Geeetech Melzi motor problems

I bought one of these some while ago and when connected it ran amock and the MOSFET for the bed started smoking. I found there was no bootloader and the MOSFET was under-rated so I bought a Melzi from Mophead which worked fine until I shorted out a thermistor terminal. Mophead has offered a solution which is currently outside my skill set, at the moment, so I went back to the Geeetech board.
I have sorted out the bootloader issue and changed the MOFSET and the Marlin sketch I had previously used uploaded OK.
The heaters worked OK and were controlled by the thermistors. However I now have motor problems.
I'm using Pronterface and initially all the axis motors juddered when an input was made using the graphics on the top left of the screen. I've now set all the motor trimmers to .4 volts and whilst they make the normal motor sound for a short period they don' move when using the graphics interface. I've checked the wiring and all is well.
The extruder motor works OK with the same wiring arrangement albeit it extrudes more filament than demanded, I can adjust this in config.h. I connected the x-motor to the extruder terminals and it motored OK, so I presume the wiring is correct.
I then tried instructing the motors through the input field below the log window and the X and Y motors work but only after I have tried to motor them through the Pronterface graphics. However, a command of G1 X5 F500 causes the x-axis to move 80mm not 5mm, with the Y-axis operating in a similar manner.
The extruder works through the graphics but not with command M302.
The Z motors screamed with no movement so I increased the voltage .6V with no effect.
All the microswitches function using the G28 code.
Any help would be appreciated.

Looks like the Geeetech is missing the micro step jumpers. On the Melzi these are solder blobs across a tiny gap in a pads. Without these it will be trying to go 16 times too fast.

---

[www.hydraraptor.blogspot.com]

Thanks Nophead. The soldering did the trick, at least for now it seems to work. I've had so many problems with these boards they are just not worth the hassle but it's all part of the learning process.

my Z axis stopped working for no aparent reason after completeing a print on Friday night. won't home can't jog with pronterface. it finished that last run just fine. checked endstop its operational, and flipped y and z and motors are fine on y. Original geeetech melzi. buying a new board isn't in my budget right now so I'm hoping there's an easy fix.

Do the motors lock and not move, remain free or vibrate on the spot?

The most likely fault is a bad solder joint on the stepper chip or the MCU. Try applying some pressure on those to see if it is intermittent.

How easy it is to fix depends on you soldering skills.

If you are not using the expansion port you could map the enable, step and direction pins to that and attach an external stepper driver.

---

[www.hydraraptor.blogspot.com]

I'll take a look at solder joints when I get home, but the motors remain free. Thanks for the input nophead. I will let you know.

If neither coil is energised most likely the enable signal is not getting through or the VREF voltage is zero.

---

[www.hydraraptor.blogspot.com]

last night while jogging Z with pronterface I gently pressed on the stepper chip for a moment I got some oscilation of the z motors, but aparently this didn't get a good enough connection to get actual movement. So now I know what it is thanks nophead. Tonight I'll sharpen up the tip on my soldering iron and give it a shot at soldering those little connections.

Gave soldering the chip a try and gained something now the z motors buzz but don't turn and do stop when the endstop switch is triggered. Am I still looking for a bad solder joint or something else?

If they vibrate but don't turn it is normally a sign of only one of the two coils being driven. That does seem different from your original problem. Odd to have two different faults.

---

[www.hydraraptor.blogspot.com]

Well I did make my way around the chip attempting to resolder but as you know they are tiny connections and my fear was bridging. it is possible that I missed something but does that make sense? or is it likely that I have more than one problem?

My best guess would be you fixed the missing enable but also opened or shorted one of the outputs.

Or maybe both power pins were missing and you restored only one. Or both coils were open and you restored one. It would be odd for two pins to both fail at the same time though.

They are hard to hand solder because most of the contact is underneath the chip.

---

[www.hydraraptor.blogspot.com]

Nophead you're the best! Back up and running as before. found a trace to the stepper driver that was pulled up. I got it connected and we're as good as we were. Thank you!!!

Now if I could warm the bed up a little without fryin the mosfet any thoughts. (my previous post [forums.reprap.org])

It's just a matter of using a MOSFET with a low enough RDSon at 4.5V. The Melzis we supply use Fairchild FDD8870, which is only 4.4mΩ It has no trouble powering a Mk2 bed.

---

[www.hydraraptor.blogspot.com]

I used this one [www.digikey.com] when I replaced it when it fried. Do you think it'll be ok? says its good for 85w and the Fairchild FDD8870 says 160W. Even if I have to set a limit on the heat I'd like to at least turn it on

The maximum wattage is not the relevant parameter here, RDSon is. The HUF76423P3 has an RDSon of 35mΩ, so it will dissipate about 8 times more power at the same current. Also it is not a surface mount part so it won't fit the PCB.

With a 10A bed the FDD8870 will dissipate only 0.44W and it uses copper on the PCB as a heatsink and barely gets warm. The HUF76423P3 will dissipate 3.5W and will need to be bolted to a heatsink.

---

[www.hydraraptor.blogspot.com]

Unfortunately 2 years ago the first time I tried to "turn on" the heated bed it majorly fried the factory surface mount mosfet and the circuit traces to it which was why I used the HUF76423P3 as a replacement so I could bend the legs to make the connections. Then I epoxied the area of the board where the legs connect as the traces were fragile. So using the FDD8870 won't work but I could look for a cross ref. Maybe I should just abandon the heated bed on this machine and save it for the next one. Thanks again for all your help.

Well Z on the Melzi quit again while my son was trying to print something for school nice huh! So Dumb question if I replace the Melzi with MEGA 2560 R3 + RAMPS 1.4 + A4988 Stepper Drivers and upload the marlin set up just as I have it now on the melzi will I be back to where I was? I've been too busy with work lately can't get my head around it. Hoping you guys can help.

You just need to change the board number to the one for RAMPS and make sure you use the correct endstop connections and jumper the unused three. I.e. X min, Y min and Z max.

You will also need to turn the current down to about 1A because stepstick format boards generally can't do 1.25A.

---

[www.hydraraptor.blogspot.com]

Thanks again Nophead. I have other questions/thoughts on the change over like maybe I can make my heat bed work now I had done away with it on the Melzi. but this is the isn't the right thread. Is there a good general thread for that sort of thing?

Yes ramps should be able to do everything a melzi can in theory. We use melzi because the ones we supply have better connectors, MOSFETs and stepper driver heatsinking on the PCB, so they don't need heatsinks and fans. On ramps you have to crimp connectors on all the wires and might have to short out the bed polyfuse as it will be right on the limit with the bed PCBs we supply that are less than 1.1R.

---

[www.hydraraptor.blogspot.com]