View Full Version : Discussion Servo Dead Band Hack?
bearcreek
Jan 09, 2009, 09:38 AM
I have 3 Hitec hs-322hd servos and they sometimes do not seem to exhibit a dead band. The specs say it has 5 usec. I am driving them using a computer and my own simple software in real dos mode without the multi-theading problems of microsoft windows. When I run it sometimes one or more servos feel tight without any dead band (this is what I want for a robot). At other times they feel loose with a few degrees of easy movement before the motor pulls it back. At the end of a 5-inch sailboat arm this is a an inch of free play which is not good enough. Could there sometimes be signal noise tricking the servos into moving all the time instead of a dead band? I have 3 feet of 22awg servo wire and a 6 foot parallel cable connected to a poorly shielded desktop computer that may affect it. Could I rewrite my computer software code to "jitter" the signal to eliminate the dead band effect?
TMorita
Jan 09, 2009, 05:16 PM
Sounds like you're using the wrong tool for the job IMHO.
You might look at these:
http://www.servocity.com/html/spg400a_top_mount.html
Toshi
bearcreek
Jan 09, 2009, 10:25 PM
I am using an aluminum arm and it is plenty strong it just needs a shorter dead band timing.
Acetronics
Jan 10, 2009, 09:44 AM
Hi,
Hitec ??? Hitec ???
What's that ??? a new Hong Kong plastic seller ???
Not Joking,
Use serious brands, ... and "Digital servos" ... if you look for precision.
Alain
bearcreek
Jan 10, 2009, 11:45 AM
Hitec makes some good servos, though everyone has their own opinion. Hitec's best "robot" servos are $114 each. If I decide to use 5 servos that would be $570, which is more than I would like to spend for this.
Hitec has a digital servo programmer that can supposedly set the dead band to zero usec.
After adding some variation to my computer timing I have been able to decrease the dead band somewhat, but it may take quite a bit more testing to see if it is accurate enough.
Tomapowa
Jan 10, 2009, 04:45 PM
Even though a programmable deadband function is good (applicable to digital servos only)... your servo "play" will also be affected by sloppy gearing, like that of the cheaper servos you are trying to use. Regardless how you set the deadband or try other means of adjusting it..., a $10 servo will give you $10 worth of performance, nothing more.
jab
Jan 11, 2009, 07:27 AM
The trick then is finding a servo with actual performance worth $10, and not some $3 servo marked up to $10. :)
Acetronics
Jan 11, 2009, 08:45 AM
"Hitec makes some good servos"
Hi, Bearcreek
I just reviewed the Hitec range REAL performances ( measured - not avertised ! ) - "Servormance" test report.
only 4 servos ( high end Digital ones ) are not qualified " deceiving and awful current draw "
lol ...
Alain
Tomapowa
Jan 11, 2009, 10:00 AM
Actually, after testing and experimentation I conducted over the years, I've determined that $10-$20 Hitec servos perform a little better than other (i.e. Futaba) servos, although I've found that mid-priced Futaba servos performed a little better than mid-priced Hitec servos. When you get into digital servos, it's probably a toss-up in performance, but Hitec are a bit cheaper (10-15% less than Futaba). Also, in the past 5-6 years of using both Hitec and Futaba servos, I have yet to have a failed Hitec servo... though I've stripped gears in 2 Futaba servos and one Futaba digital in fact died on me a year ago for no reason. Again, this is my conclusion/opinion after actual test & use of both Futaba and Hitec servos,... not a report someone else put out.
orraman
Jan 11, 2009, 11:20 AM
Alain.
Thank you again.
I was intrigued by your reference to "Servormance" and not having heard of that word went looking via an auto-translation of F3:NE:WS which itself was interesting.
And ended up here where other English only folk may find interest ~~
http://74.125.79.101/translate_c?hl=en&sl=fr&u=http://www.teaser.fr/~osegouin/aeromode/servos.phtml&prev=/search%3Fq%3DServormance%26hl%3Den%26sa%3DG&usg=ALkJrhgv_ujJAtQXT-q437l_ExHlD4mdXw
Dave
Probotics
Mar 03, 2009, 09:46 PM
The 322s have variation in the POTs and can even have a little bit of shift there. I think the issue is probably physical. Its worth the step up to 422/425 to get the ball bearings if you're sensitive about positioning.
If you want to keep everything really cheap think about replacing the POTs with more accurate external ones. You can connect the POT directly to whatever you are driving and carefully eliminate the slop. Not neat, but it works.
andrew b
Mar 05, 2009, 10:23 AM
I'm sorry to say it but IMHO the **economy** brands of servos are just that !, Inside you find sloppy, ill fitting gears, cheap pots, shaky bearings etc.
They may appear fine for a while, but after a bit of use, especially loaded up then they just get sloppier and then thier REAL worth shows up unfortunately.
Ask yourself "why is "X" cheaper than "Y"? "Is it the expensive brands of servo are just overpriced and riding on thier name"? or are the cheap brands using stuff made in worn moulds, bulk market "almost reject" or cheaper less preferable materials.
Anyway, choice is yours, just my opinion. :censored:
vBulletin® Copyright ©2000-2009, Jelsoft Enterprises Ltd.