[Solved] Sabertooth blinking

[Stephanie]

I have two signal cables connected to my Sabertooth 2x5. I use an Axon MCU.

The cables are very very bad (almost no solder, the solder is really bad, some parts are black because of the soldering iron).

When there's nothing connected to signal, there's nothing but the strong blue LED (everything is fine).

When I connect the signal cables to the Sabertooth, the LED's ok, but the motors start moving randomly, even if there's NOTHING on the other side of the cable (not even a turned-off Axon --- it's really nothing).

When I connect them to my turned-off Axon, it not only moves the motors randomly, but also the blue LED first glows dimly and then blinks strongly.

I intend to fix the cables tomorrow and tell here what happens. But, anyway, does anyone have a clue about what's happening?

PS: I don't know if this has something to do with it, but the robot chassis is made of aluminum and seems to be conducting electricity. I'm using spacers.

[sonictj]

1. Check and make sure all gnds are common.
2. What are the positions of the switches on the sabertooth.
3. What voltage are you putting into the sabertooth bat + and - terminals if your near 6V the sabertooth will act strangely.
4. How is the Axon connected to the sabertooth.

[Stephanie]

It's a 12V battery and 12V motors.

It's non-lithium R/C MCU mode.

It's connected through digital signal (s1 and s2).

I fixed the cables and now it's ok. But why did the motors move randomly whenever I put a cable there, even if they weren't connected to anything? Did the Sabertooth think it was receiving some signal?

[sonictj]

It sounds like the sabertooth's ground line was "floating".  Meaning it was not grounded to anything, or there was poor contact.  This will cause erratic behavior.

[Stephanie]

Now the motors move strangely and the blue LED blinks again.

Well, there's M1A, M1B, M2A, M2B connected to two 12V DC motors.
Then there's S1 and S2 connected to the Axon's E3 and E4.
Then there's B+ and B- to a 12V NiMH battery.

Do I have to connect something to 0v and 5v? or they're just "free to use" outputs? How do I "unfloat"?

[chelmi]

Now the motors move strangely and the blue LED blinks again.

Well, there's M1A, M1B, M2A, M2B connected to two 12V DC motors.
Then there's S1 and S2 connected to the Axon's E3 and E4.
Then there's B+ and B- to a 12V NiMH battery.

Do I have to connect something to 0v and 5v? or they're just "free to use" outputs? How do I "unfloat"?

EDIT: removed stupid advice. I'll go hide somewhere

[sonictj]

No don't connect anything to the 5+ on the sabertooth thats a bec line providing 5V.  Just connect gnd from the axon to 0V.  The sabertooth provides its own 5V vcc for its logical circuitry.

[Stephanie]

Thanks for the replies.

Don't worry, Chelmi 

Are all the GNDs the same? Or is there a distinction between the grounds from the side rows and the grounds from the ADC row?

[sonictj]

All grounds on the axon I believe are common.  You might as well get gnd from one of the power busses near the pwm lines going to the sabertooth.

removed stupid advice. I'll go hide somewhere Sad

Everything you said is generally right.  Don't feel bad everybody makes mistakes.  I hope you didn't think I was yelling.  I just used bold to make sure my comment wasn't glanced over.

[Stephanie]

Alright. 

The UART has "T, R and G".

So for serial communication I connect:

- T to S1
- R to nothing
- 0v to G (my understanding is that the UART G pin is only to make connecting the triple Molex easier, it could be any other ground pin).

Is that correct?

[Webbot]

You also need to check the Sabertooth mode of operation.
If its in simple mode (ie max of 1 Sabertooth board) then it allows you to configure baudrates etc.
But if you are using packetized mode (ie support mulitple Sabertooth boards in parallel) then you cannot config the baudrate. It uses the first character it is sent which it assumes to 0xAA to auto-config the baud rate. Sounds good - but I spent ages fixing my library so that a power on of the Axon didn't end up sending a 'glitch' character to the Sabertooth, which then made it set its baud rate incorrectly, which meant it just went AWOL.

[sonictj]

(my understanding is that the UART G pin is only to make connecting the triple Molex easier, it could be any other ground pin).

yes, but when ever you connect any device to the axon you must have gnds common. so you will always need gnd.  This is why the header contains rx, tx,and gnd.

[Admin]

All grounds on the axon I believe are common.

yeap, all common.

The G was a way to remind/help people to make sure grounds are common.

[Stephanie]

It worked. Thanks everyone.