Pololu a4983 scared me...

Hi everybody,

I'm a bit affraid that I have burn 2 of my 4 pololu, here was my process:

-I build a generation7 electronic, it seems to work fine.
-I plug my first pololu driver on a stepper motor 8.5V NEMA23, yaaaaheee, it worked!
-I do have some 4,5V stepper driver, so, before get my stepper motor working, I wanted to watch a bit closer the output signal with my oscilloscope, so I plugged it with nothing else...
By the way: The resistor of my oscilloscope is 4MOhms, I verrified...
Starting to this point, the driver went in strike..

first, I through it was a firmware (teacup) problem, so.... arrggll... I plugged the second one, and, same problem happened.

So in the result I have 2 working pololu and two not working (I believe)...

Nothing burned, no "plop noise" things, can it be something else?

I dont know if it migth help but I saw a difference between both groups (working/not), I dont know if it's a coïncidence but its is not the same chips (working:A4983ET 1049 453M, not: 4983ET 1102 718L). In fact I dont really want to try the same thing over the working group.

Somebody have an Idea of what happened?

THanks,

Ghunt

Edited 3 time(s). Last edit at 04/09/2011 08:57AM by Ghunt.

i am curious about this too.

i wish i knew of a way to test them without a stepper on them. from what i have read they need the stepper on them all the time.

---

[mike-mack.blogspot.com]

Yes but it's a bit strange to need some output filled...

I still hope a solution, perhaps they are not dead even if they seems to be, like a protections or something...
I cant find anything in datasheet. If this problem was known, we would see some : CAUTION: always plug a stepper motor on your driver! ... But it's not the case...


Thanks for your reply anyway,

If somebody have nice ideas, I take it

Ghunt

They won't be damaged by not having a stepper connected but they can be by attaching something while the power is on.

If the output of the chip gets an ESD or BEMF transient it latches up and tries to draw a lot of current from the 5V rail. If that is connected directly to a PC PSU then the current is only limited to about 20A, so the chip fries.

If you drive the 5V supply from a 78L05 100ma regulator then that seems to protect them from being destroyed in this way.

---

[www.hydraraptor.blogspot.com]

nophead Wrote:
-------------------------------------------------------
> They won't be damaged by not having a stepper
> connected but they can be by attaching something
> while the power is on.

No, I didn't act like a pig... just plugged my oscilloscope.
Perhaps, Something about an EMP or a ground problem...

Anyway, I tried several things, nothing seems to make my pololu work again, so, apparently it's fried even if nothing burn...

I must say that I am a bit disapointed about the robustness of the chips ...

Well I will make my own pololu I believe, I cant order in france but some similar allegro chip can be found for some euros.

Thanks for your answers

Hugo

Edited 1 time(s). Last edit at 04/09/2011 07:09PM by Ghunt.

You can also protect them by adding 8 Schottky diodes. The don't need them when they are running because they have synchronous rectification but the lack of them is what makes them vulnerable to transients.

---

[www.hydraraptor.blogspot.com]

i feel like a bit of an idiot for having to ask this nophead but how would you wire them up for that.? is there a circuit that i could see.?

---

[mike-mack.blogspot.com]

On each of the four outputs you connect a Schottky diode to +12v (cathode) and a diode to ground (anode). That prevents the output going above 12V or below zero.

Edited 1 time(s). Last edit at 04/09/2011 07:52PM by nophead.

---

[www.hydraraptor.blogspot.com]

One possible mode of failure: If there's no inductive load, the current would spike quite quickly, and may defeat the overcurrent protection. I.E. the overcurrent mechanism might not be able to respond fast enough to prevent damage to the chip.

Of course, that can't be the case here, because the oscilloscope has 4MOhm resistors, which would limit the current to a safe level anyway... In fact, I think that failure case (if it's possible at all) would pretty much require that there be no load at all - i.e. a dead short.

So if this failed, what IS the proper way to hook up an oscilloscope to the steppers? Insert a very small resistor in one side of each coil, and put the leads on either side of that resistor?

---

--
I'm building it with Baling Wire

Not really sure, but i think "proper" way would be to have it on the sense resistors.