[daedalus]

Hi,

I have previously built systems with drives that operate in 4 quadrant mode, where they dump power into the power supply during braking. This is fine with linear supplies, but i am considering a new build that needs to be portable, and the weight of linear power supply is a problem.

I can see a few problems with using switch mode supplies with my application, and i wanted to get some ideas of ways around them. Firstly i will need to dump power into the power supply, so are there switch mode psu designs that allow this? Alternatively does anyone know if i could somehow add a seperate dump resistor to the design?

Are there going to be problems with the voltage regulation tracking output voltage if i start increasing the output voltage due to back-emf?

My application requires the drives to be operated in torque control mode, and the motors will run with 0 torque for the most part, then switch to full torque instantly. I am concerned that a switch mode power supply would not be capable of providing sufficient instantaneous current, is this likely to be an issue? I assume i can add extra external capacitance to help with this?

[mcpltd]

Most of the drivers i have used have an external dump resistor, but we also used a linear power supply.
It will come down to the powersupply you use, but most will have over voltage protection and when it sees the voltage going up, it will throtle back.
Also if you were to use a switch mode supply it will need to be non regulated.
What driver are you using and what current and voltage is the supply?

Hi,

I have previously built systems with drives that operate in 4 quadrant mode, where they dump power into the power supply during braking. This is fine with linear supplies, but i am considering a new build that needs to be portable, and the weight of linear power supply is a problem.

I can see a few problems with using switch mode supplies with my application, and i wanted to get some ideas of ways around them. Firstly i will need to dump power into the power supply, so are there switch mode psu designs that allow this? Alternatively does anyone know if i could somehow add a seperate dump resistor to the design?

Are there going to be problems with the voltage regulation tracking output voltage if i start increasing the output voltage due to back-emf?

My application requires the drives to be operated in torque control mode, and the motors will run with 0 torque for the most part, then switch to full torque instantly. I am concerned that a switch mode power supply would not be capable of providing sufficient instantaneous current, is this likely to be an issue? I assume i can add extra external capacitance to help with this?

[daedalus]

Im using two maxon dc brush pwm motor drives, rated at 10A peak, 55v max at the moment on a 20a 50v linear supply. For the next build im considering moving to push-pull linear amplifiers, such as the ones quanser sells. The reason for the change is that we use low inductance motors, and have problems with motor heating when using pwm. The isssue with dumping power is that the robot is moved by a human, causing the motors to be back driven, and with the linear power supply unplugged (not able to sink power from the drives), you can really feel the difference.

[daedalus]

actually i thought i recognised your web address, was looking at your switching power supplies a few weeks back. Are the outputs of the supplies floating, or is the -ve tied to ground? Are they suitable to be used in series / parallel for greater output current/voltage?

[pastera]

If the power supply is no load stable, you can isolate it with a diode, add a capacitor bank for pulse loading and a resistive load dump set a few volts over nominal.

As far as going linear, have you tried filtering your drive output?
A capacitor in parallel with the motor (tuned to PWM frequency) could reduce the heating quite a bit. The capacitor would need to be AC rated and withstand a high ripple current.

Aaron

[daedalus]

pastera, do you know anyone who sells external resistive load dump circuits? or where i might find the design for one?

I havent tried adding capacitance to the drives output, but have experimented with adding external inductance without much luck. I was planning on buying one of the linear drives to do some testing with, and see if the improvement was worth the increase in cost. I might fish out a large cap and do some testing.

[Al_The_Man]

I believe A-M-C sell external resistive dump units.
You should be able to use a simple 311 comparator to switch a HexFet into a resistor at a given level, I have designed 240v field regulators for DC generators that could easily be adapted.
Al.

[daedalus]

thanks for the help everyone, and cheers for the link Al!

[pastera]

If you use a LM311, make sure you add a few volts of hysteresis to the design so it does not oscillate.

Depending on the power level, a couple of TO-220 BJTs run linear may be cheaper than power resistors.

Aaron

[mcpltd]

They are floating and yes, you can series or parallel them to get higher voltage or current.
Also the LCAM driver is 7A continuous and 10A peak, so our SP487 could be a good choice (48V max, 7A Cont and 9A peak)
Datasheet attached.
Oh and they are only £40.83. ;o)

actually i thought i recognised your web address, was looking at your switching power supplies a few weeks back. Are the outputs of the supplies floating, or is the -ve tied to ground? Are they suitable to be used in series / parallel for greater output current/voltage?