NXT not working with external power?

[inxt-generation]

Lacking an appropriate power supply, I'm trying to power my NXT with a model train controller. I've made "fake" batteries using the technique described near the bottom of this page. When I put them in, the NXT makes a "click" sound, and then won't respond. When I put batteries back in, it acts normally. I've tried this setup with a couple of my RCXs, and it works fine. Is there any reason why this wouldn't work with the NXT? It seems to me like it would work pretty much the same as this.(minus the protection elements)

Any ideas as to why it isn't working? I've tested the output, and it supplies a regulated whatever-the-dial-is-at output(made sure it was at ~9V DC), and a good 3A. Are there some power specifications of the NXT that I don't know about?

[mattallen37]

Well certainly be sure you have the polarity right!

Be very careful... anything over 10.5v has potential to permanently damage the NXT H-Bridges (motor controllers).

IIRC, at 12v, the co processor (AT MEGA 48) is supposed to shut down the NXT. At that voltage though, you most certainly shouldn't expect the NXT to ever work again (it's probably a fire preventer, not an NXT saver).

And no, I don't think there are any special requirements. Just be sure you have a good clean < 10v DC supply, and it should work fine.

[inxt-generation]

IIRC, at 12v, the co processor (AT MEGA 48) is supposed to shut down the NXT. At that voltage though, you most certainly shouldn't expect the NXT to ever work again (it's probably a fire preventer, not an NXT saver).

Fortunately, it IS a NXT saver(or the voltage wasn't high enough to break it). The controller appears to supply quite a bit more voltage when something is connected. This is probably a bad idea. Just turning the NXT on is causing the voltage to jump between 8-10V. Guess I'll have to invest in a good regulator.

[mattallen37]

Probably the supply was not filtered very well, and connecting it to the NXT supplied capacitors to filter (and raise the average) voltage.

Most linear regulators are limited to around 1A, and remember that you either need a LDO regulator, or you will need to keep the input voltage about 2.3 volts higher than the output.

Just curious, but what voltage did you read it to be?

Also curious if your motor controllers still work...

[inxt-generation]

Just curious, but what voltage did you read it to be?

Also curious if your motor controllers still work...

Nothing connected: 9.5V. NXT connected: ~12V. Dial turned to a lower setting with NXT turned on: 8-10V.

Thankfully, it still works as it should. Are the H-Bridges in the NXT that much different than the RCX? I was once (accidently) pumping 16v through a RCX 1.0, and it still works fine.

[mattallen37]

Wow, you are very fortunate!

The H-Bridges are completely different in the RCX and NXT. And by the way, the PF IR Receivers use the same H-Bridges as the NXT does for outputs B and C.

[pepijndevos]

So are they different for port A?

[inxt-generation]

So are they different for port A?

Yes. Ports B&C use the LB1836M motor driver, whereas port A uses the LB1930M motor driver. Basically, the operational difference is that port A is always supplying power to the motor except when it's in "float" mode. For more info, and a useful thing that takes advantage of this, see Philo's "Deriving power from motor port A" hack.

[mattallen37]

...Basically, the operational difference is that port A is always supplying power to the motor except when it's in "float" mode...

Basically true, but to expand on that: Motor port A supplies vcc to both sides of the motor to brake. Motor ports B and C use ground instead. There will be no noticeable difference when using standard motors (NXT motors, RCX motors, PF motors...).