Author Topic: Brushed DC motor: Resistor in series with freewheeling diode  (Read 4290 times)

0 Members and 1 Guest are viewing this topic.

Online Nominal AnimalTopic starter

  • Super Contributor
  • ***
  • Posts: 6260
  • Country: fi
    • My home page and email address
As I understand it, if one wanted to reduce the braking effect when power to an unidirectional brushed DC motor is cut off and a freewheeling diode is used, one could add a power resistor in series with the freewheeling diode, to bleed off the energy when the motor is acting as a generator.

Omitting the freewheeling diode means there is no braking effect at all, but then there is no clamping on the (negative) voltages generated either.

Similarly, on a stepper motor, one could have additional MOSFETs shorting each winding through a power resistor, as a sort of a "half-brake", when needed.  This might be useful when the motor is connected to a hand crank, as a "tactile" feedback for example.  (Short-circuiting the windings directly is usually a bit too strong, at least for the steppers I have.)

Have you ever seen this used in real life?  Aside from voltage and power dissipation requirements, are there any pitfalls here?

It's a silly question, I know; sorry.  :)
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 8172
  • Country: fi
Re: Brushed DC motor: Resistor in series with freewheeling diode
« Reply #1 on: August 21, 2021, 09:33:02 am »
You have completely misunderstood.

Freewheeling diode does not brake. It creates forward torque for the small time during every PWM cycle, by allowing the current to flow in the same direction it was already flowing when the MOSFET was turned off, utilizing the energy stored in the inductance of the motor to drive the motor as intended instead of generating a massive voltage spike.

Braking happens when you actively drive the second switch you don't have (as you use a diode, which is automatic, and never turns on to enable braking), boosting the motor BEMF voltage, and have a load on the input side which can consume this generated current, like a battery.

Braking also happens when the motor is turning faster than the input voltage would be capable of turning it. Then BEMF > Vin,  and current flows through the freewheeling diode, rising the input voltage. In order to current to flow, you need a load (think about a zener diode or active equivalent) that turns on during input high voltage condition; or a battery which would get charged.

Because information about motor control online is crap and mostly bullshit, this motor circuit is best analysed as a buck converter, motor being both the inductor, and the (very large) output capacitor which voltage is ~ RPM. This way you get the correct understanding.
« Last Edit: August 21, 2021, 09:37:44 am by Siwastaja »
 
The following users thanked this post: Nominal Animal

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21681
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Brushed DC motor: Resistor in series with freewheeling diode
« Reply #2 on: August 21, 2021, 09:36:00 am »
bEMF is in the same direction as applied voltage, it looks like a capacitor (namely, inertia), not an inductor (the winding inductance appears in series with that, however).  So yeah, no problem.  Probable "duh" moment, that's okay, I think I wondered the same thing years ago too. :)

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: Nominal Animal

Online Nominal AnimalTopic starter

  • Super Contributor
  • ***
  • Posts: 6260
  • Country: fi
    • My home page and email address
Re: Brushed DC motor: Resistor in series with freewheeling diode
« Reply #3 on: August 21, 2021, 04:51:04 pm »
Dammit, I'm an idiot. :palm:

Why didn't I check this first?

Oh well, not the first time I expose my true stupidity to the world.... :-DD
 

Offline langwadt

  • Super Contributor
  • ***
  • Posts: 4425
  • Country: dk
Re: Brushed DC motor: Resistor in series with freewheeling diode
« Reply #4 on: August 21, 2021, 05:01:12 pm »
Because information about motor control online is crap and mostly bullshit, this motor circuit is best analysed as a buck converter, motor being both the inductor, and the (very large) output capacitor which voltage is ~ RPM. This way you get the correct understanding.

and a (sync)buck converter in reverse is a boost converter, so depending on how it is switching the motor voltage can be boosted
and backfeed into the supply
 
The following users thanked this post: Nominal Animal

Online Nominal AnimalTopic starter

  • Super Contributor
  • ***
  • Posts: 6260
  • Country: fi
    • My home page and email address
Re: Brushed DC motor: Resistor in series with freewheeling diode
« Reply #5 on: August 21, 2021, 07:09:42 pm »
Because information about motor control online is crap and mostly bullshit, this motor circuit is best analysed as a buck converter, motor being both the inductor, and the (very large) output capacitor which voltage is ~ RPM. This way you get the correct understanding.
and a (sync)buck converter in reverse is a boost converter, so depending on how it is switching the motor voltage can be boosted
and backfeed into the supply
Right; good points, especially when considering the failure modes (say burning a motor controller IC due to high voltage spikes from a temporary power loss to the controller while the motor is still rotating and such).

