Ideal number of IEMs?

I want to use Integrated Encoder Modules on my drive system, and was wondering if there is any difference between using four IEMs and two IEMs. If there is a difference, how much does it matter? Also, if I were to use two IEMs, where would I put them?

@Coffee Thanks! Sometimes it’s better not to over think it. Do you know the dimensions of the red ones (just so I can plan ahead for sizing)?

Another advantage of the big red quadrature encoders is that they eliminate most backlash and are therefore more accurate.

Use -1
IME == failure

That’s not 100% true. Look at the Cyberbrains’ NBN robot. They had so many sensor fail safes including a back up encoder on their flywheel in case one broke. They were my role models and made me feel that double flywheels could work. Not that there weren’t LOTS of other teams out there, but they deserve special attention.

Anyhow, yeah, USE 0 IEM’S on your robot. They are the worst sensor made by VEX (yes, even worse than the gyro).

They worked decently for me, as long as you are careful about clearing encoder values before use in every move. Which is what you should do for both types of encoders.

I’ll talk about choosing encoders for your drive train in autonomous here.

The biggest problem I have with integrated encoders is the slack issue. The integrated encoders measure directly from the motor itself inside the housing, before the compound gear reduction inside the motor, before your drive train gearing slack outside of your motor, and before your wheel-shaft slack if you didn’t lock your wheels mechanically to the shaft. What happened a lot for me was that before the drive train even physically moved, the integrated encoder already counted 100 or 200 ticks of slack. This causes inconsistency in autonomous programming.
The drawbacks with the big red encoders are 1. take up digital ports and 2. too big. But they measure more closely to your actual robot displacement or whatever actual displacement you want to measure. If you mount them directly to your drive shaft and then lock the wheel to the shaft, probably your only source of inconsistency is wheel slipperage, which doesn’t happen a lot for omni wheels at regular speed.

So if you can work your way around not having enough digital ports and sensor taking up too much space, I’d strongly recommend big red encoders for autonomous drive train. A lot of the times for drive train you don’t even need bi-directional reading, which means you really just need one port for one big red encoder most of the time.

Some of the world’s best programming routines are done with two big red encoders on two sides of the drive train. I believe both QCC2 and AURA did that for skyrise breathtaking auton routines. 1103 curahee had that as well for the most legendary programming skills routine in VRC like ever. I’d recommend this particular drive setup for simple, reliable auton programming.

@Coffee No, I wouldn’t say they are useless. They just aren’t useful for absolute position. Definitely not good for base or for lift height. However, we used them without trouble for NBN launcher speed. The reason was that the team was updating the speed loop every 20ms and taking the difference between the present and last count only. So, if they lost their mind due to static it only lasted for one 20ms cycle.