Tongsheng TSDZ2 not providing assist. Fine on throttle

[Jez]

Hi all. Hope someone can help. I have a self build bike with Tongsheng (Tong Sheng) TSDZ2 500W mid drive purchased at the beginning of 2018. I have a home made battery producing 48V at 15A max.

The kit I purchased included a hand throttle and brakes.

The key thing to note is that if I run purely on the throttle it's fine. Works a treat. I can travel quite nicely at about 15mph (I understand that the throttle is speed limited) but the key intention for the bike is to use it for peddle assist and there is quite simply no assist at all.

I use a VLCD5 display which came with the kit. The firmware version is 3.6
Things I have tried:
Waggling all the wires (always a good start!)

Checked the hidden settings: I have a 26" wheel. Magnet setting is 1.
The next setting isn't clear. Some people call it walk assist and others refer to it as a 6Km startup limit i.e. if you are travelling at <6Km then nothing happens. Does anyone know? I have it turned off.
Current limit is 15A
There is an SD setting I've fiddled with. May have something to do with assist ratio. Tried all sorts here. Makes no difference. Default is 16 and I have it cranked up to about 45.
25 speed limit is off.

Entering the diagnostic mode on the VLCD5 shows a figure of about 80 for the standing torque. Crucially the other torque reading alters with pedal pressure so the Torque sensor seems to be working. On the bike stand I can hand push the crank arms around to about 120 (I'm guessing it's Nm?) so that message is getting to the VLCD5.

If I try the throttle whilst the bike is on the stand, the motor will not turn the wheel until some resistance is felt (by pulling on the brake)
and then off it goes. Just nothing if I try to get pedal assistance. It all seems to work a treat as an e-scooter.

Squeezing either brake leverl cuts the motor off straight away

Running without the throttle connected makes no difference.

My theory was that because I brought a kit which included a throttle the firmware was (in some way) not set up for pedal assitance? Doesn't make sense but can't think of anything else.

Is it possible to update the firmware in the VLCD5? I've found lots of info about an open source firmware setup that uses a different display. There are a lot more options for setup using this alternative but to be honest I would just like it to work.

 

[Jez]

Working!

The issue over no assist would appear to be proximity of the wheel rotation sensor to the magnet. Despite the fact that the speed was showing on the VLCD5 it would *appear* that the sensor needs to be even closer. Hopefully this will help someone out in a similar situation.

What I don't really understand is why this should fix it. If the throttle works perfectly and the VLCD5 shows a speed then signals must be coming from the wheel rotation sensor. Why should moving it closer to the magnet solve the isue of pedal assist not working? Could it be that I'm not actually on the solution at all? If it turns out not to have been the solution I will post an update.

Have fun with your bikes one and all

 

[nomidis]

Excellent!

Jez, by reading your post I understood the similar problem that I had as well.
In the first days of owning the TSDZ2 kit there had been times that I had no assist at all.
I did not ave the thumb throttle installed and I was frustrated about that..

But I saw that the speed sensor was not very well alligned and I made it closer to the magnet... and VOILA!
The next day I did not have the issue...

The thing is that I did not understand how these 2 things were related ... up to now, with your post!!

 

[Jez]

Excellent. Very pleased I could help

Now we just need to understand why on earth it behaves that way. My thinking is that if the VLCD5 can see a speed (i.e. I am going 10mph) then surely the pulses from the sensor are getting through. So if that's the case then why does the sensor need to be closer to the magnet so that (presumably) the pulses are stronger.

Theories?

(Going to change the first thread to include "Solved" as it clearly seems to be the issue here)

 

[Jez]

So for those following this thread, turns out that it was not the rotation sensor proximity after all. The bike was working really well right up until the point when the battery got low. Recharged the battery and the problem is back. The battery is a home built 48V and my current theory is that the controller is not happy at a higher voltage. Could it be that if I can get it used and the voltage drops then that will sort it? In some ways it makes more sense than the proximity issue before. Done a fairly extensive search and nothing out there helps much. New controller perhaps...