I only have hobbyist experience with brushed DC motors in small toys and such; with steppers, using drivers or driver break-out boards (Allegro A4988, TI DRV8825, Trinamic TMC2208).  I ought to know the theory, did the courses at uni.. but this is one of those things that you need both the correct theoretical understanding (at least the correct approach/paradigm to base ones understanding on), and practical experience, before you really grok the real-world behaviour.  That also explains why so much of the online information on this is bad.

Most of the break-out boards (A4988, DRV8825, TMC2208) do not have freewheeling diodes built-in, which is a bit annoying.
Then again, most appliances use a shared ground for both logic and motor power, which is even more annoying.
« Last Edit: August 21, 2021, 07:12:37 pm by Nominal Animal »
 

Offline langwadt

  • Super Contributor
  • ***
  • Posts: 4425
  • Country: dk
Re: Brushed DC motor: Resistor in series with freewheeling diode
« Reply #6 on: August 21, 2021, 07:50:30 pm »
Because information about motor control online is crap and mostly bullshit, this motor circuit is best analysed as a buck converter, motor being both the inductor, and the (very large) output capacitor which voltage is ~ RPM. This way you get the correct understanding.
and a (sync)buck converter in reverse is a boost converter, so depending on how it is switching the motor voltage can be boosted
and backfeed into the supply
Right; good points, especially when considering the failure modes (say burning a motor controller IC due to high voltage spikes from a temporary power loss to the controller while the motor is still rotating and such).

I only have hobbyist experience with brushed DC motors in small toys and such; with steppers, using drivers or driver break-out boards (Allegro A4988, TI DRV8825, Trinamic TMC2208).  I ought to know the theory, did the courses at uni.. but this is one of those things that you need both the correct theoretical understanding (at least the correct approach/paradigm to base ones understanding on), and practical experience, before you really grok the real-world behaviour.  That also explains why so much of the online information on this is bad.

Most of the break-out boards (A4988, DRV8825, TMC2208) do not have freewheeling diodes built-in, which is a bit annoying.
Then again, most appliances use a shared ground for both logic and motor power, which is even more annoying.

 diodes are intrinsic to the output FETs and usually bypassed by turning on the FETs to reduce the power dissipation

 

Online Nominal AnimalTopic starter

  • Super Contributor
  • ***
  • Posts: 6260
  • Country: fi
    • My home page and email address
Re: Brushed DC motor: Resistor in series with freewheeling diode
« Reply #7 on: August 21, 2021, 08:37:55 pm »
Are the intrinsic diodes are in the correct orientation, though, considering bipolar operation?  Using Schottky diodes with A4988 and DRV8825 does make an observable difference, you see.  For example, Stepstick Protector add-ons.
 

Offline langwadt

  • Super Contributor
  • ***
  • Posts: 4425
  • Country: dk
Re: Brushed DC motor: Resistor in series with freewheeling diode
« Reply #8 on: August 21, 2021, 11:45:38 pm »
Are the intrinsic diodes are in the correct orientation, though, considering bipolar operation?  Using Schottky diodes with A4988 and DRV8825 does make an observable difference, you see.  For example, Stepstick Protector add-ons.

yes the intrinsic diodes are in right direction (or they would short the supply ;))

only way I see the extra diodes doing anything is if the FETs on resistance is high enough that at high current and temperature the voltage is higher than the Schottky diode voltage
 

Online Nominal AnimalTopic starter

  • Super Contributor
  • ***
  • Posts: 6260
  • Country: fi
    • My home page and email address
Re: Brushed DC motor: Resistor in series with freewheeling diode
« Reply #9 on: August 22, 2021, 01:05:35 am »
Right.  I've used MOSFETs as switches, and played with ideal diode models, so I should have realized.

I took the time to look at a few circuit diagrams of full step stepper drivers, and work the conventional current flow in each phase, to make sure I understand the situation. I really should have looked at the functional block diagram in the A4988 datasheet first.  The datasheet even mentions that at 1.5A current, the body diode voltage drop is max. 1.2V.

