Page 2 of 2

Re: A-axis thermal runaway

Posted: Tue Mar 29, 2016 3:02 am
by Picky
Have you guys measured your actual coil current? Is it really as low as you expect it to be? What values are were talking about anyway?

Heat does not come from anywhere but current. Could be shorted coils in cheap motor or faulty driver current measurement. Hooking up a scope would be another thing to try and see the waveforms.

Re: A-axis thermal runaway

Posted: Tue Mar 29, 2016 4:15 am
by Picky
WayOutWest wrote:
Mark Harris wrote: As a note on this, I forgot to mention the ST L6470 - http://www.st.com/web/en/resource/techn ... 255075.pdf
Wow, no joke; thanks for pointing out that chip. Yeah, that's pretty much most of the features on my "gee why don't stepper drivers do X" list. Digital current control and automatic step-fraction-enlargement during high-speed movement for starters.
I used that driver on an eval board several years ago to drive a peristaltic pump. If I recall correctly, it had stall detection capability which I needed and was able to run a motor at a constant speed indefinitely and autonomously. Worked very well spinning a NEMA23 non-stop for a few months. Very low heat dissipation too.

Re: A-axis thermal runaway

Posted: Thu Mar 31, 2016 8:41 pm
by WayOutWest
Picky wrote: Heat does not come from anywhere but current. Could be shorted coils in cheap motor
Yes, that's my theory. Initial heating melts the insulation, causing a weak short, which causes more heat -> feedback loop.

I've killed a few solenoids and autopsy confirmed this was the failure mechanism.

Re: A-axis thermal runaway

Posted: Sat Apr 02, 2016 12:39 am
by bobc
Would be nice to play with the L6470, I've seen it mentioned before. All the low end control boards I've seen use Allegro or TI drivers. The only board I've seen supporting it is the "BAM&DICE" design, https://www.2printbeta.de/Electronics/C ... ::328.html.

The RAMPS type shield for Arduino Mega is the "BAM", plug in modules are the "DICE" I think. They also have a Trinamic driver option, which are supposedly more advanced than the ST ones.

Sparkfun have a breakout board for a single driver, it should be relatively easy to add one or two to a grbl/tinyg setup. Of course, then there is the small matter of modifying firmware to support the drivers.

I was thinking of doing a RAMPS-FD V2, maybe I will look at using L6470 or something. I also had a plan to use the small PCI connectors, they are the cheapest connector solution I could find. I was worried about mechanical stability though, it seems like they need some sort of hold down to stop them working loose.

If anyone knows of any other boards with 4 or 5 of these more advanced drivers, let me know!

Re: A-axis thermal runaway

Posted: Sat Apr 02, 2016 8:57 pm
by WayOutWest
bobc wrote:If anyone knows of any other boards with 4 or 5 of these more advanced drivers, let me know!
I'm designing a board with an L6470 and an ATTiny48 that you mount on or near the motor... string them together on an SPI+Vmot+GND chain for multiple motors.... but it's pretty far down on my priority list so it's going slowly.

Re: A-axis thermal runaway

Posted: Tue Jul 05, 2016 6:25 pm
by Danal
Hmmm.... mine is remaining cool to the touch, in a 22 or 23 degree room. I have the current dialed WAY down... just set it by eye at about one third of travel... and it works just fine like that.