What is also worth mentioning is that after a full charge nothing much works at all As I play with it on on the stand, progressively there is a much greater chance that the motor will spin up when I advance the throttle. Initially it would do nothing but seems to work more and more as time goes by. This is so weird.

 

[Jez]

Just been for a ride down the road on throttle only. By the time I got back it's back to normal. Could it be that a fully charged battery stops it working but a lower voltage is okay? Voltage dropped from 48.1V to 46V. Other than that there is no difference. Think I'll start a new thread

 

[nomidis]

Post your problem here:
https://endless-sphere.com/forums/viewtopic.php?p=1459530#p1459530

the guys over there offer much technical help.

 

[Jez]

Hi all. I have a weird problem with my eBike and wondering if anyone else has seen anything like it.

The bike is a home build with a Tong Sheng TSDZ2 500W 48V motor and a home build 48V battery.

After an initial charge the bike does not provide any assist or operate on the thumb throttle

If I take it down the road and get a bit of action out of the motor by essentially just stopping and starting until the thumb throttle works I can travel along just fine on the thumb throttle. If I attempt to slow down on throttle only (take my foot of the gas but not right off) then it cuts out.

Progressively as I travel along the throttle starts to work more "properly" (I can slow down and speed up)

Finally full peddle assist works fine

Initially I thought it was the wheel rotation sensor but I think that was a red herring. My current theory is that a fully charged battery at 48V is affecting something. Checking the battery voltage at the point where all is working shows about 46V. Has anyone come across anything like this before?

Is this all suggesting a faulty controller perhaps?

<Edit>

Done a bit more research. There appears to be talk of an upper voltage limit on the Tongsheng that stops it working at 48v. My battery when it is fully charged comes in at at about 48.4. Could this be the issue?

 

[Jez]

Pretty certain I have this nailed now so again, I hope this helps someone else.

The Tongsheng motor appears to have a software-set upper voltage limit such that the 48V motor will not work if the voltage >48V because they apparently sell exactly the same or very similar motor as a 52V system so appear to stop the 48v motor from working with higher voltages.

My battery when fully charged tops out at about 48.4v roughly. However this is enough to push the motor into stopping. However, I have a thumb throttle which I added as an experiment (came with the kit but don't really use it) and it seem that by operating the throttle, enough load is put on the system to push the battery to dip below the 48v threshold and that allows the throttle to operate. Could be that it's not all that fussy about the thumb throttle and this upper voltage limit because at this point, pedal assist doesn't work.

So after running (up a few hills) on the thumb throtle for about three minutes, the battery drops below 48v and everything works!

 

[nomidis]

Wow!
Congrats on that finding!

 

[Antenor]

Hi
It should not be that. A 48v battery has a maximum voltage of 54 volts fully charged. So, unless this is a 36v motor, which will accept a maximum voltage of 42v, the reason for not working will be another.

 

[nomidis]

Oh!
Ok.
So Why your 48v battery is only 48,4 fully charged?
Maybe there is a battery problem?

Where did you get it from? Is there any guarantee?

 

[Jez]

Hi all. Thanks for your comments. I slightly misinformed everyone. it's been a while since I built the battery and I chose to go with a dumb battery (no BMS) and use a pair of BX100 monitors to keep track of the battery voltage. For some reason my BX100s only appear to monitor the first six cells but I actually built (and forgot I had done so!) a pair of 7s batteries. So following the logic of the issues I was wrong to assume that the voltage is just over 48V it's actually 52V of course.

The motor is a "48V" motor. I brought it after being assured by the chap in China that it would work with a 52V battery. It doesn't seem to be the case but again it was a year ago and I'm not going to go very far in demanding my money back.

I still have no idea why the two BX100 voltage monitors both indicate that I have a 6s rather than a 7s battery. Have checked the feeds to the monitors seven ways from Sunday and the voltage levels being applied are fine.

So if the theory is true that the motor won't run for a 52V battery (because they sell a "more expensive" 52V motor which in reality is the same motor with a software setting that allows operation at a higher voltage) then my only option is to get down and dirty with the controller in the motor to update the firmware to hoik it up to 52V.

