TSDZ2 Torque sensor reading zero 00000 yet hall sensor works

[tsdz2UK]

I have a TSDZ2 (orginal, with original software) that works on walk mode.

Doesn't work on pedal assist

TE and TE1 in the hidden menu on vlcd5 always read 00000

I measured AC voltage on controller side coil to be 0.7v and chainring side 0.6v

Torque sensor body is not damaged or cracked

I measured DC voltage at the torque hall sensor to be DC 5v, and 2.5v on the output. I used a magnet to see the effect on the output and it changes by moving a magnet over it.

Doing this does not move the reading on display TE or TE1 from 00000

I am now wondering, do I have a controller problem or a torque hall sensor board problem...?

I understand it back induces a current in the controller side coil is there a way to measure this...as it would tell me if the controller or the torque sensor board is at fault.

Should I hook up an ammeter multimeter in series with the controller side coil coil and would I be able to see it...?

 

[blowhole]

Yours is the first post I've seen here that included any multimeter measurements from the torque sensor. Thanks for that, would be interesting to compare to a working unit, though I suspect the actual "output" is too subtle to measure without hardcore tools.

Complete shot in the dark, but I can see how the damaged red wire could have shorted out a diode or something on the hall sensor output circuit. From what I understand the hall sensor is meant to be a relatively large, variable load that causes the input voltage/current to sag, similar to how the ceiling lights may flicker when you turn on a large appliance. The controller measures the sag/flicker as a proxy for torque. And this is just me talking out of my @$$, but if the hall sensor output signal is no longer driving anything then it stands to reason that there won't be any change in the amount of sag.

In any case I would recommend just buying a new torque sensor (~$100) or having it replaced under warranty. How long have you had your kit in terms of miles ridden, etc.? Mine failed right off the bat in the first few miles with TE/TE1=0 but has been fine ever since (after torque sensor replacement).

 

[tsdz2UK]

Thanks for that.

I will see if I can lay my hands on an oscilloscope, as it would be interesting to see the signals on a working system. I just used a basic multimeter, as you say the subtleties may be lost.

Seems such a waste to replace the whole torque sensor
unit for a 50p circuit....

 

[blowhole]

Seems such a waste to replace the whole torque sensor
unit for a 50p circuit....

Exactly how I feel about the black plastic assembly on the internal motor. I did reach out to one of the US-based suppliers and he informed me there isn't enough demand and that they would need to special order in lots of a thousand from the manufacturer. I imagine it would be similar with this piece.

 

[Elinx]

...
I will see if I can lay my hands on an oscilloscope, as it would be interesting to see the signals on a working system....

In that case the article of Stancecoke on the german forum is maybe interesting for you.

 

[tsdz2UK]

...
I will see if I can lay my hands on an oscilloscope, as it would be interesting to see the signals on a working system....

In that case the article of Stancecoke on the german forum is maybe interesting for you.

Amazing thankyou for this!

I'm sure I have a USB scope (just got to find it) and will compare mine.

This is a really fascinating device, I have a degree in electronic engineering and am very impressed with the clever design.

 

[Elinx]

...

This is a really fascinating device, I have a degree in electronic engineering and am very impressed with the clever design.

You are absolutly right. It is a very simple and easy to handle clever design
Only the mechanical tolerances and material choices could be better.

 

[tsdz2UK]

Right sorted now.

I got similar readings to stancecoke.

However I found the top pin in the connector to the display was bent.

This is the transmit pin, so no wonder data was not going from controller to screen. Data clearly went the other way as walk mode worked.

Anyway, fixed now! Rookie error not checking connections!

These connectors are a bit rubbish.

 

[blowhole]

Nice! I guess the controller assumes assist is disabled if the display does not transmit the assist level? Somewhere someone jumpered some pins to get a headless motor (no display), I wonder if that could have helped diagnose the issue in this case...

And the speedometer must not have been working either then? Was it just showing zeroes or blank, out of curiosity?

 

[tsdz2UK]

Hi it came to me in a box, already faulty

I was fixing it on the bench.

So the speedometer was disconnected.

I tested it using a chain whip and one crank to get some torque. When assist kicked in it nearly jumped off the bench. Not recommended.

I'm going to fit it to a bike soon