New "TSDZ2 Torque Sensor Central Motor"

Huguenot said:
First, the clock looses time. This seem to be a mfg flaw and perhaps the display should be replaced? Not critical but annoying. I have a Luna 750c on another bike which does the same thing. Is the 860 display any better in this regard?
Second, the drive works well and silently about one out of 3 startups. 2/3 the time it works but growls rather loudly. Shutting down and restarting a couple of times usually gets it into quiet mode. We are very careful to perform the startup exactly the same way each time (pedal down, no touching) but the behavior seems random.
Third, sometimes the bike will give a short "jerk" right after the startup screen switches to the regular display. This does not seem to be linked to the growling as sometimes it is quiet after, sometimes not. This jerk happens about 1/3 of the time as well.
And last, maybe not a problem, but lately when the display switches from startup to ready it will show "21.5mph" for a brief second before changing to "0mph".
What am I dealing with here?

Hi,

The display clock seems to be an inherent issue... As for the other two issues I think you may need a firmware update, as it is likely a glitchy firmware issue. If you send an email to directly to warranty you should get a response back ASAP, not sure what method you used to contact. I think it's a bit crazy everywhere right now. Feel free to reach out by PM too, I can help you troubleshoot and/or upgrade the firmware.
 
Huguenot said:
I have installed one with OSF (6.3) on a cargo trike. Easy install which I plan to document for the GitHub pages later this weekend.

And last, maybe not a problem, but lately when the display switches from startup to ready it will show "21.5mph" for a brief second before changing to "0mph".
What am I dealing with here?
Please document!! Can I after copy to my wiki (give your credit and linking)?
I am curious to the bicycle model you used....

Please stop using that firmware version or your motor controller may burn!! There are user reports that last version is much more stable and probably solved this fatal issue.

Don't worry about the wheel speed at startup, the system is working ok.
 
The bike is a Nihola "dog bike", which is their standard family model with a door in the box for pet access. The motor mounts very nicely on this bike, a tadpole bakfiets-style trike with somewhat improved steering over the more typical "swivel-the-box" type.
Alleged weight is 72# before adding motor and battery. These bikes are made in Denmark and rather rare in the US.
Here is a pic.
The odd looking bike next to it is a Pedersen, also a Danish design (from about 1900) but a copy built by me. I'm thinking of adding a motor to it as well.
I will take close up pics and do a write-up for you. I meant to do that when originally installing it but got ahead of myself.
Seems like I need a firmware upgrade but I am Mac based and more of a mechanic than a software guy so I guess it goes back to Eco Cycles for upgrade. They have contacted me and are "on the case", I will need to send it in for reflash.
Do you think the controller could be damaged already? Should it be replaced or inspected?
Thanks for your quick response and all the work you have done on this project.
 

Attachments

  • IMG_4057 copy.jpg
    IMG_4057 copy.jpg
    482.4 KB · Views: 1,047
New feature: battery pack resistance

One important technical characteristic of the batteries is not only the capacity in watt hours (WH) but also his internal resistance - the lower, the better.

The next firmware version will automatically calculate for you this value, you just need to pedal / run the motor for some seconds at a high battery current value like 10 amps. Then, you need to insert this value on the battery resistance configuration field - you should maybe increase a bit the value to be conservative.

Q: Why is it important and how it is used by the firmware?
A: It is used to calculate with a very good accuracy the state of charge of the battery (SOC) as also in future to calculate the energy watt hours per kilometer (will be shown on a graph on 860C/850C display) and estimate the available range in kms.

Here an example of calculation of power loss inside the battery due to his internal resistance:

Assuming the TSDZ2 is pulling 16 amps from the battery, and that my battery of 350 Wh (watt / hour) has an internal resistance of 275 milliohms, the power loss inside the battery is: P = R * I * I = 0.275 * 16 * 16 = 70.4 watts. As we can see, this power loss is equal to 20% of my battery capacity!!
I know there are battery cells on the market with half of the resistance, so, by using them I would have an extra 10% of my battery capacity on the previous usage example.

