New "TSDZ2 Torque Sensor Central Motor"

[Antenor]

Is it possible for Casainho software to work with the original display?
With vlc5? Or is Kt lcd3 required?
Thank you

 

[andyme]

Is it possible for Casainho software to work with the original display?
With vlc5? Or is Kt lcd3 required?
Thank you

LCD3 required or in the near future bafang 850C

 

[sac02]

...
14s2p is not enough and 14s3p would be way too heavy unless there are some better cells out there than the GA's ?
...

Indeed there are, but it seems few people are yet willing to look outside the 18650 box, which I find surprising. I'm building my battery pack with 21700 cells. 21700s are slightly more energy dense because they have a higher battery:steel case ratio, and they have higher discharge capabilities due to the more massive battery.

Here is a comparison of Sanyo GA vs Samsung 48G. Note both the significantly higher capacity, and the significantly higher load handling capability (much less voltage sag) that the 48G cell has compared to the GA. I specifically chose to plot 7A and 10A discharges so that you can compare the GA in 3p vs the 48G in 2p at a theoretical 20A~21A drain. A 18650GA weigh about 47.4g, a 48G 21700 weighs about 67.6g.





14s3p in 18650 using GA would be 10.2Ah, and weigh 1.991g (cells only).
14s2p in 21700 using 48G would be 9.6Ah, and weigh 1.892kg (cells only).

So you sacrifice 6% capacity, and gain 5% in weight savings - but you also get improvements in load handling (very significant improvements after about 30% discharge), although that is harder to quantify easily.

There are also newer cell in the pipeline that will outperform the 48G, but have not been tested and entered into HKJ's comparator yet. The 21700 is the way of the future...

 

[Hillhater]

Whilst that discharge test is informative, it is not a true comparason of performance due to the different capacity of the cells.
It would be better to repeat the test with 3P, GAs and 2P 48Gs,....if that is the intended options.
Im sure the 48Gs will still give the better performance , but i suspect the difference in sag will be much less and make the choice less obvious....especially when prices are considered.

 

[sac02]

Whilst that discharge test is informative, it is not a true comparason of performance due to the different capacity of the cells.
It would be better to repeat the test with 3P, GAs and 2P 48Gs,....if that is the intended options.

I'm not sure I understand your comment?

At about 21A draw, 3P GA will need to supply ~7A per cell, so look at the 7A trace for GA.
At about 20A draw, 2P 48G will need to supply ~10A per cell, so look at the 10A trace for 48G.

The 48G voltages are higher both for instantaneous draw and for the duration of the discharge. I think you are implying that by having 3P, the capacity of the GA will effectively be higher than it appears on the graph (true), but it does not change the cell's characteristic at 7A draw. The GA at 7A draw will Vsag more than the 48G at 10A draw.

By the "simple" rated capacity numbers, the 3S GA pack will be 10.2Ah, and the 2S 48G pack will be 9.6Ah - but the 2P 48G will perform better than the 3S GA pack.

Im sure the 48Gs will still give the better performance , but i suspect the difference in sag will be much less and make the choice less obvious....especially when prices are considered.

Bulk price from US source for GA is $4.35, or $182.70 for 14S3P.
Bulk price from US source for 48G is $7.25, or $203 for 14S2P.

I will gladly pay the extra $20 (10%) for improved performance and 100g lighter weight at the same time.

 

[Ron Paul's Blimp]

I will gladly pay the extra $20 (10%) for improved performance and 100g lighter weight at the same time.

Is there anywhere I can buy a 14s2p or 13s2p complete battery (with BMS) made with 21700s?

 

[Hillhater]

Whilst that discharge test is informative, it is not a true comparason of performance due to the different capacity of the cells.
It would be better to repeat the test with 3P, GAs and 2P 48Gs,....if that is the intended options.

I'm not sure I understand your comment?

At about 21A draw, 3P GA will need to supply ~7A per cell, so look at the 7A trace for GA.
At about 20A draw, 2P 48G will need to supply ~10A per cell, so look at the 10A trace for 48G..

