What blows up Pololus on RAMPS

I have recently blown up a bunch of Pololu carriers (4988 chip) and was wondering if anybody knew what causes this. In each case a hole has been blown in the top of the chip - th whole sorry saga in the General thread under 24V woes, but it seems that it is not restricted to 24V. I have seen some mention of bad motor connections, but for 4 motors all at once?

The only warnings I know about are that you should not connect or disconnect motors while under power, and do not remove from or install into a socket that is powered.

These don't sound like your issue, though.

Got any pictures?

It's a good question. The paradox is that replaceable steppers often blow up, but they are replaceable. OTOH, driver chips on the motherboard rarely seem to blow, so it doesn't matter so much if they are not replaceable.

The "pololu" type drivers have several major design flaws, which is excused because they were not really designed for the use which we put them to. An additional problem is that depending on the source, some of them are badly made or inferior clones of the original design (although a few are improvements on the original).

I don't think that faults in the motor wiring should cause the driver to fail, they have various protections for that side. Also, it seems unlikely that a fault in one motor would propagate to other chips, unless perhaps Vmotor gets shorted internally to Vlogic.

In this case, it seems that Vmotor went somewhere it shouldn't. Apart from poor build quality, problems I have noticed with pololus is that copper traces can be exposed on the ends of the PCB, and when jammed into adjacent sockets could short. The pinout means that you get adjacent pins: VLogic GND VMotor GND, so it seems like it would be easy to cross a positive rail with ground.

If the Arduino stops working, that is often the voltage regulator getting zapped.

I admit I have never broken a driver chip of any sort (yet!), so don't have first hand evidence to go on.

I would like to design a better driver carrier (Allegro chips seems really cheap at the moment), but now the pololus are everywhere it might be hard to get any take-up.

---

What is Open Source?
What is Open Source Hardware?
Open Source in a nutshell: the Four Freedoms
CC BY-NC is not an Open Source license

Things not to do (applies to most boards):
- Unplug/plug stepper motors while RAMPS is powered on (instant current draw can blow drivers, despite current limiters).
- Unplug/plug stepper drivers while RAMPS is powered on (instant current draw can blow drivers, despite current limiters).
- Unplug/plug hot end/heated bed while RAMPS is powered on (can lead to arcing in connectors if the output gets enabled somehow).
- Short out any coil on the stepper motors (have seen this blow A4988's, despite the fact they are supposed to be protected against this sort of thing).

Things to note (somewhat RAMPS specific, but also applicable to other boards):
- Beware when installing stepper drivers in a socket that you get them the right way around, and that the board is in the right place (eg: 1-off pin errors can kill a stepper driver and possibly your Arduino).
- The diodes on RAMPS for reverse-polarity protection are woefully inadequate. It's possible if you connect the PSU backwards, even only once, you'll blow all your stepper drivers and/or your Arduino.
- The PTC fuses on RAMPS are in many cases useless (>20 secs to actually trip in over-current conditions), so you really want to put fuses in your input cabling (between the PSU and the RAMPS board). This is especially important if you're using >12V, particularly with the heated bed. The PTC for the heated bed is only rated to 16V.
- I would recommend heatsinks on all stepper driver chips and a small fan always pointing at the RAMPS board (active cooling). While the heatsinks don't do a HUGE amount extra, it's just enough overhead that can in some cases make a difference, especially if the board is in a warm environment.

General notes:
- Be very careful when measuring voltages and the like on the board while power is connected. It's easy to slip (especially if the RAMPS board isn't mounted on something) and short something out. Also make sure you're on the correct range first - Current ranges are effectively a short circuit between the probes, and Resistance provides a voltage out of the probes. This can cause all sorts of problems if you select the wrong range and put the probes across something.
- Check the resistance of your steppers, to make sure that they match the specs (eg: check that the coil isn't significantly lower resistance or short circuit) before plugging them into a board the first time. It's easy for them to get damaged in transit or even missed by the QA process the manufacturer uses.
- By powered on, I mean ALL POWER. This means the PSU should not be connected or plugged in, and that any lights on the PSU, RAMPS board or Arduino should be off for more than 3 seconds. This includes unplugging the USB, as some parts of the board are powered over USB. In short: Unplug all sources of power before doing ANYTHING with motors or wiring. While it may seem a little over the top, you'll never encounter a problem by following it.

This is not an exhaustive list of things people should beware of, but it's a start.

Picture as requested

Big crater is on the extruder Pololu, little craters on the others. The Pololus for X, Y & Z are Black Series and were set to 0.4V (1A) on the trim pot, extruder is only NEMA14 and was set to 0.3V (0.75A)

I have taken most of the precautions mentioned - sometimes to the extent of near paranoia. One thing that comes to mind is that the Arduino is powered from 12V through the DC connector. This 12V is derived from a 7812 Regulator drawing power from the 24V supply and also poweres two small fans. I only mention this as the Arduino 5V supply will only come up some tens of milliseconds after the 24V supply is established.

