I’m designing a mini CNC and need to know the exact gear ratio for these Gearmotors. Do the Micro Metal Gearmotors 10:1 and 5:1 have this exact ratios or are them approximate?
I’ve taken apart a 10:1 model gearbox from the motor and taken a picture of the motor pinion and then examining the photo I counted 13 teeth, but the gearbox itself is pressure assembled and I’m afraid of spoiling it if I try to disassemble it in order to apply the same procedure for counting the number of teeth.
Does someboby know the number of teeth of the different gears/pinions in the gearboxes or the global exact ratios?
Goes without saying, don’t forget that gears have backlash when you’re designing your system. I hope it goes well, as non stepper motor CNC systems are pretty innovative and interesting.
I looked into it some more, and it turns out that the exact numbers we had corresponded to older versions of the gearboxes from a different supplier. To get those numbers for our current micro metal gearmotors, we will have to physically take each gearbox apart and count all the teeth. We’ll post back here when we have those counts.
The decimal values look good. I think there might be a typo in the 298:1 fractions since they don’t equal 297.9238:1, but the 30:1 being (31/16) × (33/14) × (35/13) × (34/14) looks right.
Pololu provides awesome support - Thank you!
You are right, Clyde; sorry about that. I edited Grant’s post to fix the typos (one of the /10s should have been /13), and I think the two calculations were mislabeled.
I’m using the 50:1 gearmotor that your web page says is exactly 51.45:1. I am using the backside encoder and trying to monitor the output angle of my mechanism. I am getting a systematic error that would seem to indicate the true gear ratio is closer to 51.49:1. Can you please tell me the exact set of gears in the gearbox? I cannot directly examine it because I’d have to disassemble my mechanism to see the gearmotor.
The exact gear ratio equation we provide on the product page in the “Details for item #” section indicates the how many teeth each of the gears have and the order they are meshed from left to right, starting with the pinion gear. Here is the equation for our 50:1 Micro Metal Gearmotors:
So the 15-tooth motor pinion gear is meshed to the 32-tooth part of a compound 32-14-tooth gear, which is meshed to a compound 33-13-tooth gear, and so on until you reach the 38-tooth gear on the output shaft.
How are you determining that the gear ratio is 51.49:1 instead of 51.45:1? That amount of error seems very small; can you explain more about the error you are observing in your application?
The output stage of my mechanism is driven through a 54:11 ratio synchronous belt. I rotated the output stage through 40 revolutions and looked at the output of the encoder to see how many encoder counts I got. I then used that result to compute the measured gear ratio.
I am reading the encoder using an Arduino Uno. The A pin is connected to an interrupt pin. The beeping is sampled on each interrupt. This method gives me 3 counts per revolution of the motor.
I suppose it is possible that I am missing interrupts because they are coming too fast. Assuming the motor itself turns at 30,000 rpm, that’s 500 rev/sec which would be 1500 interrupts / sec allowing for 667 usec per interrupt. That’s about 11k instructions (at 16 MHz and 1 ipc) for handling the interrupt which seems like plenty. Especially since all my ISR does is read the B pin.
If you are calculating a gear ratio value that is a little too high like this, then I think that suggests that your system is reporting too many counts, not too few.
I think your algorithm for measuring the encoder likely is fine, but there are probably several ways error around this magnitude could accumulate, especially in a test like this where you are turning the motor through several rotations using some external mechanism. Since the encoders are attached to the motor shaft, backlash and slop from additional mechanisms can cause small errors to built up. If that is a problem for your application, you might consider looking into high resolution encoders that can attach directly to the output stage of your system.
