Double brushed motor controller, current sensing and DC link capacitance

[Kryptychon]

Hi,

this post is a follow up of my previous post in which I have asked about possibilities to increase current carrying capabilities of PCB traces (https://www.eevblog.com/forum/projects/options-for-increasing-the-current-carrying-capabilities-of-pcb-traces/).

I'm trying to design a dual brushed motor driver. My personal design goals are:

- 30V (8S LiFePO4 battery) and 30A (maybe 40A) per channel
- TTL compatible inputs, simple control scheme: one input for direction, another for PWM per channel
- robustness first
  - generously dimensioned components
  - generous thermal design
  - input protection
  - reliable protection circuitry (current & temperature), independent from MCU
- easy to understand and rebuild
  - All components are THT except LT1999 on current monitor module
  - no dense integration (-> modular design)
  - components easily obtainable
  - 4 layer stackup (Power+Signal/GND/GND/GND+Power+Signal), 35u/1oz; the inner GND layers should act as heat sinks for the adjacent high current traces
- modular design
  - current and temperature monitoring on daughterboards
  - PSU / Logic and MCU in seperate modules

I finally got to the point where I think I can share the preliminary results with you and ask for some more advice.

In the previous post I got a lot of real good advice and I tried to incorporate as much as possible.
Especially I tried to to avoid thermal issues with the high current traces.

I'm kindly asking for a (brief) review of what I did so far. I've the feeling I'm ready for revision 1
of the h-bridge and want to send the gerber files to production but before doing so I would like to
make sure I didn't mess up big time.

I feel very insecure about the current sensing circuit. I doubt, that my aproach
to make the gain adjustable like that is the way to go, but I do not know better.
I'm also not sure if the LM358 is the apamp to go for the gain adjustment.

Please find attached the KiCad projects, the circuit diagrams, the layout as well as some images.

Thank you very much! I really appreciate any feedback!


Kryp

[Picuino]

You could add a back-emf sensor with a simple resistive divider. It can be useful to control the motor speed and the maximum current to be delivered.

Current sensing can be done with a dedicated differential amplifier.
https://www.analog.com/media/en/training-seminars/tutorials/mt-068.pdf
https://www.analog.com/en/parametricsearch/11082#/
https://www.analog.com/media/en/technical-documentation/data-sheets/AD8218.pdf

Edit:
If you are going to do regenerative braking, keep in mind that the current can flow in both directions and you will need a resistor to dissipate the braking energy of the motor.

Using only THT components today is a bit restrictive.

[Kryptychon]

Hi,

thanks for your feedback.

Interesting document, you linked!
I read LT app note 105 (https://www.analog.com/media/en/technical-documentation/app-notes/an-105fa.pdf) and chose the LT1999 as differiential amplifier.
Based on the LT1999 I designed a simple current sensing and overcurrent protection circuit.

I'm just wondering, if my approach to make the gain adjustable is viable and more important: correct and reliable.

Another question I have is: Where to put the DC link capacitance. Before or after the shunt resistors?
I guess adding some electrolytic caps wouldn't hurt either...

What do you think?

Thanks!

Kryp

[Picuino]

I would put the current sense resistor between the capacitor and the bridge. Or even between the bridge and the motor (floating).

* Current Sensing in an H-Bridge (Rev. B) - Texas Instruments: https://www.ti.com/lit/an/sboa174b/sboa174b.pdf?ts=1652548586932
* DRV8876 H-Bridge Motor Driver With Integrated Current sense: https://www.ti.com/lit/ds/symlink/drv8876.pdf?ts=1652557315508
* Improving feedback current accuracy when using H-Bridges for closed loop motor control: https://www.nxp.com/docs/en/application-note/AN5212.pdf
* Current Sensing in motion control applications: https://www.st.com/resource/en/application_note/an5397-current-sensing-in-motion-control-applications-stmicroelectronics.pdf

Regarding the amplifier with variable gain, I think it is better to buy one that has it integrated.
https://www.microchip.com/en-us/products/amplifiers-and-linear-ics/programmable-gain-amplifier-ics
https://www.ti.com/amplifier-circuit/pga-vga/products.html