Roboclaw 2x15 - logic battery vs main battery

General discussion of using Roboclaw motor controllers
ecorrales
Posts: 33
Joined: Sun May 06, 2018 5:06 am

Roboclaw 2x15 - logic battery vs main battery

Postby ecorrales » Thu May 17, 2018 3:50 am

Recently purchased a Roboclaw 2x15.

-- how do I determine what board version I have?
-- how do I know if I have safely / correctly wired it so that I can use a separate logic battery?

-- what happens if the logic battery is a lower voltage than the main battery?

thank you

User avatar
Basicmicro Support
Posts: 965
Joined: Thu Feb 26, 2015 9:45 pm

Re: Roboclaw 2x15 - logic battery vs main battery

Postby Basicmicro Support » Thu May 17, 2018 10:01 am

1. THere is a revision printed on each board(V3, V4, V5, V5D, V5E).
2. If your board has the LB/MB jumper next to the LB +- pins then you have an older board.

If you wire up a logic battery before connecting the Main Battery the board will power on(but it will error out if you try to run the motors.

The logic battery only supplies power when the main battery dips below the logic battery voltage.

ecorrales
Posts: 33
Joined: Sun May 06, 2018 5:06 am

Re: Roboclaw 2x15 - logic battery vs main battery

Postby ecorrales » Fri May 18, 2018 4:33 am

thank you. one related question to wiring..

My concern is possible ground loops. In the drawing, the yellow dashed lines show the two ground paths. One path is the grounding to all the logic batteries and logic (Pi & Claw). The other is grounding for the main battery. I tried to keep logic wiring as far apart from main/motor battery as possible.

Q: internally, is the Roboclaw logic ground pin wired to the Main Battery negative pin?
(since I have the two different ground paths to the Roboclaw.

Note: not shown are other switches to the logic

User avatar
Basicmicro Support
Posts: 965
Joined: Thu Feb 26, 2015 9:45 pm

Re: Roboclaw 2x15 - logic battery vs main battery

Postby Basicmicro Support » Fri May 18, 2018 9:27 am

On the small Roboclaws(2x60A and lower) there is a common ground plane for the whole board.

1. What is in the shielded cable? Why is it conencting back to the ground bar? I assume this is a connection stricly from the roboclaw to the RPI. If so it should only connect between the Roboclaw and RPi. Im assuming signal lines and a ground line. Alternatively the Roboclaw could power the Rpi via 5v removing the Rpi connection to the Li-Ion pack as well.

Why is the Li-Ion pack connected to tthe GroundBar? Im assuming it is th elogic battery. It should be connected directly to the Roboclaw LB+- pins and optionally the Raspberry Pi(through a 5v regulator), though I recommend powering the RPi from the Roboclaw 5v.

Also you will need a power switch on the Logic battery connection unless you want to constantly power the Roboclaw logic.

wiring this way elliminates any ground loops. Main battery only connects to ROboclaw B+/B-, logic battery only connects to LB+-. RPi only connects to Roboclaw 5v/Gnd(if powered from ROboclaw 5v) or RPi only connects to Logic Battery +- through a seperate regulator. You get a common ground between RPi and ROboclaw through the battery in the later case.

ecorrales
Posts: 33
Joined: Sun May 06, 2018 5:06 am

Re: Roboclaw 2x15 - logic battery vs main battery

Postby ecorrales » Fri May 18, 2018 8:58 pm

I think I understand... I think you're saying that the center of the universe here, with respect to ground, is the Roboclaw.

Thus connect the main battery only to MB(+/-) on the Roboclaw.

And the logic batteries to (+/-) on the Roboclaw.

And the logic batteries (+) to 5v stepdown regulator, then to the Pi (+).

The 5v regulator's (-) is common to both the logic batteries (-) and the Pi (-).

However, I have already wired everything, treating the external ground bar as the common ground point. (not the Roboclaw's ground plane as the common point; since I didn't know this). (just didn't think about it in those terms).

I think this (below) is similar to what you're saying, but still keeping most of what I have already wired:

1) main battery (+) ---> switch S1 ---> fuse F1 ---> Roboclaw (MB+)
2) main battery (-) ---> ground bar
3) ground bar ----> Roboclaw (MB-)

4) logic batteries (+) ---> switch S2 --> fuse F2 ---> Roboclaw (logic +)
5) logic batteries (-) ---> ground bar
6) Roboclaw (logic -(neg)) untouched since already part of internal ground

7) logic batteries (+) ---> switch S3 --> fuse F3 ---> 5V stepdown Regulator (Vin)
8) 5V stepdown reg (Vout) --> Raspberry Pi (+)
9) ground bar --> 5V stepdown reg (-)(gnd)
10) ground bar --> Raspberry (-)

I think that's another way of saying the same thing.