The discharge "trace" is defined by voltage and capacity.
If you parallel the cells as suggested the end capacity will be much better matched (9.6 & 10.?). So the slope of the trace will be very different.
Try it and report back .!

Also, you could theoretically calculate the sag from the internal resistance of the 2p and 3p configurations.
Both cell sizes have approx the same 0.040 ohm DCIR ,
So a 2p config of 21700 would have an effective DCIR of 0.020 ohm (in theory) and a Vsag of 0.4 v @ 20amps
And a 3p config of GAs would have an effective DCIR of 0.015ohm (in theory) and a Vsag of 0.3 v @ 20amps
Theory suggests there would be less voltage sag using the 3p of GAs ? :wink:

PS: There are vendors on this site selling new GAs for $3 or less.
So your 14s, 3p , pack could cost $126 or less
That 48G pack could cost you $80+ more, and have more sag ,!
......but yes, it would be 84gms lighter !

 

[Ron Paul's Blimp]

Bumping this. Is that chain line bad news? Is the resistance normal?

Is this chain line too extreme?

https://imgur.com/a/NV6ny3Z

That's the stock TSDZ2 42T. I was planning on replacing it with a narrow-wide, but since the stock ring looks a bit dished inboard, the line would be even worse with a regular non-dished ring.

Also it feels like there's significantly more resistance when turning the cranks compared to the stock BB. I don't think it's due to the chain line since it feels like it's still there in the gears where the chain is straight.

 

[andyme]

Bumping this. Is that chain line bad news? Is the resistance normal?

Is this chain line too extreme?

https://imgur.com/a/NV6ny3Z

That's the stock TSDZ2 42T. I was planning on replacing it with a narrow-wide, but since the stock ring looks a bit dished inboard, the line would be even worse with a regular non-dished ring.

Also it feels like there's significantly more resistance when turning the cranks compared to the stock BB. I don't think it's due to the chain line since it feels like it's still there in the gears where the chain is straight.

i solved potential chain line issues by mounting a double chainring. but only as a side effect. i wanted to maximise the fast and and minimize the slow end. so i have 60T and 36T...(because my wheels are only 20")

so i can use the outer half and the inner half if i want.

re. the resistance: i am not sure if we are talking about the same thing, and i have no experience with BB:

if you are asking me: is it more difficult to pedal with mounted motor in off than without motor: yes. clearly. but that was already the case when i was still using the stock firmware

 

[casainho]

I implemented the clock time (time keeps running even when LCD is powered off). The hours and minutes are setup on LCD configurations menu, on the Display section.

Now everything is working, all the data that is shown on KT-LCD3 + the clock time are working. What is missing is design so the final main screen can be developed.

[youtube]idfb1XvqRuY[/youtube]

 

[sac02]

The discharge "trace" is defined by voltage and capacity.
If you parallel the cells as suggested the end capacity will be much better matched (9.6 & 10.?). So the slope of the trace will be very different.
Try it and report back .!

Now I understand your comment, and yes that makes sense.
I have the cells and the test equipment, and empirical trumps theoretical data in most cases, so I will run the test myself and see if 3P GA or 2P 48G performs better. You will need to give me as many as two weeks though, life is too busy... sigh.

Also, you could theoretically calculate the sag from the internal resistance of the 2p and 3p configurations.
Both cell sizes have approx the same 0.040 ohm DCIR ,
So a 2p config of 21700 would have an effective DCIR of 0.020 ohm (in theory) and a Vsag of 0.4 v @ 20amps
And a 3p config of GAs would have an effective DCIR of 0.015ohm (in theory) and a Vsag of 0.3 v @ 20amps
Theory suggests there would be less voltage sag using the 3p of GAs ? :wink:

IR is difficult to measure accurately especially for high-drain cells, is your source saying both GA and 48G are 0.040 ohm reliable? HKJ data lists GA IR as 0.050 ohm, and 48G IR as 0.040 ohm, but I take that information with significant reservation and do not make decisions or comparisons based on that.