Etiology:
Everything had been running smoothly and I had just done a "Home All" which had completed. Several seconds after this the power appeared to go off. As there was no measurable 24V, 12V or 5V I suspected the PSU; but after removal it was found to be O.K.. On reinstalling it became apparent that the PSU was folding back due to a short circuit and investigation traced this to the Pololu for the extruder. The suspect Pololu was removed and the unit turned on. LCD, SD and XYZ motors checked O.K. Performed a "Home all" and then shut down. Put in a new Pololu for the extruder connected extruder motor and switched on.

BANG, flash and white smoke.

There was no command to use the extruder motor, so there is a suspicion of something hokey with the RAMPS board. I have thoroughly checked the wiring and it looks good and does not come anywhere near any high voltages or plausible short circuits.

Other possible thoughts: There are no snubbers on the steppers, but even if there was an inductive spike it seems unlikely that it would do anything dramatic. The PSU is a 24V 350W unit intended for LED displays (similar to [www.ebay.co.uk] ) it seems very good, but could it be tempted to put out a high voltage pulse on a load other than what it was designed for? (grasping at straws department) Metal screws between Arduino and RAMPS? Yes, but worst case is they short a ground plane to a ground plane.

Any input gratefully received.

Mike

The top ground plane and the bottom ground plane may not both be ground. One may be ground and one may be +24V (or something similar).

Could the Regulator on the Arduino have failed? Putting 12V on the 5V rail would most likely kill a Pololu. Having them all go to me would mean that something happened to something common to them all, which either means Vmot (which is from the standard input) or Vcc (from the Arduino).

Since you're powering the Arduino from the 12V barrel jack, have you removed the Diode D1 on the board that powers the Arduino? It's under the X and Y stepper boards. If you don't, the Arduino will be getting 24V, regardless of your 12V regulator. There is only a small diode on the Arduino board between the DC barrel jack and AM-VIN, that is to stop you getting the DC jack back-to-front. This will stop 24V coming back out your DC barrel jack but the Arduino will still get 24V (from AM-VIN).

My printer also seems to be eating Pololus's though not dramatically with sound effects and smoke. They just seem to die gradually. The symptoms will be that the axis they control will start to act in an irratic fashion (e.g. skipping for example). I will adjust the trim pot and things will work for a bit longer. Ultimately no amount of adjusting will cause the thing to come back to life. I have not been able to pin this down to a single axis (or the extruder) as I have moved Pololus' around. I have not always unplugged the USB when doing so...I wonder if that could be the problem. I have also been using a metal jewlers screwdriver...something else I read that is bad.

I have obeyed the other rules from this thread! Any other ideas would be appreciated.

Wisar

Hi Wisar,
If I was allowed one guess, I would look at static damage. This is not common with modern components, but it can happen and the symptoms are often similar to what you described. The only real precautions that I take these days is to wear cotton - no wool, silk, nylon etc..

Hi Cefiar,
I have just been looking over the corpses. The Arduino is dead, but it's 5V regulator is working. The LCD, SD module and fitted SD card are dead - even with a good SD card it can see the directories, but not see any files in them. Diode D1 was not fitted on the RAMPS board. Other power supplies, 24V and 12V are good. All evidence points to the loom with the hot end heater/extruder/thermistor/heatsink fan, but even with 100V test between wires there is no problem. The reason that I am sure of this is that I had set up the Pololu current without the motor connected before switching off, waiting, connecting the extruder stepper and switching on.

At the moment I am ordering more Pololus, LCD displays, SD card modules etc., but I still have no idea what caused the problem.

Mike

I will endeavor to be more careful re static. I got a new stepper yesterday and after carefully plugging it in and doing an initial test everything was hunky dory. It was running at .4v as measured between ground and the trimpot. My other steppers are all at .5v but I thought I would start lower as opposed to higher. So I print some calibration objects and everything is still hunky dory. I print a fan holder in white cause the one I am using does not match the rest of my parts. It prints fine. So I print some more calbration parts as I am iterating iterating iterating. Everything fine. I go to bed. The printer does as well.

We both wake up this morning and I go to print another calibration test. The printer gets a short way into the print and the x-axis skips. I checked the voltages again and every thing was as I had left it. I have adjusted the voltage up on the x-axis and am reprinting and everything was working for the first 60 or so layers then the x-axis slipped again...though not nearly as much as above (2 cm versus 2mm).

...BUT...

