Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   Technical Discussion (http://www.chiefdelphi.com/forums/forumdisplay.php?f=22)
-   -   Gearing for a Very High Speed (http://www.chiefdelphi.com/forums/showthread.php?t=129907)

JorgeReyes 27-06-2014 13:44
Gearing for a Very High Speed
 
I am wondering if this would be a good idea. To have a robot that is 2 speed and 6cims. Low for around 7 fps and high would be 18-20 fps. The High might sound uncontrollable but what I was thinking is that we could use a button so that when it is pressed, the motors use 100% power but when it isnt it uses 50%. This would essentially give us 4 speeds with a 2 speed gearbox.

Joe G. 27-06-2014 14:03
Re: Gearing for a Very High Speed
 
20fps isn't unheard of, but it typically takes some fancy non-linear response and/or closed loop control with encoders to make it controllable for the drivers.

Jay O'Donnell 27-06-2014 14:15
Re: Gearing for a Very High Speed
 
I'm not sure how reducing the motor power to 50% is really helping you. If a high speed isn't controllable, why not just change the gearing so that is at a more reasonable speed? If you reduce the motor power by 50% then you'll barely be going faster than your low gear and you'll have half the torque that the low gear has.

Pratik Kunapuli 27-06-2014 14:27
Re: Gearing for a Very High Speed
 
During the 2014 season, our robot had a 6 cim, 2 speed drivetrain with a low gear ~4 fps and high gear ~17 fps. The only real times we used high gear was if we needed to travel the length of the field very quickly, or to maneuver around defenders and not push through them. Realistically, your driver will learn to control the high speed and will learn when to shift into high gear. Its mostly a "feel" thing and if your robot needs to go 20 fps, then he will learn to drive it well.

waialua359 27-06-2014 14:37
Re: Gearing for a Very High Speed
 
Quote:
Originally Posted by Pratik341 (Post 1391364)
During the 2014 season, our robot had a 6 cim, 2 speed drivetrain with a low gear ~4 fps and high gear ~17 fps. The only real times we used high gear was if we needed to travel the length of the field very quickly, or to maneuver around defenders and not push through them. Realistically, your driver will learn to control the high speed and will learn when to shift into high gear. Its mostly a "feel" thing and if your robot needs to go 20 fps, then he will learn to drive it well.
You guys spent most of the time driving 4 fps?:confused:

T^2 27-06-2014 14:47
Re: Gearing for a Very High Speed
 
We spent most of our 2014 season in high gear with 6 CIMs (geared to 23ft/s, ran at 17ft/s in practice). The acceleration we got was pretty great, and our driver could control it with practice. We played midfield exclusively, however, so we ran back and forth across the field a lot.

Having your motors throttled to 50% power most of the time seems completely useless. In low gear, you lose the pushing advantage of being in low gear, and in high gear, you lose the torque advantage of using 6 CIMs. I'm not sure how this would ever be useful in practice. It is pretty easy to test, though, so feel free to do that.

notmattlythgoe 27-06-2014 14:50
Re: Gearing for a Very High Speed
 
Quote:
Originally Posted by waialua359 (Post 1391365)
You guys spent most of the time driving 4 fps?:confused:
This is the same reaction I had. This can't be the whole story.

Our drive spent about 95% of his time in high gear and we were geared for about 17fps.

DampRobot 27-06-2014 14:56
Re: Gearing for a Very High Speed
 
Quote:
Originally Posted by JorgeReyes (Post 1391351)
I am wondering if this would be a good idea. To have a robot that is 2 speed and 6cims. Low for around 7 fps and high would be 18-20 fps. The High might sound uncontrollable but what I was thinking is that we could use a button so that when it is pressed, the motors use 100% power but when it isnt it uses 50%. This would essentially give us 4 speeds with a 2 speed gearbox.
We had adjusted speeds in this range this year, with 2 CIMs and a mini on each side. It was ferocious when accelerating, but ended up tripping the breaker a lot. When we took out the mini CIMs, it wasn't nearly as high performance (to my eye).

With a reasonable talented driver, a good control system, and a bit of practice, you won't need to limit power at all. We used split arcade (faux Chezy Drive), and noticed that the robot was much, much easier to control than with tank. When people think fast drives are uncontrollable, it's usually more a testament to what control system they're using rather than how fast they're going.

JorgeReyes 27-06-2014 15:23
Re: Gearing for a Very High Speed
 
Well here was where I was going with it. If we gear it to 20 fps with 3 cims, then when the robot starts moving the motors could use full power but then since we use thrustmaster joysticks, we could control the power and maybe set it to say 75% after it accelerates by releasing a button so that you get a controllable speed but still get the acceleration. The Low Speed would be full power.