Real example of my battery and measure on SW102 - Resistance est is the value automatically calculate and Resistance the value I decided to configure:

TSDZ2-battery-resistance-estimate.jpg


Note that I validated this automatic measurement with 2 different battery packs, both with the same cells but one is a 14S2P and other 14S3P. The 275 milliohms was measured on the 14S2P and 195 milliohms on the 14S3P, which for me makes sense comparing the theoretical values.
 
@casainho I have professional battery testing equipment in my shop, don't hesitate to ask if you need some test data of packs with known configuration, cell type and age to compare with your data.
 
knutselmaaster said:
@casainho I have professional battery testing equipment in my shop, don't hesitate to ask if you need some test data of packs with known configuration, cell type and age to compare with your data.
When I release the firmware, would be nice if you could compare the value measure on the display and the one measured with that professional equipment.
 
sysrq said:
casainho said:
Cristian said:
Cristian said:
I searched around, but i can't find much about play in the axle. It started today, it's about 1-2mm of play. What is the cause? What do i have to replace?
Does anyone know what is causing play in the axle?
The play on the axle, search on Google, YouTube about the axle assembly, bearings, etc. PSWPOWER seels a product related to axle...
Couldn't find any product related to axle play. Probably will gonna wait till something better comes out as messing with the axle play seems to be too much to be asked from consumers.

I think the easiest solution to eliminate axle play is to install another bearing on the non-drive (left) side of the main axle. The bearing is 6902 2RS and can be found in many shops, I bought mine from here.

It's quite simple to install, no need to remove the motor, only the left crankarm and the rubber gasket on the axle. Then chek that the inner bearing is placed properly behind the snap ring, and put the second bearing between the snap ring and rubber gasket.

I really recommend doing this, I had slight play on my axle and after driving ~1000km there was quite a lot of wear on the plastic housing of the motor, and the torque sensor coils. Luckily I managed to fix them. Axle play also means that dirt moves easier inside the main gear housing.
 
ilu said:
I think the easiest solution to eliminate axle play is to install another bearing on the non-drive (left) side of the main axle. The bearing is 6902 2RS and can be found in many shops,

Yikes! That’s a 15mm bearing. The bottom bracket spindle is too small for its job. The smallest diameter spindles I’ve seen on normal bikes are 5/8” (16mm), and that was for kid’s BMX bikes. 17mm is the common size for cartridge bearing bottom brackets, and that’s none too generous.

You’d think the designers of such components would have a look at existing parts to see what is proven to work. I guess that’s too much to ask?

I was interested to try a TSDZ2 sometime, but I’m not keen to hurt myself. I think I’ll pass.
 
casainho said:
Huguenot said:
I have installed one with OSF (6.3) on a cargo trike.


Please stop using that firmware version or your motor controller may burn!! There are user reports that last version is much more stable and probably solved this fatal .


Casainho

Please can you confirm version 6.3 is not safe?
I thought it was a later version.
6.3 is still available to download.


Sent from my iPad using Tapatalk Pro
 
andrewgateway said:
casainho said:
Huguenot said:
I have installed one with OSF (6.3) on a cargo trike.


Please stop using that firmware version or your motor controller may burn!! There are user reports that last version is much more stable and probably solved this fatal .


Casainho

Please can you confirm version 6.3 is not safe?
I thought it was a later version.
6.3 is still available to download.


Sent from my iPad using Tapatalk Pro
Well, I think there are no reports of burned controllers with the version however, the jerk and unwanted motor start sometimes, is a bad thing that was corrected on the last version. I think is safe to use the last version and I am not think to delete the older versions as users will not install them since there is a more recent version.
 
casainho said:
andrewgateway said:
casainho said:
Huguenot said:
I have installed one with OSF (6.3) on a cargo trike.


Please stop using that firmware version or your motor controller may burn!! There are user reports that last version is much more stable and probably solved this fatal .


