View Full Version : Determining Motor Speed?
Hi all,
Right now we have a CIM-U-LATOR gearbox hooked up to a modulox gearbox.
We want to get the speed of the overall output.
So, a few questions
1) Does the encoder fit on the CIM shaft?
2) If it does, then I assume we'll need to put it on the CIM shaft and perform math in the code to find the speed after all the gearing within the gearbox?
3) It's very high up on the robot, and the encoder wire seems short. Any suggestions?
EDIT:
Looking into some parts, I found
http://www.andymark.com/product-p/am-0174.htm
http://www.andymark.com/product-p/am-2040.htm
http://www.andymark.com/product-p/am-2043.htm
Are these the parts needed?
it looks like there's a circular part of the hex shaft, and it is meant to protrude from the gearbox, and you attach the encoder there, while the hex protrusion on the other side is used as the normal provided hex shaft would be.
1) I dont think so. I believe on the toughbox, the drive axle is machined down to the correct diameter on the backside to accept the US digital encoder. I have no experience with your gearbox.
2)Correct
3)You will need to solder in some extra wire and then solder some female PWM connectors on the end to connect it to the DS. (White - signal; black - ground; red- v+)
3)You will need to solder in some extra wire and then solder some female PWM connectors on the end to connect it to the DS. (White - signal; black - ground; red- v+)
DS? I thought the encoder went into the Jaguar.
DS? I thought the encoder went into the Jaguar.
There are two options to connect the encoder.
One is to connect it to the DSC; this option will require two Digital I/O channels because of the nature of the encoder output. If you take this route, make sure that you take care soldering the wires (there are alternatives to connect the wires; search the forums to find them) as they are easily damaged by the heat, and that you solder them correctly.
The other is to, as you note, connect it to the Jaguar via the encoder interface. Note that the E4P encoders provided by USDigital do not have index lines, so that pin is unused. Accessing encoder output through this option is only possible with Jaguars on CAN. Jaguars on PWM cannot react to encoder input nor send it back to the cRIO. You must be on CAN to access the encoder input. I'm not familiar with the specifics; I'll let someone with more knowledge on that elaborate.
Looking into some parts, I found
http://www.andymark.com/product-p/am-0174.htm
http://www.andymark.com/product-p/am-2040.htm
http://www.andymark.com/product-p/am-2043.htm
Are these the parts needed?
it looks like there's a circular part of the hex shaft, and it is meant to protrude from the gearbox, and you attach the encoder there, while the hex protrusion on the other side is used as the normal provided hex shaft would be.
I'm not familiar with the specific parts you're using, but yes, the cylindrical portion of the shaft is where the encoder reads the shaft's rotational speed. If you still have FIRST Choice points left, you may also want to look into getting the encoders from there (http://www.andymark.com/Encoders-2-pack-fc12-66-p/fc12-66.htm).
The US digital encoder is for a 1/4 shaft. The CIM-U-LATOR has a 8MM shaft. will not work. The MOD box adapter shaft will work. Just adjust the encoder counts by the gear ratio. Most of the AndyMark gear boxes has the encoder on the output shafts.
Encoders are sensitive to shaft end play & egocentricity. So make sure the shaft does not have a lot of float & is well aligned.
Splice longer wire to the provided encoder wire. Stranded Ethernet cable works. Cable is better than individual wires. Nothing bigger than 18 gauge. Try to keep the cable away from power wires especially motor power.
Everyone,
I'm looking at the 360 and 250 count encoders.
According to the calculation provided on the US Digital site, a lower count encoder can handle a higher Max RPM.
Our launcher is over 10,000 RPM (which is the max for a 360 step), so I'm thinking of getting the 250 step.
What do you think?
Our launcher is over 10,000 RPM
What is your wheel diameter?
Note that the FPGA can only handle ~39000 pulses per second, or about 6500 RPM with a 360 count encoder.
Our launcher is over 10,000 RPM
That is a lot of RPM for wheels. A lot more than what your typical AndyMark wheel is probably design for. Of course you might well be using something else.
Our launcher is over 10,000 RPM
Have you actually measured this with a tachometer?
Keep in mind that motors' free speeds are not the same as their speed under load.
