LitePlacer.com forums

I just happened to move the x axis a bit faster while the TinyG was NOT powered up.

To my surprise the blue power LED lit up during this movement

The steppers do create enough voltage that you can drive a connect second stepper as shown in the liteplacer unpacking video of that Australian electronics guru.

The question is can the TinyG be damanged by such movements?

The TinyG uses DRV8818 stepper driver IC by Texas Instruments. I quickly looked at the datasheet and found that it features all sorts of protection circuitry - but could not figure out if these also protect the chip from overvoltages induced by manually spinned steppers.

I have moved the gantry around quite a bit during construction (with TinyG already connected) and noticed the lights flickering, but as far as I can tell no damage has been done by that so far.

mrandt wrote: but could not figure out if these also protect the chip from overvoltages induced by manually spinned steppers.

Boy I sure hope not, because the first thing I did after hooking up the TinyG was shove the head back and forth while gawking at das blinkenlights....

I could not find the link again, but it is safe. Manually moving a cnc machine is very common use case, it would be silly if the chip would die from that.

Well I looked at the Shapeoko forum and found similar reports:

http://www.shapeoko.com/forum/viewtopic.php?f=7&t=6324

http://www.shapeoko.com/forum/viewtopic.php?f=37&t=6392

http://www.shapeoko.com/forum/viewtopic.php?f=4&t=2471

But there is no real comment on "is it safe?".

Looking at the FET circuit of the DRV8818 stepper driver data sheet I suppose the inducted voltage is fed into the 24 V motor supply via the FET protection diodes and is therefore rectified and is then regulated to 3.3V and 12V.

What could be an issues is the brown out behavior/protection of the ATMEL processor chip when the 3.3 power is "flickering".

Conclusion: I for my part will try to prevent fast moving the x and y axis .

mawa wrote: What could be an issues is the brown out behavior/protection of the ATMEL processor chip when the 3.3 power is "flickering".

Don't worry about the microcontroller. Atmel's AVR chips are made on a nowadays-ancient 0.35µm process with a ridiculously thick gate oxide. They're pretty hard to fry, all the I/Os have ESD clamps, and Atmel has excellent I/O pad designers. I used to work with their FPGAs (which could be reconfigured while active at cell granularity by the on-chip AVR, such a cool feature) and those things took a ridiculous amount of abuse. I never managed to break one.

Regarding brownout specifically, you can't damage a CMOS chip that way, at least not a 0.35µm device. In a static-CMOS topology the PFET threshold is relative to the supply rail, so it droops as the supply droops -- as long as you don't have voltage gradients internal to the chip you won't be able to get both the PFET+NFET conducting meaningful currents simultaneously (i.e. shorted) by simply monkeying with the supply voltage.

WayOutWest wrote:
Don't worry about the microcontroller.

I don't worry about destroying the AVRs ports but loosing its flash firmware content and EEPROM values.

Before I setup a suitable brownout level in one of my controller boards I once sold a couple of years ago, I had boards returned from my customers with corrupted firmware images. At that time brown out detection was disabled.

I found out that the guys had put power on the board by manually connecting the power supply wires while power was turned on.

There are many statements concerning blown-up TinyGs (even in Juhas documentation) that you should never change wires while the TinyG is powered.