Repairing 3d printer control board - a4988 stepper ic

[gjh]

My favourite 3d printer control board, I think has a bung stepper driver.  (Melzi type  clone from Wanhao v5.1 board - Cocoon Create Plus)    I tried (w hotair) desoldering a spare 4988 pin compatible chinese clone (non allegro) ic from a spare board I had and whacked it on my control board.   bugger I wished I had a microscope to confirm my work, too hard to see with naked eye.    I did some testing with a multimeter on each ic pad to make sure I didnt bridge anything.    Z axis started to do something but it was pretty erratic in its movement behaviour.    Adjusting the current seemed to make it a little better but not 100%.. not smooth motion at all as if not all coils are firing properly.   So i decided to desolder  the stepper driver ic and put it back on incase I didnt form proper connectivity with all the pins.  Now Z axis is non responsive.  I dont repair much so Im out of touch with troubleshooting methodology.      I got myself a jewelers loupe and can confirm all the pins look like they are connected.   I did have a fair bit of flux still present on the board after my handy work..  didn't realise it can be conductive until after the fact maybe that did something.    Rest of the board still works fine.   I guess I need to check signal from atmega 2580 for that particular stepper ic.   I could try desoldering one of the other known working a4988 ic's and check if it works on the Z axis... or figure out what else to buy from element14 so I can get free shipping on a real Allegro replacement part.

Any guidance on repair strategy is welcomed.

   

[fzabkar]

What was the original symptom?

Datasheet:
https://www.allegromicro.com/~/media/Files/Datasheets/A4988-Datasheet.ashx

Direct links to images:
https://i.ibb.co/zbyCnvm/IMG-20190810-235848.jpg
https://i.ibb.co/TgMwCVr/IMG-20191223-024258.jpg

[Alfons]

Also the question: what were the symptoms? It is extremely rare for a driver IC (or stepper motor) to fail due to a defect. The fault is usually: cables and plugs, power supply and software errors. Tensioned axles, stuck bearings, etc. ... In most cases, it is a mechanical error, that force the drivers to their knees.

[gjh]

Thanks for the fast reply fzabkar & Alfons.

Original  symptom = Complete failure of dual Z axis output via single driver, no movement at all.       I've confirmed Stepper motors are okay, swapped them both over to other outputs like X / Y..  they work fine.    I removed stepper driver for Z axis, replaced with chinese clone.    Showed signs of life but z steppers just buzzed without real movement.  It might of just moved one step & then just buzzed thats it.  When connected to a different output one motor at a time,  both Z axis motors moved freely.     Actually just checking the board looks like I put the original driver back on, and definitely doesnt work.   

I'm assumed logically the z axis driver was gone, and maybe replacing with a chinese clone wasnt the best of all ideas.    The most testing I've done is just checking for solder bridges before powering it on when changing over the driver.   Power on testing besides checking driver outputs on the JST XH connectors seems a bit fiddly with the fine pitch.

[Alfons]

Did the driver become very hot during operation, was everyone cooled? However, there can be completely different reasons, for example if an end stop fails, the Z axis stops moving and would no longer move. That should also be checked.

[Mechatrommer]

I've confirmed Stepper motors are okay, swapped them both over to other outputs like X / Y..  they work fine.    I removed stepper driver for Z axis, replaced with chinese clone.    Showed signs of life but z steppers just buzzed without real movement.  It might of just moved one step & then just buzzed thats it.  When connected to a different output one motor at a time,  both Z axis motors moved freely.

fair enough.. please note the driver IC has bottom pad you have to solder to pcb to transfer heat from power element. without this connected properly, IC can burnt again. i suggest next time with new chip.. clean the pcb flat very well first, put just a little of solder and flux in the center and then solder the pins on each side, finally reflow the ic so bottom pad will be connected with solder as well. put heatsink on top to ensure best treatment to the IC. ensuring bottom pad connection can be difficult with diy hobbiest style, so if me i will see if i can drill the center bottom pad (for 2 layers pcb only) to the other side so i can make manual soldering from below. remember, use decent amount of flux for best soldering result.

or maybe you can just get the A4988 module in ebay (factory reflowed) and make bodged flying wires connections to the mother melzi board by removing redundant components first and connecting to corresponding nodes.

ymmv.

[Alfons]

For a few euros there are boards on which the drivers can be plugged. This is definitely the best solution. Soldered driver ICs are bad when it comes to tracking down malfunctions.

[gjh]

Did the driver become very hot during operation, was everyone cooled? However, there can be completely different reasons, for example if an end stop fails, the Z axis stops moving and would no longer move. That should also be checked.

Board ran pretty cool.   Good pickup, I'll check the endstop

[gjh]

For a few euros there are boards on which the drivers can be plugged. This is definitely the best solution. Soldered driver ICs are bad when it comes to tracking down malfunctions.

Yep I have a replacement board, but even then due to custom firmware this is my favourite board.. 

[gjh]

]fair enough.. please note the driver IC has bottom pad you have to solder to pcb to transfer heat from power element. without this connected properly, IC can burnt again. i suggest next time with new chip.. clean the pcb flat very well first, put just a little of solder and flux in the center and then solder the pins on each side, finally reflow the ic so bottom pad will be connected with solder as well. put heatsink on top to ensure best treatment to the IC. ensuring bottom pad connection can be difficult with diy hobbiest style, so if me i will see if i can drill the center bottom pad (for 2 layers pcb only) to the other side so i can make manual soldering from below. remember, use decent amount of flux for best soldering result.

or maybe you can just get the A4988 module in ebay (factory reflowed) and make bodged flying wires connections to the mother melzi board by removing redundant components first and connecting to corresponding nodes.

ymmv.

Yep I've taken the bottom pad into account.  I tried to keep it as flat as possible with minimal solder using hotair and always generous with the flux.   then afterwards I used soldering iron on the sides with a little extra solder to make sure pads had good connections.    Then triple checked with multimeter to make sure I didnt bridge anything.  bottom side had 6 vias underneath the center pad.   I probably need to go through the datasheet and check if I have every necessary.     All ebay 4988 modules would be using the chinese clone equivalent which apparently runs rough.

Flux used - https://www.jaycar.com.au/solder-flux-paste-56g-tub/p/NS3070