Modding Servos for Continuous Rotation..... / Computer Science / Forums

Forums

4hv.org :: Forums :: Computer Science

« Previous topic | Next topic »

Modding Servos for Continuous Rotation.....

previous 1 2 3 next

Move Thread

LAN_403

Patrick

Wed Sept 12 2012, 05:36PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

... ok ill try just the two 2.2k resistors and then ill try a basic stamp and play with the PWM,then post the results here.

Patrick

Thu Sept 13 2012, 06:58PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

the problem im having now is:

when i first click the switch to run the Basic program, it starts rotating really fast (2x) then over the course of 2 seconds it slows to 1x, if i re-run the program, it wont happen again. even if i unplug the battery for 10 seconds.

If i plug the battery back in and leave it of for 5 or more minutesand then run the progam. it will be fast then slow again.

its really confounding, the way i see it, it must be the HS-55 servo circuit board. ?

EDIT: yep its the HS-55, it has a non repeatable, non-linearity, when static inputs are made. when i put a airtronics servo in parallel both servos run and only the HS-55 shows the problem.
Ok if the hardware or the coding is crap, then ill chop out the PCB from the HS-55, then put in an SMT H-bridge. since i only need direction and one speed...

Carbon_Rod

Fri Sept 14 2012, 08:33AM

Registered Member #65 Joined: Thu Feb 09 2006, 06:43AM
Location:
Posts: 1155

Note the servo control loop is trying to "match" the input reference pulse signal with a PLL.

Usually cheaper servos are less forgiving of the timing ranges, and that is why you can form a beat-frequency on the input. Try 20Hz & 50Hz around 20% on times, and slowly reduce the on-time through the entire range.... you will find the acceptable ranges vary widely among brands.

Deviating from the tolerances too much can impact the center band, as sometimes the axial variable resister is not simply a voltage divider reference in some brands....

Also, sometimes it is better to calibrate your controllers software with an oscilloscope than assume the delays are consistent with the observed ranges found by sweeping the operational ranges.

IntraWinding

Fri Sept 14 2012, 12:46PM

Registered Member #2261 Joined: Mon Aug 03 2009, 01:19AM
Location: London, UK
Posts: 581

When I looked at servos recently I found they just replace the pot with a couple of equal value resistors which add up to about what the pots end to end value is.
They generally refer to a servo modded like this as a "360Â° servo", which is very annoying if, like me, you're looking for a servo that actually works as a servo over 360Â° rather than as a speed controlled motor. I gave up trying to get a search engine to sift out what I wanted from the others...

Carbon_Rod

Wed Sept 19 2012, 03:44AM

Registered Member #65 Joined: Thu Feb 09 2006, 06:43AM
Location:
Posts: 1155

I just remembered some cheap knock-off mini servos can use a logarithmic scale Audio turn pot.

@Patrick:
You may wish to measure the sweep on another unit to see if the resistance changes linearly or not...

Cheers,
Rod

Patrick

Wed Sept 19 2012, 03:52AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Carbon_Rod wrote ...

I just remembered some cheap knock-off mini servos can use a logarithmic scale Audio turn pot.

@Patrick:
You may wish to measure the sweep on another unit to see if the resistance changes linearly or not...

Cheers,
Rod

i said : "crap", when i first read this, but remember carbon rod, it only happens once on power up, then never agian, its almost like a cap charging...

Conundrum

Wed Sept 19 2012, 06:07PM

Registered Member #96 Joined: Thu Feb 09 2006, 05:37PM
Location: CI, Earth
Posts: 4061

Wonder if anyone has attempted to use a small magnet from an optical deck assembly with a linear HE sensor?
That ought to work, and be highly repeatable.
-A

Carbon_Rod

Thu Sept 20 2012, 05:08AM

Registered Member #65 Joined: Thu Feb 09 2006, 06:43AM
Location:
Posts: 1155

Patrick

Thu Sept 20 2012, 06:37AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Ive seen the squiggle motors before, but how much $, and can i get one 26 inches long?

Carbon_Rod

Fri Sept 21 2012, 05:47AM

Registered Member #65 Joined: Thu Feb 09 2006, 06:43AM
Location:
Posts: 1155

I am sure they could be sold by the furlong, but the price may be out of budget.

Small Firgelli linear actuators are $9 a pair if you know what to buy:

Some people even make them (note the relative size of the resister in the background):

If precision isn't an issue, than telescoping two pieces of greased 8' box channel aluminum should be trivial.

However, hobby-shops should also sell the <1/4" (its mostly metric) hard-brass versions which could have a metal follower-nut brazed into one end of a inner hollow piston, a roller-bearing cup set in the outer sleeve for the leadscrew, and a "belt-drive" wheel assembly with quadrature interrupter patterns cut into it.
...note adding a zero/home switch is trivial if the frame is grounded

A note about direct-drive gears, the cheap ones are only as good as their weakest tooth.
This is why small CNC machines & lathes prefer belt drives on power output sections.

Note small linear servos are cheap:

previous 1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob