TSDZ2 Strange Behaviour - How to debug/identify the issue?

Hello everyone,
recently I build up this beauty:

It's an amazing bike to go on the trails, however, I'm experiencing very strange behavior.

Most of the time, the motor does not work or outputs very little inconsistent power, it feels like it's not sensing that I'm pedaling.
Sometimes it's outputting some little power, then stops for a couple of seconds, then outputs again, then stops at all.

Other times, it does the opposite, outputs a lot of power even on the slightest pedal and even when I've come to a full stop it's "jumping around". I must admit when it does that it's quite fun to ride, feels super powerful although a little unsafe.
Few times it has been working as I would expect, but that was like 5% of the time...

The most common scenario is, I turn on the motor, everything lights up, and on pedaling after some delay the motor spins slightly a couple of rotations and stops...
I try restarting, changing the amps settings and keeping it between 16-20, usually without much effect.

I've ordered the 48V 750W VLCD5 model from PSWPower,
however, when opened the motor this is what I saw:

Have they send me the wrong motor? I've read that the motors are the same, only the firmware is different.

The battery is 48V 13,8Ah made using Samsung cells, and can output up to 40amps.

Now I'm looking for a way to find the cause of the issue and fix it.

Has anyone experienced similar issues? Is this something that's happening to the motor when misused somehow?

Is it possible to be related to the battery?

Any ideas on what to try, in order to identify the cause of the problem?

Thank you guys,
your help is much appreciated!

------------------------------
EDIT
What I've learned so far:
------------------------------

-The speed sensor doesn't matter, I can even disconnect it.
-The AMP setting is irrelevant, it's set in the factory
-Does the 48V motor has the 36V sticker inside? - PSPower claim it's normal, the controller is the same.

What I'm trying to find out at the moment:
-How can I check if my motor really is 48V or it's 36V?
-Is it the torque sensor? Will try replacing it to see if that helps

My first reaction to your problem was that it is possibly the magnet & wheel sensor spacing either being incorrect or dirty.
If the Sensor/Magnet spacing is more than 3mm or dirty that could cause intermittent signal loss. My next step would be to check all the other connections & then go into the back menu on your LCD Display. This link does a good job of explaining how to set the back menu items. - https://www.youtube.com/watch?v=RlMZTA36elQ
If everything checks out OK then it sounds like a faulty Torque Sensor.

You can try posting your question in the tsdz2 master thread too. There are a lot of people monitoring that thread and even a user (casainho) who is currently developing hacked and re-written firmware.

It looks like they sent you a 36v engine by mistake. With a hacked bios the 36v motor can be used with a 48v battery, casainho recently accomplished this. Mostly the 36v and 48v motors are the same except for firmware differences, however the motor windings are different with the 36v and 48v motors. The 36v motor is designed for 90 cadence at 36v, the 48v is designed for 90 cadence at 48v. If you run a 36v motor at 48v with a hacked firmware it will spin at roughly 120 cadence. Some users actually prefer the faster cadence.

Perhaps try running your battery down below say 40v and see if the motor works correctly at that voltage?

Here is the tsdz2 master thread where you should try posting your question/issue:

https://endless-sphere.com/forums/viewtopic.php?f=28&t=79788

Thank you very much, Daytriker!
What you are proposing is exactly what I did initially, double checked the speed sensor to make sure it's working correctly, and I'm playing with the AMP settings. Sadly, it doesn't seem to help much.

John and Cecil, thank you as well.
I will try with low voltage as you suggest.

I was thinking that maybe all motors have the 36V sticker inside, as I've read here(https://opensourceebikefirmware.bitbucket.io/kunteng_lcd3/) that the "SDZ2 motor controller is just the same for the 36V/48V/52V"
Is this correct, or if the motor was 48V it was going to say 48V on that inside sticker? (maybe someone with 48V motor can say whats on his sticker?)

Let's say it's the 36V, is this the expected behaviour if you try 36V motor with a 48V battery?

I am not sure. I know when I try using a battery with higher voltage than permitted in the firmware it will spin very slowly (maybe 1mph) and then it will have brief moments of full power occasionally depending on how far the voltage is above the voltage threshold. The closer I get to the max voltage the more the motor will kick on. This behavior is present with the throttle, I do not know if it is the same when using assist.

