanwhile you’re sitting at home doing more than they ever did lol. It’s awesome

Thanks I am just having fun while locked up with covid and winter. I agree that hobby servos are a problem. I think the work Mickey and PTP are doing with FOC BLDC motors and Planetary drives will enable a lot more capabilities.  I am hoping the low cost off the shelf BLDC will be the game changer for robotics. The controllers will be commoditized and they should come down in price.

My fingers are crossed as well. It seems robotics is such a slow industry. It’s suffering from broken promises of past robot companies. Expectations were high but nothing useful or reliable was created. And of course the cost of full stack product development bankrupted everyone

I was hoping during covid that we’d see a bunch of useful robot products - also from this community. There’s a higher usage than normal of ARC but not many products in development.

I feel like it’s related to lack of hardware - and as you said, the cost is too high for what is available.

Maybe 2021 is the year? The media has been saying that for many years this is the  year of robots LOL. Okay I’ve done my part by making robots super easy to program, now someone has to take care of the hardware! Anyone?! LOL

Am trying ! A cost effective robot hardware platform , either assembled or in kit form, that can actually be useful in the home is something Ive wanted to do for sometime, like giving arms and legs to an Alexa. A basic off the shelf platform for "some" basic simple home maneuvers. Which of course would require a quick easy to do navigation setup for the home that the robot will reside in. Just getting it to take the trash bag to the trash can would be money maker alone I think. Carry grocery's, home sentry, some education and entertainment function, verbal companionship, would be some other great features, who knows... can buy your home robot from Sharper Image, Best Buy or Walmart in the future It just needs to be  simple design and do simple task, maybe add the ability for the robot owner to add on stuff over time. That's all I ask.

Well for small and cheap I use the ubtech jimu smart servo's. They use a daisy chained 115200 baud 8,n,1 serial interface with 5v power and a 3.3v single wire data line. They turn through +130 to -120 degrees and I've picked up a fair amount cheaply on ebay. They're great for mounting sensors and it's not too hard to adapt the jimu fittings to fit to another robot. I'll try to attach some photos.

any videos on what it can do?

Hi Ezang, sorry don't have any videos yet but it only does object avoidance and follows the wall at present. I don't wish to hijack the thread, just showing how easy these servo's are to connect. The ultrasonics have been in place for about a month (tie wrapped to a couple of jimu connectors) but just got hold of the tof10120 distance sensors. As you can see in the side photo I simply screwed these to another jimu fitting and clipped them on top of the ultrasonics. Now to try some 2D area mapping using the laser sensors.

All the best skidroe

@skidroe.  Also don't want to hijack the thread, but how are you connecting the tof10120 distance sensor to ARC?  (or are you using ARC for it?).  They look interesting, mostly because they are cheap, and being laser I think they may avoid the problems of both ultrasonic and IR for object avoidance.  If you can actually map, even better.

Alan

HI Alan, I'm not using ARC on this robot (or I would have started a new thread about the tof sensors), if you look at the side view picture you'll notice a raspberry pi zero WH which is small, cheap, easy to power up and easy to code. The Tof10120 uses serial (9600,8,N,1) which I use for setting up the sensor or i2c which I use for reading the sensor as a serial read will return a string ie 'L=1000mm' whereas the i2c returns a high and low byte ie (high x 256) + low millimeters. Also the default setting on the 10120 is active reading which means it automatically sends a serial distance every 100ms (this frequency is adjustable) but I have set mine to passive reading which will only supply a distance when requested. The sensor operates on both 3.3v and 5v.

Steve

@skidroe

Thanks.  I imagine I could do something to get the data from the Pi into ARC (mqqt or something), but I have other priorities for now.  Good to know about the possibilities though.

Alan

Here is video of Roomba cleaning up.  I guess if you created a list of all known objects that are often left on the floor you could eventually train it to pick most things up.

Nink, that should be the next upgrade for roomba!!  It clears up it's own obstacles before vacuuming.

Now that's what we need, lol

EzAng

By the way if you struggle to wake up your roomba or need to stop it from freaking out, then you need one of these. An EZB HDD servo that presses the start button.

Now if only I could work out how to make the roomba find its base when it is all of 6 inches away and pointed directly at the base without having it go crazy running around the room and smashing into everything first.

That’s brilliant!!! Haha