Tongsheng TSDZ2B noise / fixed pitch whine

[lepetitporteur]

Please help me diagnose the noise, which you can hear in the following video:

https://imgur.com/a/DOfQnsZ

The motor works quiet and assists perfectly up to the legal speed of 25km/h.
The assistance correctly tapers off at 25km/h and I can increase my pedalling power if I want to speed up above 25kph.
As soon as I accelerate to 26km/h or more, the noise starts while I am pedalling.
It sounds like the electric motor spinning with a fixed RPM, without any resistance.
The pitch of the noise is fixed.
On the included video my cadence is around 70-74rpm, the noise would be the same if the cadence was higher.
The noise stops only when I stop pedalling or the speed goes down below 26km/h.
As soon I slow down to 25km/h I can properly feel the assistance from the motor.

To summarise: Everything is working as expected in the range of 0-25kph. Also nothing feels wrong above 26kph when I use my legs to keep the speed up. It's just the noise / whine but it doesn't cause any additional resistance, the feeling is always smooth.

Conditions to reproduce the noise:
1. Speed above 26km/h.
2. Torque continuously applied with legs.
3. Any cadence.
4. Any gear.
5. Any assistance mode.

 

[dgc]

I will start with the easy question is this new or used?, I have not had this type of noise / whine from the TSDZ2(B) myself and I have had quite a few. So yes I am asking daft question(s) just to clarify, the mid drive works quiet and fine up to 25km/h, it is set to cut out at 25km/h, once over 25km/h the noise starts as the power from the drive motor backs off, when the speed drops below 25km/h the drive motor restarts, at this point doe's the noise / whine disappear? So just a shot in the dark may be the double clutch on the main gear is causing the whining on the overrun where you are pedaling and the clutch(s) release. I am assuming once over the 25km/h the motor has stopped along with the reduction gear and the outer section of the main gear now stationary leaving the inner section of the main gear free to rotate whilst you are pedalling. Look up differences between the TSDZ2 and TSDZ2B. Might be barking up the wrong tree but my first thought would be the clutch(s).

 

[dgc]

Daft idea, ride the bike as a normal bike with power off.

 

[lepetitporteur]

Daft idea, ride the bike as a normal bike with power off.

I will inspect the bike further in 4 days, on July 31st.

This bike has about 600km, it served as a demo in a bike shop. Almost new.

What makes me puzzled, if there was any issue with either of the two clutches, it would not work properly. The sound reminds me an electric motor spinning freely. For this two happen I see only 3 possible scenarios:
1. Motor side clutch slipping: Extremely unlikely, because the assist is working flawlessly.
2. Crankset spinning at a cadence greater than the maximum supported cadence. I am excluding that, because the noise appears even at a cadence below 70rpm.
3. Motor spinning backwards above maximum assisted bicycle speed due to some weird firmware bug: This is the only logical explanation I have thus far.

 

[lepetitporteur]

I am assuming once over the 25km/h the motor has stopped along with the reduction gear and the outer section of the main gear now stationary leaving the inner section of the main gear free to rotate whilst you are pedalling.

The big problem with this theory is the pitch, which is fixed. If the clutch between the stopped motor and the rest of rotating crankset was making that noise, the pitch would be modulated by my cadence. The pitch of this noise is fixed, unrelated to my cadence, unrelated to bike speed. It is on/off to fixed pitch with a noticeable speed up/down at the start/end.

 

[lepetitporteur]

I found video of a free spinning TSDZ2B online, same pitch.

A few common frequencies to my sound on a spectrum analyser: 500Hz, 1KHz, 4KHz.

Only my sound is smoother.

 

[dgc]

I understand what you are saying, one would expect a rise and fall in the pitch of the noise / whine. Still thinking about riding the bike powered down or say when over 25km/h when the whine start powering down. Are you into going thro' the display to turn the cut out speed up or down or is the setting fixed. Never thought about if the motor carries on spinning after passing the cut off speed until it's needed again to take up the slack and drive again, that could then be still the clutch(s) or the reduction gear(s) shaft or a bearing that skidding on a shaft or in a housing. Myself I would be at the stage of having a look inside for anything obvious, even something like lack of grease or the other extreme water or rust if its been ridden in the wet.

 

[dgc]

That video is interesting. maybe the motor doe's continuing running off load, as previous I would tend to give that drive assembly a good grease up and see what it was like after that.

 

[dgc]

Distance wise that not a lot, my TSDZ2 has covered has covered well over 6500km and still going strong.

 

[dgc]

Food for thought according to AI

"No, a TSDZ2 motor does not keep running when you exceed the set speed limit; it cuts off power smoothly once the wheel speed sensor detects you have reached or gone past the programmed maximum speed.

