PDA

View Full Version : can mach3 burnout transistors on a CNC control board?



kevinb
23-11-2010, 05:06 AM
hi
i did have major problems but luckly sorted them today,
basicaly the story goes, got a used trend smart cnc, with vac bed

3397
not had it long, done a few basic jobs on it that came installed on the pc, seemed ok worked as it should, so getting a bit more confidence I thought I would try some 3D carving, got a design of the net that used a diffrent code to what this one uses (.cnc), so thinking that the installed software (easycnc) was a little out dated, I installed mach3, set it up, zeroed the axis, started the program, all went well for about 2 minuets then nothing, thought at first the pc had crashed, but that seemed ok, so gave it about 10 mins to see if it would recover, well it never did,

now the real problems start, went to the onboard control panel to move the axis to recover the work, X & Z moved up and along no problem, Y axis was not good, the stepper drive motor screamed as if slipping, gave it a push and it moved normal for a while then stuck again and screamed, a push and back to normal, when moving a slight touch could stop it, at first thought the motor had gone bad, so to test swapped plugs with the X axis motor Y now worked as normal with full power and cant physically be stopped, X was totally dead,, put the plugs back to normal and was back to being weak and screaming,:cry:

well I took the covers off to have a look inside to check fuses and connections and saw a couple of transistors had blown 1 ( L6203 0-48VDC motor controller) and 1 (BDW93C 12A darlington transistor) so being fairly handy with the soldering iron, I thought why not have a go replacing them, anyway ordered bits and replaced them, but when put back together I still had the same faults, so thought there must be more damage to the board that cant be seen, so i left it, had another look at it today and went over with a test meter and found another resistor that wasnt working, so a quick trip to maplins and a bit of soldering and its back up and running,:dance:

now to the point of this post,

can mach3 burnout transistors on a CNC control board?
all was well until i used mach3 so thinking this was the cause? :confused:
installed mach 3, had a read through the instructions, set the axis travel calibration in the settings, and thats about it,
thinking back im sure i read some where about it damaging circuit boards if not set up correctly?
had a look back through instructions but cant find it now,
anyone know anything about this :question:
what else would i need to change in the settings :question:

dont fancy risking using it again if it going to blow the board again, i might not be as lucky next time,
the software that came with it (easy cnc) only runs .cnc and .dfx files, so do you think mach3 can be configured to work with this machine (think it's about a year 2000 build)
or should i give it up as a bad job :question:

is there such a thing as file converter software that can convert other file code types to .cnc :question:
or a more machine frendly program to run other file code types with no risk :question:

kevin :smile:

Web Goblin
23-11-2010, 07:21 AM
Kevin,
I havent used Mach 3 software but I suppose if it was not correctly setup it could have give problems with drive controllers but I think it would be more with stability issues rather than damaging boards.
What is wrong with the supplied software for controlling the machine? If it is purley for handling different file types then a decent post processor could be what you need. What code type does the machine run, ie Gcode or ESSI?
If you design your drawings in say Autocad then the post processor should convert the file to the correct code type you need.
There are many cnc code converters out there, some free and some not. Searching for cnc code converter on google should find you something useful.

regards

Ian

Gary
23-11-2010, 10:21 AM
Mach 3 cant have caused this problem, it is only controlling the axis, just like any other CNC software by controlling the clock and direction.

Robin Hewitt
24-11-2010, 12:56 AM
I may be teaching Granny how to suck eggs here, but that's never stopped me before :smile:

Catastrophic transistor failure in CNC is usually down to an inductive load.

eg: An NPN transistor sinks a coil to ground and current starts to flow. The transistor switches off but the current doesn't want to stop flowing.

Ohms Law V=I*R

You have current from the inductance. The resistance from the switched off transistor is enormous so the Volts become even more enormous. The breakdown voltage of the transistor is exceeded. Pop.

The solution is a diode across the inductive load so the unwanted current goes through the diode back to the top of the coil. Round and round the coil it goes until resistance takes care of it.

If transistors start going pop check the diodes across your inductive loads.

