This problem came up recently & was solved by adding a 2 ohm R in series with motor at startup then bypassed with FET.

 

Same problem that " GoatJames, Will stalling DC motor blow any fuses? "
had. From Chat, Nov. 4 2020. Problem solved by adding 1.5 ohm R in series with motor at startup, then bypassing R. Drawing attached.
Might find it in " Search Forums ".

 

From the posted specs:



Perhaps that is the problem.

 

Ultrafire batteries failed the soonest in all the tests on the web. Here is a video of a common fake one:

 

I misspoke, should have re-read old post. He was having problem with not starting on SMPS. On old test, 3,000 uF not enough.

 

If the batteries do in fact have a 4.1 amp internal limit, then they will not be able to provide enough current to start the motor. Adding that 2 ohm (or the current meter) in series reduces the peak current draw allowing the batteries to stay out of current limit.

Indeed, a large capacitor across the batteries might provide enough current to get the motor started, but it would need to be a rather large cap. Or you could use batteries without such a low current limit, or parallel some of the ones you do have. 'Soft-start' as implied in previous posts would also be a possibility.

 

How much do three genuine 18650 cells cost? $10? $20 for the high end stuff?

 

Same motor as used in post#22 will start with paralled 8,000 uF. cap. or a 1.5 ohm R in series. Current limited battery might be similar to current protection of my SMPS rated at 2 A. Motor resistance about 1 ohm.