If so, that leaves one unanwered: the USB conection between the Pi and the Roboclaw. Since I'm assuming the neg(-) of the USB across them would constitute another ground path, thus a loop.

If so, should I use opto-isolation

I want to retain the external common ground bar.

And I don't know that in the future I will stay with using the USB as the communications link.

And I don't want to power the Pi from the Roboclaw. I want them separate , only having a common ground.

?

thanks again

User avatar
Basicmicro Support
Posts: 965
Joined: Thu Feb 26, 2015 9:45 pm

Re: Roboclaw 2x15 - logic battery vs main battery

Postby Basicmicro Support » Mon May 21, 2018 10:01 am

After the web address change the wiring image link was broken so I removed.

1. Yes and no. Each area of communication should have a single ground connection. In your case for all inteszt, yes the Roboclaw is the center of the universe since it is also the distribution point for power(to the motors or to the RPi).

2. Not really. You are adding asignificant lenths of wire to your grounds and you do still need a logic battery ground to the roboclaw. Relying on the MB- as the onyl ground is not recommended. The battery + and - should be as short as possible to the Robcolaw(Logic and Main) jjust liek the motor wires should be as short as possible.

Putting everything through the bus bar may be making everything neat but its increasing the size of your high current paths. Those are the main source of radiated noise.

3. Theoretically the USB ground is another loop. You could use ONLY thbe USB ground from the Roboclaw to the RPi and not connect a ground pin on the Roboclaw to a ground pin on the RPi if you are really worried about it. Id' probably wait until I see a problem before going that route though.

ChrisA
Posts: 10
Joined: Mon May 22, 2017 4:32 pm

Re: Roboclaw 2x15 - logic battery vs main battery

Postby ChrisA » Sat May 26, 2018 12:08 pm

If you are worried about "'correctness, grounds and safety, I see an error. You batteries are wired (+) -> (switch) -> (fuse). The fuse needs to the as physically and electrically close the battery is is possible. In the ideal case the battery would be inside a box and the fuse inside the same but. Tis way wire exiting the box is protected. In your case the switch is unprotected from over current and switches are a failure point.

Yo think is is "only a battery" but if it is a modern LiPo then it can vaporize even #8 wire.

Remember the purpose of a fuse is to protect the wire not the load

One problem with the bus bar is that if make it impossible to use twists pair power wires. To power large loads you real wan the positive and negative leads to be in physical contact. Separated only by the thickness of the wire insulation. Twisting them ensures this (so does using zip cord. The reason is that the current in the two wires is ecavty the same but in opposite directions. The fields cancel nearly perfect if the wires are the exact same length and in contact. Using the bas bar you open a physical loop that will radiate.

The motors are PWM'd so the high current power leads have all kinds of high frequency components on them up to many times the PWM frequency.

Loads like the Pi don't pull much current and are nearly perfect DC so they don't matter so much

ecorrales
Posts: 33
Joined: Sun May 06, 2018 5:06 am

Re: Roboclaw 2x15 - logic battery vs main battery

Postby ecorrales » Tue May 29, 2018 7:38 am

Hi, thanks for the tips.

yes, have the main fuse coming right out of the main battery (no Li-Po; it's lead-acid).

No, dont have twisted pair, but have shielding. some I wrapped myself, some was bought. shielding is grounded to chassis. Also an over abundance of ferrite chokes. Also have a capacitor across both motor terminals. I have done some scope waveform viewing - the the spikes on the pulse edges seem really small even at the Roboclaw motor outputs .. the supply rails seem pretty quiet. The logic (Raspberry, etc) is pretty physically separated from high current cables, etc. The Raspberry and Roboclaw logic is using separate voltage sources from the main battery.

Been running the robot over the weekend under different speeds, conditions, etc... haven't had any issues.

User avatar
Basicmicro Support
Posts: 965
Joined: Thu Feb 26, 2015 9:45 pm

Re: Roboclaw 2x15 - logic battery vs main battery

Postby Basicmicro Support » Tue May 29, 2018 9:46 am

Good catch, ChrisA.

Also, lead acids can supply crazy currents as well so ChrisA's statement applies to them as well.

ecorrales
Posts: 33
Joined: Sun May 06, 2018 5:06 am

Re: Roboclaw 2x15 - logic battery vs main battery

Postby ecorrales » Tue May 29, 2018 9:53 am

oh.. right.. (I got lucky) lol... wired the fuse in first next to (+) terminal, then some wire down to switch (switch at convenient place).


Return to “General Discussion”

Who is online

Users browsing this forum: No registered users and 2 guests

cron