I have not taken my motor apart yet so I don't know what it states inside. I also got mine from pswpower and unlike everyone else that has a sticker on the outside that states the model power mine just had a sticker that stated pswpower.

The tsdz2 does not require the speed sensor for operation so I suggest removing/disconnecting it until you sort out your issue.

The amp settings are useless and do not work, at least they don't on mine and on many others. Perhaps it has been fixed on newer ones but as far as I know that setting is bogus...

Thank you very much, that is really helpful! Now I know to just ignore the speed sensor and the AMP setting as they are irrelevant.

Now the question is:

How can I check if the motor is 36V or 48V?

Does the 48V motor has 36V sticker inside on the controller?

If my motor is indeed a 36V, is this the expected behavior?

I've just ordered a throttle, I plan to use it in the debugging process. If the motor works ok with it, then it's the torque sensor.

Thank you, guys!

FYI the throttle only works with the VLCD-5 display and not the small side mount one. It looks like that is the one that you have now but difficult to tell from side image.

The best way I can tell you to determine if you have a properly functioning 36v motor is to discharge your battery to under 42v and then see if it works properly...

I have a 52v battery and a 48v motor with 48v controller. With my battery fully charged it does not work properly. I must discharge my battery to under 56v (58.8 v is the power output when fully charged). Once the battery is down in the mid 55's and lower it runs perfectly.

ps - if you post the 36v sticker question in the other thread you will probably get an answer. There are a couple of people that have disassembled a bunch of these motors.

A big flaw with the TSDZ2 is the lack of ability to check the basic set up via the "standard" vldc5 display, or a simple usb /pc link and user interface.
It ought to be possible to check the voltage settings (max, min, etc) and even alter them, without having to mess with firmware, as is possible with most motor controllers.

"There are a couple of people that have disassembled a bunch of these motors."

I have had a number of them apart that were 48v from the factory, mostly to install the factory supplied 52v controller and replace blue gears with factory supplied brass ones, and never have seen a 36v sticker.

The theory that replacing the blue gear with the brass one will cause issues putting more stress on the rest of the reduction system doesn't fit with my personal experience. My blue gear chipped and made noise during my second ride. Replaced with the brass gear to no ill effect after 700 miles of mostly strenuous single track riding.

Don't just ignore the speed sensor,.. unplug it to see if its a problem. Mine gives erratic operation when connected regardless of magnet gap.

Have you noticed if the inconvenience happens in the plain or uphill?
My engine in the plain does not give much assistance (almost nothing) because the torsion gauge feels the resistance of the rear wheel to the advance.
In the plain there is no resistance to progress and the engine does not give power! it's a fairly serious problem for me.

AWD said:

The theory that replacing the blue gear with the brass one will cause issues putting more stress on the rest of the reduction system doesn't fit with my personal experience. My blue gear chipped and made noise during my second ride. Replaced with the brass gear to no ill effect after 700 miles of mostly strenuous single track riding.

Click to expand...

The question I would ask is "why" did they choose to install a nylon gear in the first place? It is a known fact that metal on metal wears the metal down faster than if one gear is metal and the other nylon. In that case the nylon gear wears and the metal gear remains mostly unworn. Perhaps the nylon gear also acts as a shearing pin. It was put there for a reason, a weak link to give way if something happens that should not. Otherwise why not just put in a metal gear, there is no cost savings between the nylon and metal gear...

The problem isn't the nylon it is the lack of engagement it has with the metal shaft gear off the motor not being deep enough. Awhile ago we received a picture from our supplier of the revised blue gear for the next intineration of the TSD and it addresses this. Why they didn't just replace the blue gear with their own brass one knowing the issues on the TSDZ2 is a mystery.

Question is when the revised version will be made available as I expected it to be out by now.

AWD said:

The problem isn't the nylon it is the lack of engagement it has with the metal shaft gear off the motor not being deep enough. Awhile ago we received a picture from our supplier of the revised blue gear for the next intineration of the TSD and it addresses this. Why they didn't just replace the blue gear with their own brass one knowing the issues on the TSDZ2 is a mystery.

Question is when the revised version will be made available as I expected it to be out by now.

Click to expand...

Wow ! Thanks AWD...this is important intel !
Please keep us updated if you hear more on this, such as the improved gear part# and supplier.
Is there a simple way of telling the difference..diameter, ?, old part #, markings ?

Hillhater said:

AWD said:

The problem isn't the nylon it is the lack of engagement it has with the metal shaft gear off the motor not being deep enough. Awhile ago we received a picture from our supplier of the revised blue gear for the next intineration of the TSD and it addresses this. Why they didn't just replace the blue gear with their own brass one knowing the issues on the TSDZ2 is a mystery.

Question is when the revised version will be made available as I expected it to be out by now.

Click to expand...

Wow ! Thanks AWD...this is important intel !
Please keep us updated if you hear more on this, such as the improved gear part# and supplier.
Is there a simple way of telling the difference..diameter, ?, old part #, markings ?

Click to expand...

It's not a part that will retro fit to the TSDZ2 as far as I know. Tong Shen is re-doing the TSDZ to address the issues it has had in the past and the blue gear is the only part I have seen so far of that effort, not sure what else they are doing. They must have run into some issues because we thought we would have seen it by now.

Ahh !.. that is a bigger issue then than just a bad gear design or manufacture.
I should have realised , you cannot fix a poor gear mesh by a simple size adjustment,.. you have to change the shaft centerline positions to do it properly.
Maybe the "Lingbei " motor is the re-engineered Tongsheng ..?
TSDZ2 on the left ....Lingbei on the right..

Hi all,
thanks of being part in the discussion, I'm learning a lot and hope you are too

Specifically on my case, I've drained the battery, however there was no significant change, so I guess it's not the voltage...
Contacted PSPower and asked about the 36V sticker, and they have said that the controller is the same, just programmed for 48V.

What we have agreed, is to send me a replacement Torque Sensor, and I've also ordered the KT-LCD3 display.
One they arrive,
I will replace the torque sensor, and will flash the motor with the OpenSource Software from casainho, hopefully this will bring it back to life.

I've also got in touch with the "Lingbei" motor, and they will have a new, updated version of the motor in 1-2 months, with better torque sensing and other improvements.
I will probably order this one just to compare them both.

Thanks, everyone!

Yes, sounds like the torque sensor may be the issue. I would get it working properly before flashing it though. Once you modify the internal firmware you can kiss away any assistance or return support from the manufacturer, seller, and credit card company.

The nylon gear is ok if ridden properly. I do not know your intended use (street riding vs stair climbing) but when ridden sensibly on the street it can last a long time. We have about 1000km on ours and so far it feels fine, and we are pushing 275lbs including the bike and we ride in a very hilly environment. I ride in economy a lot though, especially near hills and then use the throttle for a power boost when needed. When it is fairly flat I ride in turbo and use the throttle as a clutch, but I try to limit motor output on the initial takeoffs. I guess only time will tell but so far so good.

Good luck, I hope it is fixed soon. Please let us know how it works out.

Hi Alex,
It does sound as if you might have a bigger problem, but lets go back to the basics for a moment:
A couple of questions:
1) Are you aware that when you turn the system on, you should keep your feet off the pedals until it has gone through the whole start-up procedure? Each time it starts, the controller looks at the voltage coming from the torque sensor and sets that as "zero torque", so only applies motor power once the pedal torque exceeds this figure. Therefore having torque applied to the system (force on the pedals) during start-up confuses the system and decreases performance until the system is powered off and re-started.
This can make it seem that the system works at full power sometimes and reduced power other times.

2) Have you fully checked the battery contacts? Quite often, intermittent performance can be caused by bad contact between the removable battery and its base (or somewhere else in the system). I once had to modify / upgrade the setup on a removable battery quite similar to yours to solve an intermittent issue. Your battery might have four contacts, two positive and two negative.. check that the matching contacts in the battery base are all connected.. sometimes only two of the four are wired.

If you need it, I posted a bit of information on changing the torque sensor on the main TSDZ2 thread a few weeks ago (May 27th, Page 88 of the thread). It isn't too hard but there are a couple of tricks that it helps to know about before you start (I didn't, so made the mistakes...).