We're using the 2011 FIRST Wheels, 8" Diameter, with a Banebots RS-775 up to a CIM-U-LATOR.
This CIM-U-LATOR is then stuffed into a Modulox gearbox, geared for 1:3
We're using the 2011 FIRST Wheels, 8" Diameter, with a Banebots RS-775 up to a CIM-U-LATOR.
This CIM-U-LATOR is then stuffed into a Modulox gearbox, geared for 1:3
How do you get 10,000 rpm out of that?
They may actually have a free speed RPM of 10000. But that would be only if the gear ratio really was the 2.7:1 in the cimulator followed by the 1:3 that he is talking about. I don't think that it will ever get that fast though.
They may actually have a free speed RPM of 10000. But that would be only if the gear ratio really was the 2.7:1 in the cimulator followed by the 1:3 that he is talking about. I don't think that it will ever get that fast though.
You're not suggesting that they're gearing down 2.7:1 and then gearing back up 1:3 are you? What would be the point of that?
But suppose they were.
775-18 free speed is 13000 rpm.
13000/2.7*3 = 14444 rpm.
So if the output is 10000 rpm then the 775-18 would be spinning at 69% of its free speed. It would be pulling 28 amps and burning over a hundred watts of waste heat. (I'm assuming here that the OP meant 10000 rpm at 12 volts; maybe that's not what he meant).
In any event, the tangential speed of an 8" diameter wheel at 10000 rpm is 238 miles per hour.
jwakeman
13-02-2012, 23:48
We can get our 8" shooter wheels up to 112 feet/sec (76 mph) and it is scary!!
Grim Tuesday
14-02-2012, 00:47
We had ours at about that speed, and we realized we really don't need it. You should be at something around 20 Feet/sec to be making half court shots. Having them too fast makes them draw more power, and raises the spin up time for repeating shots. Furthermore, it makes your realm for adjustment considerably lower. I would suggest changing your gearing, if you can.
Our dual 775's running through cim-u-laters freespin our 8" wheel up to just over 3600 RPM, as measured with a photo-tach. Works out to about 125 fps, or 85 MPH. We use around half power to shoot from the key, but since we're using a single axle we only get half the wheel velocity as ball velocity (the rest is backspin).
At full power it's max range was right around 30 feet. We're using a high angle fixed hood, so it's not optimized for max range.
jwakeman
14-02-2012, 08:52
Our dual 775's running through cim-u-laters freespin our 8" wheel up to just over 3600 RPM, as measured with a photo-tach. Works out to about 125 fps, or 85 MPH.
Am I doing my calculation incorrectly? For 3600 rpm with 8" wheel I got 40 fps.
(3600 / 60) * (8/12) = 40
We're using a high angle fixed hood
We are also using a hood with a fixed angle of 65 degrees.
Am I doing my calculation incorrectly? For 3600 rpm with 8" wheel I got 40 fps.
(3600 / 60) * (8/12) = 40
3600 rpm / 60 = 60 rps
8 inches * pi = 25.13 inches per rev
60 rps * 25.13 ipr = 1508 inches/sec
1508 inches/sec /12 = 126 feet/sec
Our dual 775's running through cim-u-laters freespin our 8" wheel up to just over 3600 RPM, as measured with a photo-tach.
free speed of 775-12 at 12 volts is 7300 rpm. CIM-U-LATOR gear reduction is 2.7:1
7300/2.7 = 2704 rpm So you can't be using 775-12.
free speed of 775-18 at 12 volts is 13000 rpm
13000/2.7 = 4800 rpm
3600/4800 = 75%
775-18 at 12 volts at 75% speed is drawing 23 amps, and cranking out over 200 watts. And you say you have 2 of them. Something doesn't add up.
jwakeman
14-02-2012, 11:31
3600 rpm / 60 = 60 rps
8 inches * pi = 25.13 inches per rev
60 rps * 25.13 ipr = 1508 inches/sec
1508 inches/sec /12 = 126 feet/sec
Oops that's right, circumference not diameter. I did have it right in our code so we max out at 112 feet/sec which is about 3200 rpm.
We're using 775-18's.
Something doesn't add up.