Casainho

Please can you confirm version 6.3 is not safe?
I thought it was a later version.
6.3 is still available to download.


Sent from my iPad using Tapatalk Pro
Well, I think there are no reports of burned controllers with the version however, the jerk and unwanted motor start sometimes, is a bad thing that was corrected on the last version. I think is safe to use the last version and I am not think to delete the older versions as users will not install them since there is a more recent version.
Maybe they should be removed if they are dangerous. I can’t be the only one with a policy of staying a couple of versions behind on experimental/beta stuff.
 
very wise

if it ain't broke don't fix it

but keep up with at least six months / three versions back
 
Balmorhea said:
ilu said:
I think the easiest solution to eliminate axle play is to install another bearing on the non-drive (left) side of the main axle. The bearing is 6902 2RS and can be found in many shops,

Yikes! That’s a 15mm bearing. The bottom bracket spindle is too small for its job. The smallest diameter spindles I’ve seen on normal bikes are 5/8” (16mm), and that was for kid’s BMX bikes. 17mm is the common size for cartridge bearing bottom brackets, and that’s none too generous.

You’d think the designers of such components would have a look at existing parts to see what is proven to work. I guess that’s too much to ask?

I was interested to try a TSDZ2 sometime, but I’m not keen to hurt myself. I think I’ll pass.

Agreed, that's still the only fairly discreet mid drive kit on the market with a torque sensor wich has some serious flaws. Wtf is going on with the ebike industry!?
 
ilu said:
sysrq said:
casainho said:
Cristian said:
Does anyone know what is causing play in the axle?
The play on the axle, search on Google, YouTube about the axle assembly, bearings, etc. PSWPOWER seels a product related to axle...
Couldn't find any product related to axle play. Probably will gonna wait till something better comes out as messing with the axle play seems to be too much to be asked from consumers.

I think the easiest solution to eliminate axle play is to install another bearing on the non-drive (left) side of the main axle. The bearing is 6902 2RS and can be found in many shops, I bought mine from here.

It's quite simple to install, no need to remove the motor, only the left crankarm and the rubber gasket on the axle. Then chek that the inner bearing is placed properly behind the snap ring, and put the second bearing between the snap ring and rubber gasket.

I really recommend doing this, I had slight play on my axle and after driving ~1000km there was quite a lot of wear on the plastic housing of the motor, and the torque sensor coils. Luckily I managed to fix them. Axle play also means that dirt moves easier inside the main gear housing.

Probably will gonna wait until it becomes sure that it's one of the best ways to doo it. Installation instructions with pictures or a video would be preferred just to be sure.
 
I did a testing release for TSDZ2 motor firmware that solves the popular overrun issue - on my tests it works quite well as on this video from mbrusa (because it uses the same solution - thanks mbrusa!!).

Can you guys please give feedback to help me make it for a new stable release?

Download here (should be used with the latest display firmware version 0.7.0): https://github.com/OpenSource-EBike-firmware/TSDZ2-Smart-EBike/tree/coast_overrun/releases/0.56.2

[youtube]RIa_0ZvcBQE[/youtube]
 
Huguenot said:
First, the clock looses time. This seem to be a mfg flaw and perhaps the display should be replaced? Not critical but annoying. I have a Luna 750c on another bike which does the same thing. Is the 860 display any better in this regard?
Today I tested a new 860C display and after about 36h the time drifted. On the other side, the one installed on my bicycle is keeping the time correct but I am using it every day.

Since there are supercapacitors on this type of RTC circuits, I now assume it is an supercapacitor and not a battery and so that is the reason why it holds correctly the time only for about one day and half.

Here an example of RTC circuit with a supercapacitor: https://wiki.52pi.com/index.php/Raspberry_Pi_Super_Capacitor_RTC(English)

image.png
 
casainho said:
Huguenot said:
First, the clock looses time. This seem to be a mfg flaw and perhaps the display should be replaced? Not critical but annoying. I have a Luna 750c on another bike which does the same thing. Is the 860 display any better in this regard?
Today I tested a new 860C display and after about 36h the time drifted. On the other side, the one installed on my bicycle is keeping the time correct but I am using it every day.

