Solving Troublesome Noisy Treadmill Motor

  • Thread starter Thread starter Guineafowl
  • Start date Start date
  • Tags Tags
    Motor
Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
Sign up Log in
6 replies · 5K views
Guineafowl
Messages
927
Reaction score
420
TL;DR
Someone has posted this fault on a woodworking forum I’m on. The motor makes a humming noise when electrically, but not mechanically driven.
Link to thread: https://www.ukworkshop.co.uk/threads/very-noisy-motor.125693/#post-1398215

The treadmill is powered by UK standard 1ph 240V AC. The motor is a 180V, permanent magnet, DC brushed type and speed control is by PWM. There is an audio file in the thread; to me, the pitch of the sound stays relatively constant, suggesting the excessive noise is related to the PWM signal.

OP has ruled out mechanical problems such as motor bearings, and the treadmill belt/roller system.

Changing the motor solved the problem, but only for six months. As far as I can see, this is either a pattern failure of motors of a similar age, or a problem with the driver board, perhaps inducing noise from distorted PWM. I’ve done my best with remote diagnosis, but would welcome opinions on this.
 
Engineering news on Phys.org
I once helped maintain treadmills at a fitness center. I found it was very important to keep the belt lubicated to reduce the motors load.
 
Based on an elaborate probability/cost algorithm, my next troubleshooting step (it's free) would be to stretch the springs on the brushes (or shim them, if that's possible). PWM-controlled motors vibrate. Spring-loaded brushes have less seating force as the brushes wear.
 
Here is a spectral analysis of the noise at 10sec and 33sec after startup.

The upper graph in each display is a Spectrogram with Time on the X-axis and Frequency on the Y-axis. Color/Brightness shows amplitude (loudness).

The lower graph in each display shows Frequency on the X-axis and Amplitude on Y.

The first one, at 9.9sec, shows some signals around 5kHz and 17kHz which are probably from the PWM switching, although there is a small possibility those could be a bad bearing... or even LED or fluorescent room lighting. Most of the energy is in the 300-1500Hz band, which would be mechanical.
..
Treadmill at 10sec.png

..
The second display, at the 32.9sec mark, shows much energy over the whole spectrum with a broad peak around 1kHz. I suspect this is from the large belt slipping on something as mentioned by @dlgoff. This is also supported by the amplitude increasing with speed.
The upper graph shows (spectrogram) shows pulses in the 3-4kHz bands repeating at just under 1sec intervals. That may be one (or both) of the large rollers that the belt rides on. Check that the belt is not too tight and the rollers are clean.
Treadmill at 33sec.png

..

Please let us know what you find.

Cheers,
Tom
 
Last edited:
  • Informative
Likes   Reactions: dlgoff
Many thanks for the replies. I have passed them on and will update.
 
A scope trace of the V and I to the motor may be very telling.
 
  • Like
Likes   Reactions: Tom.G
Windadct said:
A scope trace of the V and I to the motor may be very telling.
I did ask the OP if he knew of anyone who could come and scope the motor drive, but no luck. The frustrations of remote diagnosis...