What doesn't seem to add up? Running the two motors at 100% power through two Victors gives us the 3600 RPM as measured at the wheel. Looks like we're losing about 25% of the potential motor free-spin RPM through the speed controllers, gear train, shooter axle bearing misalignment, etc.
jwakeman
14-02-2012, 11:37
free speed of 775-12 at 12 volts is 7300 rpm. CIM-U-LATOR gear reduction is 2.7:1
7300/2.7 = 2704 rpm So you can't be using 775-12.
free speed of 775-18 at 12 volts is 13000 rpm
13000/2.7 = 4800 rpm
3600/4800 = 75%
775-18 at 12 volts at 75% speed is drawing 23 amps, and cranking out over 200 watts. And you say you have 2 of them. Something doesn't add up.
We are using two 775's (I guess the -18's) and we max out at 3200 rpm as mentioned above. We are using chain drive with sprockets. They look like 1" sprockets on the motor shafts and 6" sprockets at the wheels (I'm not sure, I just help with the software :) ). It takes something like 10 or 12 amps on each motor just to get the thing moving and I think it was drawing 20 or 30 amps each at full speed.
We are using chain drive with sprockets. They look like 1" sprockets on the motor shafts and 6" sprockets at the wheels
That's the part you left out.
What doesn't seem to add up? Running the two motors at 100% power through two Victors gives us the 3600 RPM as measured at the wheel. Looks like we're losing about 25% of the potential motor free-spin RPM through the speed controllers, gear train, shooter axle bearing misalignment, etc.
775-18 motor at 12 volts and 75% of free speed is generating over 200 watts of output power. Times two is 400 watts. If it's taking well over one-half horsepower to drive your unloaded free-spinning shooter wheel and you're happy with that then go for it.
If it's taking well over one-half horsepower to drive your unloaded free-spinning shooter wheel and you're happy with that then go for it.
That's what makes the "real world" vs paper theory interesting! ;)
That's what makes the "real world" vs paper theory interesting! ;)
What's "paper theory" ?
What's "paper theory" ?
Calculating theoretical system performance without testing as validation.
Is the battery delivering the full 12 volts? We hope so, we haven't bothered to measure it under load.
Are the Victors really delivering the full battery voltage? They're off our 2007 robot, who knows? There's a definite audible difference between running the motors hooked up directly to a battery and running them through the FIRST control system (there's probably some reason that there's a fan on the Victors).
Are the 775's performing to spec? They're the same ones we used on our forklift last year, so they already have one full season plus a year of demo's on them. They aren't even getting warm after 3 or 4 hours of practice, so we plan to keep running them.
Lots of system variables in there that aren't accounted for if you're using a nominal P=IR^2.
It's currently scoring well over 90% from the key. Due to the ball variables, the best we could hope for with a lot of additional effort is a couple of percent improvement if we're lucky. Plenty of other things higher on the priority list before we worry about that!
theprgramerdude
14-02-2012, 13:25
You're not suggesting that they're gearing down 2.7:1 and then gearing back up 1:3 are you? What would be the point of that?
But suppose they were.
775-18 free speed is 13000 rpm.
13000/2.7*3 = 14444 rpm.
So if the output is 10000 rpm then the 775-18 would be spinning at 69% of its free speed. It would be pulling 28 amps and burning over a hundred watts of waste heat. (I'm assuming here that the OP meant 10000 rpm at 12 volts; maybe that's not what he meant).
In any event, the tangential speed of an 8" diameter wheel at 10000 rpm is 238 miles per hour.
I'm pretty sure the photosensors for counting shots will still register individual chunks of ball landing in the nets, even if the rest of the ball is obliterated on launch :yikes:
It's currently scoring well over 90% from the key. Due to the ball variables, the best we could hope for with a lot of additional effort is a couple of percent improvement if we're lucky. Plenty of other things higher on the priority list before we worry about that!
That's pretty darn good. Don't mess with success!
What's "paper theory" ?
Calculating theoretical system performance without testing as validation.
Ah. Never heard that phrase before. If that's what "paper theory" means I would not rely on it. Always validate simulations before trusting their predictions. That's what real-world engineers do.
they aren't even getting warm after 3 or 4 hours of practice
Does anyone know how long a 775-18 will run at 12volts & 9750rpm before it gets warm?
vBulletin® v3.6.4, Copyright ©2000-2017, Jelsoft Enterprises Ltd.