Mindboards

I know that for RS-485, only two wires are required; the data wires. Although not 100% necessary, Gnd is highly recommended to be common, as a voltage stabilizer/reference for the two (or more) systems (NXT's).

What I am wondering about, is the affect of sharing the supply (4.n volts). When using just a normal NXT wire to connect the two together, all six pins are connected to each other's. In practice, both NXT's will not be powered at exactly the same times, so it is quite probable that one will be powered at times, when the other one is off (assuming only two NXT's for now). Does it matter that you would then be supplying power from the On NXT to the NXT that is off? I am not sure if it matters, and I haven't seen any warnings about it, but I always try to keep them either both on, or both off. Would it be better for me to cut a wire, and only have three or four of the pins connect through (data0, data1, Gnd, and maybe the other Gnd)?

On a slightly different note, does it matter if I have the NXT turned off, with active pullups (to a different power source) on an I2C bus?

I'd almost be more worried about having them both on at the same time. Slight allowances in component manufacturing and other such things could make one NXT put out 4.68V, while the other is supplying 4.76V (example only). I'm no expert, but that doesn't seem like it'd be very good for the NXT. If you could wire a resistor in-between the two power rails you'd most likely be fine, but cutting the wire does seem to be a quicker/easier solution. I'd buy some NXT plugs from Mindsensors, half a Km of two or four strand telephone wire, a crimper and get to work . But, that's probably not what you wanted to hear...

About your second question. I think it'd be preferable to power both circuits off at the same time however, the NXT is fairly well protected and shouldn't go boom!

I hadn't really thought about that issue, but I have heard that the voltage can be anywhere from 4.3-4.7 volts (I never measured it). Because of this, it doesn't seem to be very regulated, so I think it should be fine (if they can just give or take a little). I will probably cut another cable though, and play it safe from now on.

One of my robot arms ran with that back-to-back configuration for about 3 weeks while I programmed it and then again for several weeks while on display. It never had a problem.

- Xander

mattallen37 wrote:I hadn't really thought about that issue, but I have heard that the voltage can be anywhere from 4.3-4.7 volts (I never measured it). Because of this, it doesn't seem to be very regulated, so I think it should be fine (if they can just give or take a little). I will probably cut another cable though, and play it safe from now on.

If I remember correctly, the voltage is regulated to 5V, but somehow it droops, something to do with other components or something like that. I may very well be wrong, so take this with a grain of salt .

mightor wrote:One of my robot arms ran with that back-to-back configuration for about 3 weeks while I programmed it and then again for several weeks while on display. It never had a problem.

Well that's nice to know . As the NXT is fairly well protected, it should be fine long term... but then again, I'm no EE!

Ok, thank you for the information.