We kind of did this this year with our mecanums. We pressed a button( we called it the turbo button) for full power but then released it so we could go a bit slower when there were obstacles close by.

Edit: Also with our joystick it is a lot easier to control than using logitech controllers and I drove this year so I am pretty confident I should be able to control the robot at these speeds with practice

T^2 27-06-2014 15:48
Re: Gearing for a Very High Speed
 
Quote:
Originally Posted by JorgeReyes (Post 1391375)
Well here was where I was going with it. If we gear it to 20 fps with 3 cims, then when the robot starts moving the motors could use full power but then since we use thrustmaster joysticks, we could control the power and maybe set it to say 75% after it accelerates by releasing a button so that you get a controllable speed but still get the acceleration. The Low Speed would be full power.

We kind of did this this year with our mecanums. We pressed a button( we called it the turbo button) for full power but then released it so we could go a bit slower when there were obstacles close by.

Edit: Also with our joystick it is a lot easier to control than using logitech controllers and I drove this year so I am pretty confident I should be able to control the robot at these speeds with practice
If you're using joysticks, why not just ease up on them? That way your driver will get an infinite degree of control between 0% and 100% power.

JorgeReyes 27-06-2014 15:53
Re: Gearing for a Very High Speed
 
I guess so. I think we are just going to do a lot of testing this summer with different gear ratios and control set ups.

John 27-06-2014 16:14
Re: Gearing for a Very High Speed
 
Quote:
Originally Posted by Pratik341 (Post 1391364)
During the 2014 season, our robot had a 6 cim, 2 speed drivetrain with a low gear ~4 fps and high gear ~17 fps. The only real times we used high gear was if we needed to travel the length of the field very quickly, or to maneuver around defenders and not push through them. Realistically, your driver will learn to control the high speed and will learn when to shift into high gear. Its mostly a "feel" thing and if your robot needs to go 20 fps, then he will learn to drive it well.
This is similar to what we did: we stayed in low gear (~8-9 fps) for the majority of the match, mostly using high gear (~19 fps) to make full field runs. Many teams greatly overestimate how much time is spent traversing the field (suitable for high speed), and underestimate how much time is spent lining up passes and shots (suitable for low speed). ~4 fps seems really slow, even to me though. What shifter did you guys use that has a 4.25:1 spread?

A really fast high gear is sometimes just what you need, though. For example, in Galileo Match 100 (no video) our alliance partner missed a would-be match winning shot with less than 20 seconds left, and the ball bounced all the way back to our defensive zone. We were able to race back, get the ball, and score it just before time expired.

Some years full field runs happen all the time (2013, 2011 come to mind), others they almost never happen (2012). Whether the ability to save a couple of seconds on each run is worth it depends a lot on your team and the year. It's worth maybe one more cycle, but only if your team can score effectively and quickly.

Quote:
The High might sound uncontrollable but what I was thinking is that we could use a button so that when it is pressed, the motors use 100% power but when it isnt it uses 50%. This would essentially give us 4 speeds with a 2 speed gearbox.
I question why you want to have a secondary "software shifter" when you are planning to have a real shifter on your robot. Why not just use low gear when you want to go slow? In our experience (if I have time I might add the math to back it up) it should give superior acceleration, control, and power consumption. The only drawback I can think of is shifting time, but the COTS shifter options are so good now its not really a big deal.

JorgeReyes 27-06-2014 16:32
Re: Gearing for a Very High Speed
 
I guess the question I have then is if it is better to automate shifting to shift into low gear when we are drawing too much current and high when we are not, or have the driver shift manually.

notmattlythgoe 27-06-2014 16:34
Re: Gearing for a Very High Speed
 
Quote:
Originally Posted by JorgeReyes (Post 1391381)
I guess the question I have then is if it is better to automate shifting to shift into low gear when we are drawing too much current and high when we are not, or have the driver shift manually.
We plan on monitoring the current draw and shifting down when we sense that we are in a pushing match. Otherwise we will stay in high gear.

JorgeReyes 27-06-2014 16:38
Re: Gearing for a Very High Speed
 
Quote:
Originally Posted by notmattlythgoe (Post 1391382)
We plan on monitoring the current draw and shifting down when we sense that we are in a pushing match. Otherwise we will stay in high gear.
See but the problem I see with this set up is when you want to do something precise like lining up to shoot. If you try to do this on high gear it will be harder to do this. What I was thinking was that it shifts down when it draws to much current to automatically shift in pushing matches like you said.

But also have a manual override option to shift when necessary like the situation I mentioned.