Hope you get it sorted soon.. that looks a great bike!
Dave

That is crazy - I just took my new 750W 48V TSDZ2 (purchased from ebird w/ VLCD2 display) out for its first ride today and I am having nearly identical problems.

For the first mile everything seemed to work as it should. Then I noticed the motor was cutting out for a few seconds causing my ride to be a lot less smooth. Around mile 5 the motor cut out completely. The VLCD5 display did not indicate any problems and still registered full battery power and accurate speed. But the motor did nothing.

I tried turning things off and then back on. The motor started to work again, barely. It only seemed to function when I stood up and gave full force to the pedals.

• Dont think it is speed sensor. It registered accurately the whole time.

• Don't think it is the battery. The voltage reads fine.

• Don't think this is normal behavior. I have another 500w TSDZ2 that has worked perfectly for 2000+ miles. Very smooth. Never cuts out.

I contacted ebird to let them know I think the motor may be faulty, but I have a feeling I will have trouble getting a replacement shipped. Please let me know if you find a solution.

wesblog said:

That is crazy - I just took my new 750W 48V TSDZ2 (purchased from ebird w/ VLCD2 display) out for its first ride today and I am having nearly identical problems.

For the first mile everything seemed to work as it should. Then I noticed the motor was cutting out for a few seconds causing my ride to be a lot less smooth. Around mile 5 the motor cut out completely. The VLCD5 display did not indicate any problems and still registered full battery power and accurate speed. But the motor did nothing.

I tried turning things off and then back on. The motor started to work again, barely. It only seemed to function when I stood up and gave full force to the pedals.

• Dont think it is speed sensor. It registered accurately the whole time.

• Don't think it is the battery. The voltage reads fine.

• Don't think this is normal behavior. I have another 500w TSDZ2 that has worked perfectly for 2000+ miles. Very smooth. Never cuts out.

I contacted ebird to let them know I think the motor may be faulty, but I have a feeling I will have trouble getting a replacement shipped. Please let me know if you find a solution.

Click to expand...

Although you do not think it is the speed sensor did you try disconnecting it to see if it still exhibits the same behavior? The motor does not require the speed sensor to function and it is very easy to disconnect it and be certain. Also try charging the battery to full and see if that has any effect. These 2 things are very simple to test. Good luck!

I've gone on a few more test rides.

The issue with needing extreme pressure was solved by not putting feet on the pedals when turning on the motor. However, the issue with motor power cutting out is still occurring.

If I unplug the speedometer or remove the magnet the motor does not provide any power.

Did the original poster ever find a solution?

Did you ever figure out what was causing this issue. My bike is still having problems. It rode great for about 200 miles and then it started cutting off power intermittently again. Removing the speed sensor won't fix it.

Do you think maybe it could be mistakenly sensing the brake levers? I didn't install them on my build, but it is behaving as if the engine is cut off (display remains on with full battery, but no assist power).

wesblog said:

Did you ever figure out what was causing this issue. My bike is still having problems. It rode great for about 200 miles and then it started cutting off power intermittently again. Removing the speed sensor won't fix it.

Do you think maybe it could be mistakenly sensing the brake levers? I didn't install them on my build, but it is behaving as if the engine is cut off (display remains on with full battery, but no assist power).

Click to expand...

Not sure if this is related, but once after opening up my unit to add some grease I had a similar problem. The display would be on but the battery would go to zero (on a full charge) and the motor would provide no assist. It turns out that when I put it back together I routed the 5-pin connector harness too close to the phase wires. Once I moved it out of the way it worked like new.

Im also having this problem! Usually mine will work on initial start up/turning on. Then 100m to half a km it will go from truning on to off and the whole ride is just a headache and a struggle. A lot of times turning it off and on it would work again but only as far as "about 100m". The bike was actually working fine until I went up a steep muddy hill, the motor really threw me up there until just before the top when my tires and chain were full of mud and sticks the thing went CONK! It was the torque sensor which broke so I replaced that. Not going up and muddy hills anytime soon. Are we suppose to use grease on and around the torque sensor shaft? I did..? Anyway after putting back together has been a headache.


If I find the problem I will post it on here!