How the Speed Limit Works

• • Power Cutoff: When your speed goes over the limit set on your display, the controller stops sending power to the motor.
  • Pedal Feel: Because the TSDZ2 uses a torque sensor, you will only feel your own raw leg power once the motor assistance stops.
  • Sensor Dependence: The system relies entirely on the spoke magnet and the speed sensor. If the sensor loses signal or gets disconnected, the motor triggers an error code and cuts assistance entirely.

If you are experiencing unexpected motor behavior or trying to adjust your top speed limits, let me know which display model you have (such as VLCD5, XH18, or SW102) and whether you are running stock or open-source firmware so I can provide exact configuration steps."

 

[lepetitporteur]

I have already spent a considerable amount of time debating with ChatGPT about this issue.
We need more information and more tests. I will have the access to the bicycle in 3 days.
I will do some stationary tests, I will change the max assistance speed if possible. The bike is using a small VLCD7 remote display. According to its manual, the assistance speed is adjustable.

The symptom is pretty unique, as I have not found any trace on the internet of anyone experiencing the same.

 

[dgc]

Yep, as before new one on me and I have been using the TSDZ2(B)s for years, before I used a AKM mid drive which was more like a Bafang clone or original. Was running a new TSDZ2, after a week or two I was getting a drone from the mid-drive when it was driving, thought bugger fault with the TSDZ2 so took it off and refitted the AKM, the surprise was the drone did not disappear, the mid drive loads had stretched the rear wheel spokes, needed half a turn on each spoke to pull the tension up.

We have thrown ideas around all you can do is do stationary and riding experiments and report back. I assume this drive is not running the OSF.

 

[lepetitporteur]

The drive is stock Tongsheng TSDZ2B + VLCD7 setup on Le Petit Porteur Longtail

When adjusting maximum speed, the freewheeling motor noise starts as soon I exceed the assist speed limit.
If I set it to 11km/h, I can trigger the noise starting from around 12km/h.
Setting the assistance mode to "0" (no assist) stops the noise.

For now setting the limit to 27km/h helped in my case, as I rarely ride faster than that.

Feels like a firmware bug that spins the motor in the opposite direction when the conditions are met.

 

[dgc]

May not be something you would consider. Myself I would use the OSF software to reprogram the controller to the standard setup to see if that cleared the problem. Wrongly or rightly still wonder about the clutch system.

Question, just to clarify, if riding with power at 27km/h to the point where the drive assistance drops out and the noise starts what happen if you drop the assistance mode to "0" (no assist) or is this what you have done?

 

[dgc]

Not me but an AI response. I put in "tsdz2b noise starts when speed cut out point is reached". I know one has to be careful with the speed sensor setup. You may be right all along, I know with inverter / variable speed drives that you can get hunting or oscillation.

When a TSDZ2B motor starts making a fixed-pitch whining or whirring noise right at the speed limit cut-off point (like 25 km/h), it is usually caused by controller firmware behavior where the motor core keeps spinning at a fixed idle RPM or a minor sensor feedback mismatch.

Causes and Diagnostics

• • Firmware and Speed Sensor Glitches: If the magnet and the Speed Sensor are misaligned (ideal gap is 10 mm to 15 mm), the controller gets erratic speed signals right at the threshold, causing the motor to switch rapidly on and off or hum at a fixed pitch.
  • Controller Transition Lag: The stock or open-source firmware (OSF) sometimes fails to cleanly drop the PWM signal to zero power at the exact cut-out boundary, leaving the internal components energized just enough to produce an audible whine without adding real propulsion.

How to Fix It

• • Adjust the Sensor Magnet: Check that your spoke magnet lines up with the indicator arrow on the sensor and sits precisely 10mm to 15mm away.
  • Check Display Settings: Enter your display's hidden settings menu (holding power and info buttons) to verify that your wheel diameter and magnet count are configured accurately for your specific wheel size.
  • Test Without Sensor: Unplug the speed sensor temporarily to see if the behavior changes, noting that the speedometer will stop working during this test.

 

[lepetitporteur]

Question, just to clarify, if riding with power at 27km/h to the point where the drive assistance drops out and the noise starts what happen if you drop the assistance mode to "0" (no assist) or is this what you have done?

The moment I drop the assistance mode to "0", the noise stops. It only appears when the following conditions are met at the same time: 1. The actual bicycle speed exceeds the maximum assistance speed set. 2. There is torque continuously applied on pedals using cyclist's legs. 3. The bike electric engine is enabled in one of the four active modes.

At some point I will mount camera close to the motor and I will document everything more thoroughly.

 

[dgc]

OK, but try out the "How to Fix It" suggestions on my last reply, they do seem to make sense, worth a try, the speed sensor magnet could have moved.