John 27-06-2014 16:42
Re: Gearing for a Very High Speed
 
Quote:
Originally Posted by JorgeReyes (Post 1391381)
I guess the question I have then is if it is better to automate shifting to shift into low gear when we are drawing too much current and high when we are not, or have the driver shift manually.
This thread may be of interest to you, then.

notmattlythgoe 27-06-2014 16:43
Re: Gearing for a Very High Speed
 
Quote:
Originally Posted by JorgeReyes (Post 1391383)
See but the problem I see with this set up is when you want to do something precise like lining up to shoot. If you try to do this on high gear it will be harder to do this. What I was thinking was that it shifts down when it draws to much current to automatically shift in pushing matches like you said.

But also have a manual override option to shift when necessary like the situation I mentioned.
We will also have a manual way to shift down. But our driver has never had an issue lining up to shoot in high gear. Both this season and last our driver was in high gear in all situations except pushing matches.

Pratik Kunapuli 27-06-2014 17:51
Re: Gearing for a Very High Speed
 
Quote:
Originally Posted by John (Post 1391380)
~4 fps seems really slow, even to me though. What shifter did you guys use that has a 4.25:1 spread?
Quote:
Originally Posted by waialua359 (Post 1391365)
You guys spent most of the time driving 4 fps?:confused:
Quote:
Originally Posted by notmattlythgoe (Post 1391368)
This is the same reaction I had. This can't be the whole story. .
Being in high gear is different from trying to drive at top speed. While we may have been in high gear, I wasn't trying to go very fast a lot of the time, only for the full field dashes. We used low gear when there were defenders near us, we needed to push (partners and opponents), and when we were lining up shots. We had a custom 2 speed gearbox.

Ether 27-06-2014 21:04
Re: Gearing for a Very High Speed
 
Quote:
Originally Posted by John (Post 1391380)
if I have time I might add the math to back it up
Here's some quick rough calculations using some simple numbers to illustrate a point.

Say you have a CIM with 10.4 volts applied and loaded with 206.1 oz*in of torque. It will be spinning at 1416 RPM, drawing 80.5 amps, generating 215.8 mechanical output watts, and running at 25.8% efficiency.

Now add a 3x speed reduction gearbox. For sake of simplicity for this simple calculation, ignore gearbox losses. To produce the same output speed and torque as above, the CIM needs to produce 68.7 oz*in torque at 4248 RPM. To generate that torque at that RPM, the CIM will need 12 volts applied. It will draw 28.8 amps and be operating at 62.5% efficiency.

That's why you want to operate in low gear when possible appropriate.

Code:
Motor Calculator  build MCALC_2014d 2/3/2014 1255pm


Select motor:
1)CIM    6)FP0673  d)FP2719  f)FP9012    g)FP9015    h)FP9013 j)MiniCIM
a)am-0912 b)am-0914  7)DensoL  8)DensoR    c)Denso0160 k)BAG    u)UserDefined
2)RS395  3)RS540    4)RS550  e)RS775-12  5)RS775-18  m)RS555  n)am-0915
r)am-2193 s)am-2235  t)RS390  v)RS545      p)am-2161&2194
w)VEX2177hi  z)VEX2177std  1


CIM FR801 001, AM802 001A @ 12.00 volts:

@ free (no load):
  oz-in      Nm    rpm    rpm%    amps  watts out  watts heat    eff%
    0.0  0.000    5310  100.0    2.7        0.0        32.4    0.0

@ stall:
  oz-in      Nm    rpm    rpm%    amps  watts out  watts heat    eff%
  343.4  2.425      0    0.0  133.0        0.0      1596.0    0.0

@ max power:
  oz-in      Nm    rpm    rpm%    amps  watts out  watts heat    eff%
  171.7  1.212    2655    50.0    67.8      337.1      477.1    41.4

@ max efficiency:
  oz-in      Nm    rpm    rpm%    amps  watts out  watts heat    eff%
    42.8  0.302    4648    87.5    18.9      147.1        80.2    64.7

Select input:
1)oz-in  7)watts_in    3)rpm    5)amps  b)rpm&amps  8)eff%
2)Nm    6)watts_out  4)rpm%  9)volts  a)rpm&ozin  m)main menu  x)exit a


Enter motor rpm & ozin, and the program will calculate the motor voltage
required to produce the specified ozin at the specified rpm.