For example, (again, HKJ data because it is well organized, thorough, and respected):
Sanyo GA IR is listed as 0.50 ohm
Sony VTC5A IR is listed as 0.40 ohm.



but look at traces of 15, 20, and 30 A draw for the two cells.
GA with "0.050 ohm" IR is not capable of 20 or 30A, and I would argue not suitable for 15A without significant heating and damage to the cell in short order.
VTC5A is fully capable of 30A and we know it is one of the best (possibly THE best, currently) high-drain-high-capacity cell available - but it has IR of 0.040 ???

VTC5A IR is essentially the same as GA? I think not.

IMO IR as measured by our methods is not a reliable metric for these high performance cells.

PS: There are vendors on this site selling new GAs for $3 or less.
So your 14s, 3p , pack could cost $126 or less
That 48G pack could cost you $80+ more, and have more sag ,!

What is the source for $3 GA? That is very good, especially if it is a trustworthy source.

I chose a popular US source that sells both cells, so it could be compared apples-to-apples (same-same). Price for 48G from the first (trustworthy) Chinese supplier I checked is $4.75/cell, so again - more costly, but only 5%. Not 50%.

......but yes, it would be 84gms lighter !

Not sure if you are mocking me or not? :?

1. note: it's not 84g, it's 99g just for the cells
2. After 50% additional ni/cu mass for 3p instead of 2p (ok, make it 40% due to larger 21mm vs 18mm C-C distance), longer balance wires, potentially longer discharge wires (depending, for the pack shape I prefer the discharge wires can be shorter, but not always), and more massive wrapping, the difference will be closer to 200g.
3. 200g may not matter to you or most people on this forum who seems to build 50kg ebikes that are more similar to small electric mopeds than bicycles and consider a bike "light" at 25kg. But my ebike weighs under 30lbs (~13.5kg) without lights and it makes the bike so much more fun to ride. It feels almost like a real pedal bicycle, which is my goal. The way I accomplished this is by choosing lightweight components whenever possible. So if you say 200g doesn't matter - well it does matter to me, and possibly others.

 

[andyme]

very embarassing....i have broken half of the surrounding magnet....it is part of the 850C display. (don't worry, it does not show that it will be difficult to flash it...i had other special ideas that i better not share... :lol:

so my question: can the device still be working or is it impossible? if anyone knows enough about this kind of coil...it should be just the shield..so maybe I am in luck..

 

[casainho]

so my question: can the device still be working or is it impossible? if anyone knows enough about this kind of coil...

I think it may keep working but you need to try.

Try to power on the LCD while it still has the original firmware and see if it works.

I think that coild is part of the power circuit that has as input the battery voltage and then outputs 5V to the circuit. LCD will not work if that 5V is missing.

How did that happen?

 

[Ron Paul's Blimp]

Has anyone used https://www.electrifybike.com/store/p65/42_Tooth_Narrow%2FWide_CNC_7075_T6_Chainring_%2810mm_offset%29_for_TSDZ2.html ? Looking at my motor and the pictures I wonder if it'd interfere where the cables come out of the controller.

i solved potential chain line issues by mounting a double chainring.

I'm more asking if people who know about this kind of thing think it looks like the chain line I have now will be a problem.

if you are asking me: is it more difficult to pedal with mounted motor in off than without motor: yes. clearly. but that was already the case when i was still using the stock firmware

Yup, that's what I'm asking. I don't know why, but I thought that the motor wasn't supposed to add much if any pedaling resistance when off. I thought only direct-drive hubs did that.

 

[qwerkus]

In case anyone is wondering: I found why my tsdz2 is so noisy: it's the bearings before an after the blue gear. Does someone know a quality replacement for them ?

 

[sac02]

In case anyone is wondering: I found why my tsdz2 is so noisy: it's the bearings before an after the blue gear. Does someone know a quality replacement for them ?