So, I definitely now think it's the lower voltage drop compared to the intrinsic diodes, and the reduced thermal stress, that explains why these stepsticks tend to fare better (fewer issues and breakdowns) with those extra Schottky diodes.  The datasheet mentions the package thermal resistance is 32°C/W on a four-layer PCB (these ICs are cooled via the bottom, with vias), with PD=4.0W at 20°C ambient temperature; they cannot really dissipate much power at all.
 

Offline langwadt

  • Super Contributor
  • ***
  • Posts: 4425
  • Country: dk
Re: Brushed DC motor: Resistor in series with freewheeling diode
« Reply #10 on: August 22, 2021, 01:22:42 am »
Right.  I've used MOSFETs as switches, and played with ideal diode models, so I should have realized.

I took the time to look at a few circuit diagrams of full step stepper drivers, and work the conventional current flow in each phase, to make sure I understand the situation. I really should have looked at the functional block diagram in the A4988 datasheet first.  The datasheet even mentions that at 1.5A current, the body diode voltage drop is max. 1.2V.

So, I definitely now think it's the lower voltage drop compared to the intrinsic diodes, and the reduced thermal stress, that explains why these stepsticks tend to fare better (fewer issues and breakdowns) with those extra Schottky diodes.  The datasheet mentions the package thermal resistance is 32°C/W on a four-layer PCB (these ICs are cooled via the bottom, with vias), with PD=4.0W at 20°C ambient temperature; they cannot really dissipate much power at all.

I exepect the chip to turn on the FETs so the instrinsic diodes are only during the transition, so the voltage will beRdson * I 

 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21681
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Brushed DC motor: Resistor in series with freewheeling diode
« Reply #11 on: August 22, 2021, 05:16:53 am »
The winding inductance is often used as the filter choke in a buck converter; depending on how the controller deals with this, it could be that it's causing body diode conduction from time to time, in which case schottky diodes could help.  This can be prevented by working in forced continuous current mode (one transistor always on).

I haven't checked if these controllers typically use this mode or what.  I do know older stepper drivers for example, tend to use a single switch, and catch diode, to implement the buck control as a hysteresis current limiter (e.g. L297+L298) -- as a result, the winding is simply open circuit when it's not driven, so there's some advantage to be had with schottky.  (But also such old drivers as L298 have relatively huge voltage drops too, making schottkys a pretty obvious win. :P )

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline langwadt

  • Super Contributor
  • ***
  • Posts: 4425
  • Country: dk
Re: Brushed DC motor: Resistor in series with freewheeling diode
« Reply #12 on: August 22, 2021, 12:42:34 pm »
The winding inductance is often used as the filter choke in a buck converter; depending on how the controller deals with this, it could be that it's causing body diode conduction from time to time, in which case schottky diodes could help.  This can be prevented by working in forced continuous current mode (one transistor always on).

from the datasheet they either reverse the output of the hbridge or short the coil using the lower transistors, for either fast or slow decay


I haven't checked if these controllers typically use this mode or what.  I do know older stepper drivers for example, tend to use a single switch, and catch diode, to implement the buck control as a hysteresis current limiter (e.g. L297+L298) -- as a result, the winding is simply open circuit when it's not driven, so there's some advantage to be had with schottky.  (But also such old drivers as L298 have relatively huge voltage drops too, making schottkys a pretty obvious win. :P )

afaikt from the datasheet they are bipolars so no diodes, and in both cases even in forced conduction there will be some dead time where the diodes will conduct

 

Offline Ernest1

  • Newbie
  • Posts: 9
  • Country: ca
Re: Brushed DC motor: Resistor in series with freewheeling diode
« Reply #13 on: December 06, 2023, 09:48:01 pm »
I have a  drill here that has a diode in parallel with motor through a switch on the trigger. When the motor trigger is not pressed the switch is closed with the diode. when the trigger is pressed the diode is disconnected and the motor is being driven by pwm. When you let go of the trigger, the switch is connected while the motor is running and the diode motor brakes.

The original diode was shorted (fx2000g) or equivalent (DST2045AX). I placed a much smaller 1n5404 instead because that's the biggest I have and it stops the motor instantly but I'm not sure about longevity. Would it help to add a big resistor (wattage) in series with it to reduce the stress on the diode while keeping some braking?
« Last Edit: December 06, 2023, 09:50:42 pm by Ernest1 »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf