Motor electrical noise? Seriously...?

I just recently got my extruder assembled, but then run into problem of random resets when I have the motor connected.

Actually it goes like this:
I had one gear DC motor, 12V 6ohm. I am using Magnetic Encoder, everything is fine except that it's like 30rpm and keep slipping on the filament.

So I decided to buy another gear DC motor. Similar size, totally different brand. 12V 55ohm, 6rpm. But when I have this motor connected, the magnetic encoder will give me random value when motor start spinning. Within 0.5 to 10 seconds the communication link will break, or the micro controller will be auto reset.

I soldered 0.1uF on the between the motor terminal, and 1 0.1uF each from terminal to the case of the motor. I twisted the motor leads, I add a little magnetic ring on the motor leads...neither of this help.

I happen to have a high power 12V motor driver, controlled by parallel port signal (used to drive my mini CNC spindle), and this new motor works fine on that beast.

Oh, btw, one of my DC Driver chip on the extruder controller just blow off while I was testing. I have absolutely no idea what's went wrong. I don't think I shorted anything but...anyway. I thought they have thermal protection inside?

Did anyone have similar experience?

Or should I just go ahead to build an opto isolation + DC driver...

I had a very similar problem with my GM3 motor, which I ended up resolving by using a different power supply. I did find a massive improvement when I put a hefty capacitor right next to the chip, I think in my case the problem was that as the motor switched on and off the voltage given to the chip fluctuated quite a bit. Check that any comms cables are either far away from any switching power cables (ie from the motor to motor driver), or they are at 90 degrees to each other. When you say blow off, did they just stop working?

Yes DC motors generate an incredible amount of RF noise. See my experiences here: [hydraraptor.blogspot.com] and the RepRap solution here: [hydraraptor.blogspot.com]

---

[www.hydraraptor.blogspot.com]

General rule of thumb when using brushed dc motors is that you solder 100nf (or smaller) across the terminals and from each terminal to the chassis of the motor

Cool...I skimmed through nophead's post and seems it might be the cure..A good thing to try over the weekend.

By "blow up", i mean literally, really!. It went into smokes. The chip surface has a hole! Something wrong with the schematic?

The poor thing is I don't have a scope. A Fluke multimeter is all I have...

Nophead, I saw you said that your 100nF disc capacitor doesn't work, but 1nF works.

Is that because of the disc problem? or 100nf VS 1nF makes the differences?

I have solder 3 * 100nF between the terminals, and terminals to the shell of the motor, though I am using the 0805 SMD i have on my hand. Would that help?

Though a bigger issue is I don't have any Diode on my hand, neither the Extruder Controller board does...I think it's the back EMF killing the A3949.

Hm...if I have to buy the diode...I think I would just go ahead and buy a OPTO+L298N driver board, the design totally isolate the logic and motor (both powers and group), I just saw one selling at 35 RMB.

Have you checked out the datasheet?
[www.allegromicro.com]

The L298 does however require external diodes for shoot through protection.

One thing that will likely break an h-bridge, but I dunno about chipped ones, but if two lines go high at the same time, that WILL blow up the h-bridge, trust me on this, i've done it, and learn't, just don't do it.

The chip already has shoot through diodes so you shouldn't need to add external diodes.

As for caps, this is more of a guessing game, especially without a scope, you need something less than 100nF, so say if you have a few 10nF around use those, you can try 100nF. Essentially the cap is just filtering the noise to a small level.

This is real basic electronics/robotics.

Also mentioning this from the conversation about the revised 3.0 driver board and making it EMC, further noise reduction is increased by using inline choke filters, all the major suppliers of electronic components sell them, and you can use them, having them is not going to damage your circuit in any way, its just a coil with a ferrite bead which is connected to earth.
Further reducing any noise is to use ethernet cable with the end being open very close to the motor, this is something i'd like to see in r4.0 electronics, more ethernet cable usage.

Right I am buying a L298 PCB & Parts, which has OPTO for signal lines, 4 Diodes for Back EMF, caps and such...I hope that will do the magic.

"One thing that will likely break an h-bridge, but I dunno about chipped ones, but if two lines go high at the same time"
You mean high side and low side both goes enable? Oh no of course I was not doing that. The A3949 control this logic internally anyway and it only accepts PWD+DIR, no way that I could enable both high / low side at the same time.
I think it's the back EMF that may be reaching 50V+ and break the chip.

I used to play some 24V 80W motor when I was in the college, building big robots, but by then we were putting a lot more protection and over-killing components, and the motor was much more expensive...I just didn't anticipate 12V 2W mini-motor could also cause big enough problem.

The reason for using lower value capacitors is that they have less inductance so are better at suppressing high frequencies.

The A3949 has built in clamping diodes and synchronous rectification, so it should not need external diodes. I haven't used that chip but I have noticed with their stepper driver chips that if you connect or disconnect the motor with the power on it causes the chip to latch up and take a lot of current on the 5V rail. I have a current limited 5V supply, but if it was connected direct to a PC 5V rail I could imagine it might destroy the chip.

The fact that disconnecting a stepper can upset these chips must mean they don't like fast high voltage transients. Since an unsuppressed DC motor generates such transients, then perhaps it could take out the chip, but the circuit is used with the Makerbot Cupcake which has a DC motor and I haven't heard anybody else have problems.

One thing to note is that a 12V 6R motor will take 2A at startup but if you reverse it while it is running you can get more like 4A.

---

[www.hydraraptor.blogspot.com]

My opto isolated + L298 came. Same-day delivery! It happens the seller is also Shanghai based (where I am living).

I hooked it up, change the code to use the Servo PWM pins, and yahoo, It works like charm.