Relay not switching correctly, controlled from ATtiny85, output to motor

Question

I have an ATtiny85 switching an NPN transistor, which will in turn switch on a relay that will run a motor. Everything is run off a 3.7V 18650 Li-Ion battery, but I have since moved over to a variable power supply with same problem:

Without the motor, the relay switches fine, ATtiny85 will switch the relay on for one second, and then off for one second, and repeat, and the oscilloscope (connected to relay output and GND) and clicks are showing correct behaviour for the relay.

When I connect the motor, the relay still clicks, but the oscilloscope is showing zero voltage. Seems like relay tries to switch, but doesn't. I connected an LED to Pin0 of the ATtiny85, and it flashes instead of staying on for one second and going off for 2 second.

Seems to me that the problem could be:

1. The ATtiny85 is resetting because not enough current going through. This doesn't make sense because I've switched over to a bench power supply and same thing happens, voltage is still the same.
2. Some sort of feedback from the motor? The motor runs fine if I connect it directly to battery so I don't think power is an issue.

Some additional information:

• Relay switches on for 1 second then off for 1 second. I am not doing PWM, but just have this on/off behaviour for testing.
• This is the motor: 3V-12V water pump motor
• Power supply is set at 3.9V with 2A current available. Increasing it to something like 5V still doesn't work.

EDIT: Although the answers all helped me figure out my problem, I chose the one that helped me the most. Found that it wasn't back-emf or your standard type of decoupling required. The motor was dropping the voltage down hard and no capacitor was going to fix it (already had these in place), although these are common issues.

More findings here

Answer 1

I have two suggestions for you:

1) check to make sure that the relay is able to handle the current the DC motor is pulling.

2) use the variable power supply to run the control circuit, and the battery to run the motor with relay. if this proves to be working, you can then work your way back to re-integrate the motor back to your system.

Answer 2

Depending on how the motor is wound, it could be giving an inductive kick onto the power supply rails as the brushes disconnect from the coils. This is why you see sparks at the brushes of a spinning motor. This kick could glitch the chip and cause exactly the kind of stuff you are seeing (relaTo remedy this, you can try putting another flyback diode across the motor leads just like the one across the relay coil.

It is also good practice to use a decoupling capacitor across the leads of the chips. This will filter any sudden power spikes or dips that could otherwise cause the chip to glitch. Try putting a ~1uF capacitor physically as close to the leads of the chip as possible.

Answer 3

Something like this:

^{simulate this circuit – Schematic created using CircuitLab}

Are you actually PWM'ing?
Why relay?
Why not FET?