So what with one thing and another it has been an interesting voyage here. I'm scribbling away again in the hope that someone finds this useful. There is so much I have learned from others on this thread and this is the best I can offer in return

To answer your question specifically it's a home made battery. The design is an interesting one which I'm testing out before I share It relies on a series of copper pipes to both carry current and remove heat from the batteries. But not sure how it's going to stand up to the bouncing about that it will get on a bike. Once I get the other issues sorted out I'll share some designs in the battery thread and everyone can have at it

 

[john61ct]

Great result.

I wonder what a more consistent / rigorous method for pulling top-voltage down enough would be, ideally not wasting the energy?

Loading up a discharged capacitor maybe?

 

[kortina]

Hi all. Thanks for your comments. I slightly misinformed everyone. it's been a while since I built the battery and I chose to go with a dumb battery (no BMS) and use a pair of BX100 monitors to keep track of the battery voltage. For some reason my BX100s only appear to monitor the first six cells but I actually built (and forgot I had done so!) a pair of 7s batteries. So following the logic of the issues I was wrong to assume that the voltage is just over 48V it's actually 52V of course.

The motor is a "48V" motor. I brought it after being assured by the chap in China that it would work with a 52V battery. It doesn't seem to be the case but again it was a year ago and I'm not going to go very far in demanding my money back.

I still have no idea why the two BX100 voltage monitors both indicate that I have a 6s rather than a 7s battery. Have checked the feeds to the monitors seven ways from Sunday and the voltage levels being applied are fine.

So if the theory is true that the motor won't run for a 52V battery (because they sell a "more expensive" 52V motor which in reality is the same motor with a software setting that allows operation at a higher voltage) then my only option is to get down and dirty with the controller in the motor to update the firmware to hoik it up to 52V.

So what with one thing and another it has been an interesting voyage here. I'm scribbling away again in the hope that someone finds this useful. There is so much I have learned from others on this thread and this is the best I can offer in return

To answer your question specifically it's a home made battery. The design is an interesting one which I'm testing out before I share It relies on a series of copper pipes to both carry current and remove heat from the batteries. But not sure how it's going to stand up to the bouncing about that it will get on a bike. Once I get the other issues sorted out I'll share some designs in the battery thread and everyone can have at it

I had a 52v battery with my TSDZ2 (VLCD5). Opening the motor it was the 48v version. The VLCD5 is programmed for the 52v.

Do yourself a favour and checkout the open source firmware with the KT-LCD3
https://github.com/OpenSource-EBike-firmware/TSDZ2_wiki/wiki/How-to-flash-the-Flexible-OpenSource-firmware-on-KT-LCD3

This makes the motor awesome, and highly configurable.

 

[Jez]

Thanks for that. Very useful but in another context! The open source thing is definitely on my list but was having some issues with battery life.

The VLCD5 has a relatively crude indicator of battery level but what I don't understand is how does it know what "full" is? The spec for the VLCD5 talks about a voltage input level from 11V to 60V. So the starting voltage could presumably be anything from 36V 48V 52V etc.

What you imply is that the VLCD5 has a programmed level which presumably goes with the kit i.e. a 48V setting for a 48V motor so that it knows what full is.

My issue here is that after about 20 miles my BX100 monitors start to beep indicating that one of the cells is below 2.7V but if I then plug the battery into my wonderful iCharger all the cells look great. The VLCD5 shows that the battery is "full" so I'm now beginning to suspect the BX100s. Anyway I'm rambling here.

Really just wanted to say thanks for your response and plan to look into the open source thing once I'm happy with my battery and I have the battery all in a nice plastic box to make it look pretty :lol:

 

[gfmoore]

@Jez
Hi, I know this is an old post, but what you described is what I am experiencing right now with a new motor and battery. I haven't let the motor drain as yet, my next job.

I just wondered if you had gained any more insights or had more thoughts on this issue?

Gordon