Since there are supercapacitors on this type of RTC circuits, I now assume it is an supercapacitor and not a battery and so that is the reason why it holds correctly the time only for about one day and half.

Here an example of RTC circuit with a supercapacitor: https://wiki.52pi.com/index.php/Raspberry_Pi_Super_Capacitor_RTC(English)

image.png
Could the supercapcitor be replaced by a coin cell battery that motherboards use for that function?
 
ilu said:
sysrq said:
casainho said:
Cristian said:
Does anyone know what is causing play in the axle?
The play on the axle, search on Google, YouTube about the axle assembly, bearings, etc. PSWPOWER seels a product related to axle...
Couldn't find any product related to axle play. Probably will gonna wait till something better comes out as messing with the axle play seems to be too much to be asked from consumers.

I think the easiest solution to eliminate axle play is to install another bearing on the non-drive (left) side of the main axle. The bearing is 6902 2RS and can be found in many shops, I bought mine from here.

It's quite simple to install, no need to remove the motor, only the left crankarm and the rubber gasket on the axle. Then chek that the inner bearing is placed properly behind the snap ring, and put the second bearing between the snap ring and rubber gasket.

I really recommend doing this, I had slight play on my axle and after driving ~1000km there was quite a lot of wear on the plastic housing of the motor, and the torque sensor coils. Luckily I managed to fix them. Axle play also means that dirt moves easier inside the main gear housing.

After installing the second bearing in this way, have you tried to remove the bearing. If it is installed outside the snap ring I foresee a problem removing the bearing?


Sent from my iPad using Tapatalk Pro
 
ilu said:
I have the 6v light cable coming from the dual connector speed sensor, and I'm only using one if the two wires for a basic front light. I'm not putting a rear light on this build, so was thinking about using the other end to make a USB output for charging a phone or something. I'll use a switch so only the light or USB is working at a time, but does anyone know if I can just connect a USB terminal to those 6v wires (I know USB power generally is 5v) or do I need a voltage regulator of some sort to safely charge electronics?

I know the amperage is also quite small so this would only be a backup solution, usually I won't need it.

Apparently this can cause issues. To be safe, I soldered a simple DC-DC 6-24V to 5V buck converter onto one of the 6V lines (in my case to charge a small LoRa GPS tracker). They are fairly cheap and tiny and already feature a USB-output. It charged - albeit rather slowly. I later removed it because I wanted minimal outside wiring and it solved a problem that wasn't really there.

One similar to this:
s-l300.jpg
 
Hasse67 said:
I know it's been warned about this before, by casainho and maybe others, but I'm putting out a new heads-up for this.
I was careless enough to just mount the rear torque arm and forgot the 2 outer screws closest to crank arm, and after only 3 trips with my new motor, chinesium housing was cracked :( so it is very important all fixing points are used :warn:
Does anyone know if I can buy a replacement anywhere in Europe?
Tsdz2 Cracked.JPG

Wow, serious damage. I have a couple of spare housings (old style) in decent condition so I can send you one from Germany.
 
andrewgateway said:
After installing the second bearing in this way, have you tried to remove the bearing. If it is installed outside the snap ring I foresee a problem removing the bearing?

When you remove the snap rings and main gear from the drive side, you can push the axle out of the left side and removing the bearing should be no problem.

Lateral play can also be fixed by putting thin washers on the inner side of the snap ring, but I didn't find suitable washers locally so adding the second bearing was simpler option for me.
 

Attachments

  • Eco Cycles TSDZ2 Lateral Axle Play Fix.pdf
    217 KB · Views: 148
ilu said:
.......
When you remove the snap rings and main gear from the drive side, you can push the axle out of the left side and removing the bearing should be no problem.....
Andrea_104KG has added the second bearing and has made a photo-report on the Italian Jobike forum
 
Back
Top