Nothing changed between last night on the x-axis. I did tighten the y-axis timing belt but that is the only change. Could that have caused enough additional current pull from the y-axis to cause the x-axis a problem given my lower amperage adjustment there? I am concerned because, as stated above, I seem to be eating steppers and this working, not working behaviour, has been observed before!



I am calibrating for layer consistancy on x-y axis. Looking great on the Y-axis and all but two layers of the X-axis. Then I can work on the whiskers.

Thanks for advice,
Wisar

Edited 3 time(s). Last edit at 02/25/2014 02:20AM by Wisar.

Hi Wisar,

The only suggestion that I can make is to reduce step speed, acceleration and jerk by 10% or 20% and see if it goes away. As far as the failed Pololus goes, I have no other thoughts.

On my own blown Pololu problem, I have replaced the Arduino, all Pololus and the SD module, no new RAMPS or LCD having arrived. So far I have only been running it through a number of cycles on a 30V 20A bench power supply set to limit at 8A. I am still worried about running the complete setup without knowing what blew up one Pololu and then the whole electronics set. It feels a little like tempting fate to do the final connections and turn it on without having something to blame it on.

Mike

Something to remember with steppers:

Any fluctuations in your PSU voltage, will affect how your motors move. This includes slight drops due to the heated bed turning on (as it draws so much current), or just the PSU not quite performing on the day (temperature maybe?). It might not be much of a change but you might find that it's JUST enough to cause the odd misstep.

If you do run another calibration of the motor voltage, do it while the heated bed is turned on. Also it's usually better to err on the side of a slightly higher reference voltage (eg: 0.41v rather than 0.39v) for the same sort of reason.

I was wondering about power from the PSU. Slowing things down a touch seems to have helped. Have been running without a slip for a bit now (knock on wood).

[UPDATE of 4/3/2014]

Another two slips! One on the z-axis, which I have never seen before, and one on the x-axis. After the z-axis slip I added a big old honking cooling fan for the electronics thinking that it may help and can not hurt. Unfortunately the x-axis then jumped. At this point I have raised my main power supply voltage to 12.65 or so...and lowered my steppers to 0.43 from 0.5. I, know, two changes at one time, bad.

Iterate, iterate, iterate.

Wisar

Edited 1 time(s). Last edit at 03/04/2014 04:33AM by Wisar.

Sigh!!! it happened again

After replacing everything but the RAMPS board I was trying to run a strange missing X steps problem to ground when there was a flash from the extruder motor Pololu and the same old story as before: all Pololus dead, Arduino dead, SD card and SD module dead.

The extruder wasn't being used at the time - although the motor could still have been enabled. I was simply using the LCD/controller to cycle left and right on the X stepper. In the event that it rings a bell with anybody: The X stepper was erratically losing steps. It was doing this without mechanical or buzzing noise. I had tried currents from 0.4V (1 Amp) to 0.65V (1.625A) No more missed steps at either settting. Slightly fewer missed steps at higher speeds (25mm/sec vs 10mm/sec) Acceleration and jerk had been reduced to 1000 and 10 and everything was mechanically free without binding belts or bits catching. The mass of the moving parts is about half of that on a standard Prusa.

My main suspect, the loom to the extruder stepper, hot end heater and thermistor and heatsink fan has now been completely stripped to look for short circuits or bad connectoins and everything was perfect.

If this rings a bell with anybody and they have any clue what is going on it would be helpful. The only thought that I have left is that there is an intermittent fault in the RAMPS board.

Mike

leadinglights: I would check the motor itself. You may find that it has a bad winding or that at some place in it's movement it shorts out somehow (ie: not all the time, so usually you don't hit it). That's about the only thing I would think that would cause such a problem.

FWIW: Such motor faults are also quite capable of being the culprit for losing steps, and very hard to track down sometimes.

BTW: Which stepper drivers were you using this time and where from? While the A4988's are supposed to have short circuit protection, they only protect against shorts on each coil, and not shorts between coils.

Edited 2 time(s). Last edit at 03/04/2014 04:50PM by Cefiar.

Hi Cefiar,
Thanks for the suggestion. While I had checked for shorts between the motor windings, I had not checked for shorts caused by the rotor shorting them together - the first new possibility in a week. Sadly, even this has not been found - even at 100V between windings I can't find a short with rotating, pushing and pulling the shaft or even tapping it quite hard.

I think that this marks the end of my trials with 24V on RAMPS. I will have to go back to 12V which I have used on 5 or 6 setups without anything more dramatic than easily solved missed steps - too little current, too much speed, mass, acceleration or jerk. Certainly nothing like I have been getting since trying 24V

You asked which drivers I had used: in each case the extruder motor driver - the one that seems to have fed 24V into the 5V system was a genuine Pololu A4988. Other destroyed drivers have included Black series A4988 Pololus and what I think are clone 4982 drivers from Denmark.

Mike