Subject:
|
Re: Adder/Subtractor revisited
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Sat, 6 Feb 1999 03:58:40 GMT
|
Viewed:
|
1074 times
|
| |
|
|
In article <36B6F795.4B9310A6@nada.kth.se>, johannes@nada.kth.se says...
>
> Cyberia wrote:
> >
> > This is not what we did with our models.
> >
> > In the method we, and others have used, you apply the motor power to the
> > carriers of the differentials. There are two possible places to apply the
> > drive force to each carrier (24 teeth or 16 teeth). You do not have to use
> > the same one for each carrier, this permits you some degree of control of
> > the steering ratio.
> >
> > The output shafts can drive wheels directly or through reduction, or tracks
> > through some serious reduction.
> > Tracks require more torque, particularly when turning. Although the A-S
> > transmission seems ideal in this respect, the single motor for
> > forward/reverse requires toom much reduction to be useful (unless your bot
> > is not in a hurry).
Ingenious as it is, this thread and previous similar ones have pointed out the
primary problems with the A-S drive approach: power transmission through the
LEGO differential, which is minimal at best, and a rather high frictional loss
due to complex gear train. For these reasons, I decided that the independently
driven track approach is preferable, using software, instead of gears to
synchronize the system. My current design uses independent motors to drive
each track. The two drive gear trains are also interconnected through a single
differential, the sole function of which is to synchronize the two sides. Since
the additional gearing used to drive the differential is not under any load, it
contributes negligibly to friction. The outer housing of the differential is
connected through gearing to a LEGO rotation sensor. Each track gear train
drives one shaft of the differential. Gearing is arranged so that when both
tracks are driving in the same direction, the two differential shafts
counter-rotate, resulting in no net rotation of the outer housing, as long as
both sides are synchronized. A combintation of three RCX tasks monitor the
rotation sensor and dynamically adjust motor power when differential rotation
is detected. So far, this system is working *very* well. As soon as I get
some decent drawings done, I'll post the mechanical design. I've invested much
effort into optimizing the gear train design and have reached the point where I
don't think it can be further simplified, but I'd love to be proven wrong...
|
|
Message has 2 Replies:
 | | Re: Adder/Subtractor revisited
|
| That sounds very interesting. I hope to see some diagrams or photos soon... Your approach solves the biggest problem introduced by the A-S drive, that of insufficient torque when driving tracks. Two motors is the way to go for tracks. -- SeeYa ! (...) (25 years ago, 6-Feb-99, to lugnet.robotics)
| | | Re: Adder/Subtractor revisited
|
| (...) Actually, I've had pretty good luck with independently driven left and right drive trains; I haven't had a lot of problems with robots curving when I want them to go straight. I find the adder/subtracter useful for a different reason: it makes (...) (25 years ago, 6-Feb-99, to lugnet.robotics)
|
Message is in Reply To:
| | Re: Adder/Subtractor revisited
|
| (...) Right. This is what I was trying to get at. If one motor gives too little power, you can add the power of two together trough a differential. Power the axles of the differential by two different motors, and the added power comes out through (...) (25 years ago, 2-Feb-99, to lugnet.robotics)
|
17 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
