Linear actuator help
Jonathan Cousineau wrote 08/16/2018 at 15:25 • 0 pointsHello,
In search of knowledge, I am using a RoboClaw powered by a switching power supply to run two Linear Actuators (ServoCity SDA 12-263) and I am running into trouble with transients. I am pretty sure that I need to make either a voltage clamp or a snubber to deal with these spikes but I don't know which is appropriate or how to calculate the required values. My google foo has failed me and I was hoping to have a kind passer by send me in the right direction.
More specifically the caps in side the motor keep exploding. My understanding is this is cause by over voltage and so it must be transients from when I change directions quickly..
Thanks in advance
Jon
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.
I think you're heading the right direction. However there are a few things I want to mention. (maybe you don't need it, but maybe other readers do)
The RoboClaw's manual starts with "Motor Selection". Your actuator can draw more than 26 Amps. Are you sure, the right driver model was selected? You didn't specify.
As agp.cooper mentioned, there can be a great amount of energy in the system when you stop driving it. Most of that will go back to the power rail, because the motor acts as a dynamo. How much energy is there, depends on the system: moving mass and its velocity. The motor's rate of conversion, the power, is similar both ways, so 26A*12V= 312 W. In practice, you probably won't need a 300W resistor (that's beefy), but still a big one.
It seems, the RoboClaw is a pretty smart driver. I suggest using its advanced features, like the current limiting function. You can also use some ramping in your software to limit the power (I guess that's the same as "appropriate braking"). Avoid rapid changes of speed, especially direction changes. Or select really high power parts that can handle it.
Are you sure? yes | no
Hello Everyone,
Thanks for your answers. Food for though to be sure and I have done more reading of both the roboclaw manual (which is a whopping 105 pages) and just motor control in general.
@Jarrett I didn't include a schematic because A. I don't have a digital one and B. I have literally just connected the roboclaw to an arduino and the motors with no peripherals or additional components so I didn't think it would help anyone.
@Tonkabot @agp.cooper Thanks for your input, I have dug deeper into the roboclaw's manual and it doesn't explicitly say how it works. I suspect it is a "fancy" H-Bridge it come with a lot of features. I did not explain before but I inherited this project and so the major hardware components were not my choice. After further digging there is an option to turn on a voltage clamp (this requires a MOSFET connected to the pin of your choosing). That being said I do think I can save myself a lot of heart ache by coding some appropriate braking into the program when switching directions.
Also if you care to follow along its page 46 of the roboclaw manual found here: http://downloads.ionmc.com/docs/roboclaw_user_manual.pdf
For now I'll order the mosfets and power resistors to make a good test of this as I dont have any resistors or fets that can handle the wattage.
Cheers
editted typo
Are you sure? yes | no
I kinda figured you didn't have a schematic :)
My point stands, in that with just words alone, it is nearly impossible for us to figure out if you have some sort of wiring issue. With a schematic (even if it's just handwritten), it does three things:
A) Allows YOU to sanity check yourself, because no one is as good at circuit-designing in their head as they think they are :)
B) Allows us to see what's going on at a glance. It could be a really simple mistake that you're overlooking, but take about three seconds for a third party to spot it and point it out.
C) Provides a base for us to mark it up and send it back, if we do spot something that is difficult to explain without words.
So basically, all of the reasons that you would post your code in a programming question still apply here.
Hope you have it figured out, though!
Are you sure? yes | no
A quick check of the manual (p10) says to use a BATTERY! That pretty well says it all.
If you have to use a SMPS then:
1) use the H-Bridge to Brake the motor, that is cycle the EMF current through the motor (as previously discussed).
AND
2) use the voltage clamp circuitry, refer to pages 38, 46 and 47
AlanX
Are you sure? yes | no
I am not an expert here but have run into similar problems in the past with servos.
Basically your motor use lots of power (>20A at 12v) so the inductive kickback is pretty significant. The inductive current has to go somewhere to be stored or absorbed.
If you are using a battery then great, a couple of high current rectifiers back to the battery will do the job. It is not really practical to store this level of power in a capacitor.
Tell us about your power supply and switching circuit.
---
As Tonkabot suggests you may be able use the "H" bridge as a "fly-back diode" and absorb the power in the motor. But you have to wait for the inductive currents to die down or you will blow something up.
AlanXAre you sure? yes | no
If you can, you should 'brake' your motor before you reverse voltage.
If your powering it with an H-bridge, that means turning off both high sides and on both low sides.
The motor will quickly stop, and should not hurt the capacitors.
Then you can reverse the voltage.
Are you sure? yes | no
Do you have a schematic?
For electronics, it's like asking a programming question without posting your code :)
Are you sure? yes | no