Slightly trickier to wire if you are driving the coil in both directions but there will be a circuit diagram showing where to put the diodes somewhere.

kevinb
24-11-2010, 05:43 AM
hi
ill give it another go at the weekend
the software that came with it was trend draw and engrave pro and easy cnc,
easy cnc only handels .cnc files as far as i can tell, the reason for mach3
so am i right in thinking easy cnc and mach3 are post processor's ? im new to this so still trying to get my head round it :confused: can you recommend a good one,
i ve been playing with a trial version of aspire that gives a .cnc toolpath,

sorry robin thats way over my head, if i see some thing that needs replacing i can do it, but understanding what all the bits do is something else, how would i test a diode with a meter,
i found the last faulty part with the continuity setting on a test meter, by following and compairing the circuits from each motor controller transistor, simple sience, but thanks,

if it blows again ill think about replacing with more up to date boards and drivers

thanks
kevin

Robin Hewitt
24-11-2010, 11:00 AM
sorry robin thats way over my head, if i see some thing that needs replacing i can do it, but understanding what all the bits do is something else, how would i test a diode with a meter, i found the last faulty part with the continuity setting on a test meter, by following and compairing the circuits from each motor controller transistor, simple sience, but thanks,

Think of it as water moving down a pipe. If you turn the sink tap off too quickly the water doesn't want to stop moving so the pressure goes exceedingly high momentarily. It stresses your plumbing and makes a loud clonk.

The diode acts like a magical one way valve that connects the two ends of the pipe work together with no distance in between. Now when you turn the tap off too quickly the water doesn't have to stop. It can go through the magic valve to the pipe start and keep moving round and round the circuit. No clonk.

Do you have diodes? If not, that is why your transistors are exploding.

Wobblybootie
24-11-2010, 12:00 PM
Hi Robin, Like Kevin your first post went over my head :confused:, but the analogy ... well, pure genius, even I now understand it ... well done!!! :beer::beer:

Robin Hewitt
24-11-2010, 04:39 PM
pure genius, even I now understand it ... well done!!! :beer::beer:

I like being called a genius, doubt you could convince my wife though :rofl:

Wobblybootie
24-11-2010, 09:06 PM
doubt you could convince my wife though :rofl:

Know that feeling ....

irving2008
25-11-2010, 07:19 PM
Does MACH3 drive the axis faster than the old software?

MACH3 on its own cant blow up a transistor but it could potentially stress the system more as it is far more capable.. High deceleration rates will generate much higher back EMF (Robin's clonk) and if that happens on two or more axis simultaneously with a older less regulated power supply the surge could take out older transistors which may already be running hotter than designed for due to MACH3 working them at a higher duty cycle..

kevinb
26-11-2010, 04:24 AM
hi
gave it another go today on a little 4" x 6" carving, just using the easy cnc software that it came with and it worked perfect,
reading what irving's wrote think that's it, so will leave mach3 alone, and look for a code converter,


3401

here's a photo of the board robin, arrowed are the bits that blew, plus a Darlington transistor that was on another little 2"x2" board that was joined to this by the yellow and pink wire,
don't know if there's diodes or not, didn't know what the other little bit was either, but the lad at maplins said something about resistor,

kevin

irving2008
26-11-2010, 07:50 AM
The line of black items 3rd row down are the flyback or catch diodes... therefore the issue probably wasnt back-EMF spiking but decelleration surge... that makes more sense if other parts died that are not directly related to the output transistors.... of course all has been stressed now... so budget for a new set of drivers sometime...



:lol: and you'd believe what a lad at Maplins said?:whistling: mind you i'm surprised he even knew what a resistor was, so maybe hope for them yet! :confused:

kevinb
29-11-2010, 04:24 AM
of course all has been stressed now... so budget for a new set of drivers sometime...
yes think if i have any more problems with parts blowing thats what i'll do


:lol: and you'd believe what a lad at Maplins said?:whistling: mind you i'm surprised he even knew what a resistor was, so maybe hope for them yet! :confused:

lucky he did, there were thousands of bits on the racks, i could of been there for days:eek:

thanks guys

kevin