99% chance they are standard size bearings. If you have an industrial supply place nearby (Grainger, McMaster, Fastenall, etc) just take them in with you and they can get you set up with high quality replacements. The markings on the bearings will denote their size, so you can use google if you prefer.

 

[sac02]

Does anyone make an offset chainring with higher tooth count, ideally 52t?

(pictured chainring is 42t)

 

[Ron Paul's Blimp]

In case anyone is wondering: I found why my tsdz2 is so noisy: it's the bearings before an after the blue gear. Does someone know a quality replacement for them ?

99% chance they are standard size bearings. If you have an industrial supply place nearby (Grainger, McMaster, Fastenall, etc) just take them in with you and they can get you set up with high quality replacements. The markings on the bearings will denote their size, so you can use google if you prefer.

https://www.mcmaster.com/ is great and with some good measurements you can usually find the exact thing you need.

 

[crun]

very embarassing....i have broken half of the surrounding magnet....

so my question: can the device still be working or is it impossible? if anyone knows enough about this kind of coil...it should be just the shield..so maybe I am in luck..

20181211_171239~01.jpg

Glue the core back together with superglue or epoxy.
The inside part is the most important part of the inductance, so it will probably still be fine.

 

[andyme]

so my question: can the device still be working or is it impossible? if anyone knows enough about this kind of coil...

I think it may keep working but you need to try.

Try to power on the LCD while it still has the original firmware and see if it works.

I think that coild is part of the power circuit that has as input the battery voltage and then outputs 5V to the circuit. LCD will not work if that 5V is missing.

How did that happen?

is there a simple way to power it on without connecting it to my bike? because the connection between lcd3 and motor is not a plug but soldered wires. so i will only change it when it is time for change

how it happened is embarassing and childish, but ok, i can explain it: i have a strong cold, so i need to sit at home. i like the case of the DPC 18 much better than the one of the 850C and since i have no real use for the DPC 18 and cannot sell it, i decided to transplant the 850C into the body of the DPC 18...i know: it is ridiculous and childish...but well... and then there was not enough space in the case behind the magnet, and while trying to find out where and why and what, the magnet broke. i really did not push very hard, but it was quite fragile...and it broke into pieces, so i cannot glue it back on. i hope it will work nevertheless. worst case i think i should be able to get a replacement inductance and change the ring...i would guess...anyhow: sorry for not being serious and doing childish stuff. but you wanted to know although i had said i did not want to talk about it.... :lol:

 

[casainho]

sorry for not being serious and doing childish stuff. but you wanted to know although i had said i did not want to talk about it.... :lol:

I was worried that It could happen when opening the LCD, as users will have to open it to program. Good to know it was not at opening.

 

[andyme]

sorry for not being serious and doing childish stuff. but you wanted to know although i had said i did not want to talk about it.... :lol:

I was worried that It could happen when opening the LCD, as users will have to open it to program. Good to know it was not at opening.

that is why i had mentionned right away that it had nothing to do with the flashing process...

so how can i power it up?

 

[casainho]

so how can i power it up?

Connecting to a battery, see here the wires (just like LCD3): https://github.com/OpenSource-EBike-firmware/TSDZ2_wiki/wiki/Wire-KT-LCD3-to-TSDZ2

I added to the table at the end of page the 850C wires.

 

[andyme]

so how can i power it up?

Connecting to a battery, see here the wires (just like LCD3): https://github.com/OpenSource-EBike-firmware/TSDZ2_wiki/wiki/Wire-KT-LCD3-to-TSDZ2

I added to the table at the end of page the 850C wires.

:thumb: :wink: it will get a beautiful coverglass and i will love it :lol:

i am ready including connector for the st-link.

so whenever you are :wink: .....

 

[andyme]

Does anyone make an offset chainring with higher tooth count, ideally 52t?

(pictured chainring is 42t)


s263235063426757404_p65_i2_w1024.jpeg

could this maybe also be solved with some spacers?