Is there an easy way to extend the 6 pin JST encoder wires from NEO to Spark max? It is our first time using these and we ordered the 36" extender cables from REV but they are male-male so it is looking like we’ll need to splice them. Is there an easier way? How do others solve this problem?
3 Likes
Thanks!! That makes it easy! Should have seen that when I ordered the others ones.
I would be very careful about extending those wires. The reason they’re so short is because of latency. The motor controller needs to be able to quickly switch power between the three wires on brushless motors and as such needs very fast encoder feedback from the motor. If the cables get longer, it increases latency, and it will eventually cause a loss in performance and possibly damage to the motor.
3 Likes
You’re not going to run into that until the wires get absurdly long for FRC scale.
10 Likes
What would you consider too long
1 Like
I don’t want to pick a specific number and promise it will work without knowing the components at both ends, but I’d be surprised if you run into issue with hall extensions shorter than 20’.
It’s easy to test on the bench and gain confidence.
Latency isn’t the concern, it’s voltage drop due to resistance from wire length such that the halls no longer trigger the digital inputs on the controller.
6 Likes
Yeah, the speed of light is not your enemy here.
5 Likes
Boom.
A motor spinning at 5x00 RPM and the speed of an electron - I know where my money’s going…
The extensions do make it easier to damage part of the cable without realizing it, because there will now be so much more cable to route and protect.
You’ll also want to protect the boards pretty well with how they’re mounted so that they don’t wear through their conformal coating and start shorting against the frame.
(We ran into this same design issue last night and bought a pack of 300mm hobby-grade 6-pin JST PH cables and boards from Amazon instead of the Rev product, at about half the “cost per foot” - though we might end up with two extension boards to get to our intake motor, which isn’t ideal.)
1 Like
I would not expect issues, don’t the SparkMax’s stop using the hall sensors and switch to back EMF for power switching at higher RPMs?
I’m not sure; I have just heard that from a few mentors on my team as reasoning as to why the encoder wires are so short.
Makes me curious which method is used at higher rpms on those motors, obviously at low rpms and high torque situations the hall sensors will be used but the lower rpms means latency has less of an effect.
The real reason these wires shouldn’t be extended is not because of latency, that shouldn’t be a problem with quadrature encoders because of their pretty basic design. The real problem is EMI. When you have a tiny wire with a sensitive signal running over a massive electromagnet switching on and off a few hundred times a second, things get hairy quickly.
1018 used NEOs last year. Had one motor that required extending the sensor cable.
We built our own, and used REV’s adapter board. To my knowledge, the extension was perhaps 2 feet. Didn’t have any issues with operability.
Wondering if there were other experiences with extending the sensor cable.
Did teams twist the wires?
Did teams shield the wires?
Did teams twist and shield the motor power wires?
Signals travel along wires at the speed of light (299 792 458 m / s).
The Free Speed of a NEO is 5676 RPM. If we assume that the Hall sensors puts out one pulse per rotation, the signal would travel 3169053 m in the time of one rotation. I am not sure how long the cables are since I don’t have any NEOs. The latency can be calculated by dividing 299 792 458 m by the length of the cable. My guess is that the latency would be shorter than the clock period of the microprocessor used in the SPARKMAX making it undetectable.
1 Like
Latency is not an issue for the wire lengths involved. Significant latency would affect any sort of encoder.
The size of the wire (AWG) is irrelevant to it’s ability to pick up EMI. The wires running between the motor and the motor controller and between the motor controller and the PDP are more likely to cause EMI problems than the motor windings because they run long distances, and because they are more likely to run near the sensitive wires. The magnetic structure of the motor also captures a lot of the EMI coming from the motor windings.
1 Like
Twisting signal wires together only helps if the signal is differential.
Most encoder wires are serial not differential.
If you happen to twist (or run alongside) power wires you may, in fact, induce EMI
Twisting wires around each other reduces (or even eliminates) the area of your unintentional wire loop antenna. It helps independently of the signals you’re trying to pass.
3 Likes
Only in the case of the 2nd wire being a true ground and not another signal wire (which a encoder using grey code would be)
1 Like