search

JorgePe / BOOSTreveng

Reverse engineering the LEGO BOOST Hub
MIT License
168 stars 35 forks source link

Train motor AutoID/resistor? #31

Closed nathankellenicki closed 4 years ago

nathankellenicki commented 6 years ago

Hi Jorge,

Do you happen to know/are you able to measure the resistance for the AutoID on the train motor?

I ask as using a 2k2 resistor should mimic a WeDo motor, but when used with a Powered UP controller, this will result in "bang-bang" style on/off operation. If we can mimic a train motor instead, then it will allow continuous variable power to be provided when using a PUP controller.

Cheers, Nathan

JorgePe commented 6 years ago

Yes: 3 to 6 and 4 to 5 (no resistors, just shunts)

On Thu, Aug 9, 2018 at 1:45 PM Nathan Kellenicki notifications@github.com wrote:

Hi Jorge,

Do you happen to know/are you able to measure the resistance for the AutoID on the train motor?

I ask as using a 2k2 resistor should mimic a WeDo motor, but when used with a Powered UP controller, this will result in "bang-bang" style on/off operation. If we can mimic a train motor instead, then it will allow continuous variable power to be provided when using a PUP controller.

Cheers, Nathan

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/JorgePe/BOOSTreveng/issues/31, or mute the thread https://github.com/notifications/unsubscribe-auth/AICXa9VNhR0lXv6WTmNmM_0TmRF9xBWwks5uPC7xgaJpZM4V1oEB .

nathankellenicki commented 6 years ago

Aha thank you!

Do sensors and the Boost motor adhere to the same resistance based system, or do they have a more advanced handshake?

JorgePe commented 6 years ago

Uff...! It's not that simple. They do have an AutoID but they also have some sort of microcontroller inside and it seems that they negotiate something before they can be used.

I got a poor man's logic analyser. Spent last weekend looking to the BOOST motor. It was the hottest week here in Portugal, I got quite an head ache... and almost zero progress. The BOOST motor (as in an Eurobricks post, train section) as small microcontroller inside. Pins 5 and 6 are connected to TX/RX (not sure if in this order, need to re-check). The microcontroller can also use it as A/D inputs. And from the LEGO hub side it clearly uses them also as A/D pins for the simple AutoID.

There is 300 bps talking on pin 6. No sense what it means yet. But there is no talking on pin 5. Or at least almost, perhaps just a bit every 2 seconds or so. Very strange.

It's easier to sniff BLE than UART with this gadget, I need more time. But if what I got is correct, I see two options:

I hope it's last option because half-duplex is a pain. And there is no sense in doing half-duplex when you have a full UART available... unless you plan to create some kind of network with just one data wire. But that is also crazy when you already have I2C or SPI on modern microcontrollers - the NXT and the EV3 use it very well.

On Thu, Aug 9, 2018 at 2:06 PM Nathan Kellenicki notifications@github.com wrote:

Aha thank you!

Do sensors and the Boost motor adhere to the same resistance based system, or do they have a more advanced handshake?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/JorgePe/BOOSTreveng/issues/31#issuecomment-411749904, or mute the thread https://github.com/notifications/unsubscribe-auth/AICXazp9Ex99Gs8DC1M871I-eFqaS6Eqks5uPDPKgaJpZM4V1oEB .

nathankellenicki commented 6 years ago

I see, I figured it was slightly more complex for those types of devices. I don't have much hardware experience so I'll have to make myself happy with the simple AutoID's of the wedo/train motors. :) Mimicking a train motor will suffice for the conversion cable I plan on making (since Lego has disappointingly said they won't be making one...)

Thanks!