I have a 6x2 30T Zumo-tracked bot with ultrasound and optical obstruction sensors. A TB6612FNG is the motor driver, with VMOT served by a 6-pack of AA NiMHs. I dropped a current sensor across VMOT for those times the bot tries to get pushy against an immovable object, wall corner or table leg, when the PING and optical sensors fail. I tried a Pololu +/-12.5A sensor, but the detection range is too great for the little SP motors. In one attempt, I set the threshold in code so low that current spikes from the motors triggered the evade subroutine. Increasingly higher thresholds failed detecting the bots stoppage. So, I am thinking of changing out the SP’s for HP motors. I can also drop down to a +/-5A current sensor (on wishlist), with its 187mV/A output. I am thinking the HP motors will have higher current demands making this a workable combination. Another thought I have is that the HP motors will allow better performance at lower duty cycles.
Can someone advise on my thinking in these matters?
(I have had little success with ramping speeds in code, to reduce current spikes, so will post separately regarding this.)
Your robot seems pretty cool. The TB6612FNG motor driver can be permanently damaged by a stalled high-power micro metal gearmotor, so I do not recommended upgrading to more powerful motors if you plan to keep using the TB6612FNG. However, if you want to upgrade your motors, you might consider replacing that motor driver with either two DRV8838s or a single DRV8835, both of which can handle the stall current of the micro metal gearmotors at 6V. (We have used both of those drivers on our Zumo robots with HP micro metal gearmotors.)
As for your current sensing issue, unfortunately I think you would still run into the same issues with the +/-5A current sensor; and we do not have any other current sensors more appropriate for measuring several hundred milliamps. I also do not have any recommendations on where to find any of those sensors.
Aloha Jon and thank you your reply –
I picked up several Pololu TB6612FNG’s as the specs seemed respectable, so let me get this straight:
TB6612FNG driver: per channel, 3A peak, 1A continuous (2A parallel)
DRV8835 driver: per channel, 1.5A peak, 1.2A continuous (2.4A parallel)
OK, I’ll go with your recommendations on the DRV8835. I have other projects for the Toshiba’s. Just to make note of something, I have never felt the Toshiba board warming up or get hot. The bot never truly stalls, just bogs a bit, as it runs on faux hardwood flooring. Carpet floors might require a different strategy.
I picked up the +/-12.5A current sensors for monitoring some 37D’s in a rope machine, so those will go where intended.
I want to take the random wandering out of my kid’s bot, add an IMU, wheel encoders, &c. Current sensing seemed a worthwhile (safety) feature to incorporate for those odd times the environmental sensors failed. To accommodate all this I got a 40-pin DIP 1284P-PU in the mail last week.
Again, thanks Jon! A hui hou.
Say Jon –
Might the MP micros be a better choice if I wanted to stay with the TB6612FNG board? I am looking at the 100:1 micro MP for encoders.
The DRV8835 and TB6612FNG drivers do have similar current capabilities, but the DRV8835 is an overall more robust driver because it has better internal shutdown features for over-current and over-temperature protection. I suspect the MP micro metal gearmotors would be fine with the TB6612FNG.
Placed the order.