Enter rpm and ozin, separated by a space: 1416 206.1

CIM FR801 001, AM802 001A @ 10.40 volts:

@ free (no load):
  oz-in      Nm    rpm    rpm%    amps  watts out  watts heat    eff%
    0.0  0.000    4603  100.0    2.3        0.0        24.3    0.0

@ stall:
  oz-in      Nm    rpm    rpm%    amps  watts out  watts heat    eff%
  297.7  2.102      0    0.0  115.3        0.0      1199.3    0.0

@ max power:
  oz-in      Nm    rpm    rpm%    amps  watts out  watts heat    eff%
  148.8  1.051    2301    50.0    58.8      253.3      358.5    41.4

@ max efficiency:
  oz-in      Nm    rpm    rpm%    amps  watts out  watts heat    eff%
    37.1  0.262    4029    87.5    16.4      110.6        60.3    64.7

CIM FR801 001, AM802 001A @ 10.40 volts:
  oz-in      Nm    rpm    rpm%    amps  watts out  watts heat    eff%
  206.1  1.455    1416    30.8    80.5      215.8      622.0    25.8

Select input:
1)oz-in  7)watts_in    3)rpm    5)amps  b)rpm&amps  8)eff%
2)Nm    6)watts_out  4)rpm%  9)volts  a)rpm&ozin  m)main menu  x)exit a


Enter motor rpm & ozin, and the program will calculate the motor voltage
required to produce the specified ozin at the specified rpm.

Enter rpm and ozin, separated by a space: 4248 68.7

CIM FR801 001, AM802 001A @ 12.00 volts:

@ free (no load):
  oz-in      Nm    rpm    rpm%    amps  watts out  watts heat    eff%
    0.0  0.000    5310  100.0    2.7        0.0        32.4    0.0

@ stall:
  oz-in      Nm    rpm    rpm%    amps  watts out  watts heat    eff%
  343.4  2.425      0    0.0  133.0        0.0      1596.2    0.0

@ max power:
  oz-in      Nm    rpm    rpm%    amps  watts out  watts heat    eff%
  171.7  1.213    2655    50.0    67.9      337.1      477.2    41.4

@ max efficiency:
  oz-in      Nm    rpm    rpm%    amps  watts out  watts heat    eff%
    42.8  0.302    4648    87.5    18.9      147.2        80.2    64.7

CIM FR801 001, AM802 001A @ 12.00 volts:
  oz-in      Nm    rpm    rpm%    amps  watts out  watts heat    eff%
    68.7  0.485    4248    80.0    28.8      215.8      129.4    62.5

asid61 28-06-2014 02:55
Re: Gearing for a Very High Speed
 
I see nothing wrong with gearing for a very high speed, provided you provide some kind of current detection for autoshifting or speed limiting in emergencies.

Aren Siekmeier 28-06-2014 05:40
Re: Gearing for a Very High Speed
 
Quote:
Originally Posted by Ether (Post 1391397)
Lots of good math showing considerable differences in current draw
This is really good to see, and demonstrates the importance of gearing properly for the designed load.

The loads on an FRC drivetrain can be to accelerate the robot, to maintain speed by overcoming drag, or to push with some force at a constant speed (perhaps close to zero).

In the third case, the load will be continuous, so you should definitely heed the current figures Ether's shown. You don't want to be pulling 80A per motor continuous, or you will pop some breakers (a 40A breaker in about a second, or for 6 six motors for a total of 480A, the main breaker in a few seconds as well).

In the second case, the load will be continuous as well, but much much lower.

In the first case, the acceleration is only transient. 6 CIMs geared 5:1 on 4" wheels get to 90% of top speed (about 18 ft/s with no loss) in about half a second. They are stalled briefly for the beginning of this, but then quickly approach free speed, so there aren't too many amps in too few seconds. In comparison, 6 CIMs geared 13:1 on 4" wheels get to half the high gear's top speed in about a quarter of a second, so while you are putting less load on the system (getting to 90% of low gear's top speed in just 0.1 seconds), the performance improvement due to starting in low gear and shifting at the right time is hardly noticable at all. Even at a distance of 5 ft, the low gear takes 50% longer to complete the traversal than the high gear, and of course it's even more in favor of high gear for longer distances. For this reason, we often remain in high gear unless we see that we're applying load continuously (pushing).

These numbers are all pretty approximate (and also coincidentally almost match our 2014 drive), but I think they show why high gear is important, too. Note that the ability for this high gear to still accelerate as it does is in part due to the 2 extra CIMs in the drive. For a 4 CIM drive, you would not get the same performance over short distances at these reductions, and low gear would be more favored.

JoshD99 29-06-2014 22:09
Re: Gearing for a Very High Speed
 
I am the main driver for team 4327. The robot we built this year moves at about 16 FPS. At first, it was difficult to drive. But now, after becoming very familiar with the robot, I can drive it well, others on the team who don't usually drive, however, do have issues. So I don't think, with practice, a robot with a speed of 20 FPS would be an issue.


All times are GMT -5. The time now is 01:50.

Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
Copyright © Chief Delphi