New "TSDZ2 Torque Sensor Central Motor"

[gfmoore]

At boot up I get this :

1.JPG

Then, after 50 seconds I get this one :

2.jpg

I had this same error a few days ago. Are you using the brake sensors? That is the ones with sensor and magnet? I used adhesive to more firmly fix the sensor to the hood on my sti levers (road bike). I must have moved it a fraction from the magnet. Hence the error. Once I moved the magnet to touch the sensor and switched on all was good.

Just a thought.

@woly, sorry seemed you had similar thought, missed it on reading.

 

[sysrq]

All modifications to reduce noise discussed in this forum are for users who are using their TSDZ2 engine to its maximum. And it is quite correct.

I have the TSDZ2 myself, and at low power (100-250w), it is definitely audible (not particularly bothersome though). I have also briefly tested two different bafang motors, and both of those were waay quieter at similar power levels.

I've regreased my motor (with SHC100), which helped a little. I also tried the brass gear (when I was using it for more off-road), which was much noisier, so I went back to the nylon one (no issues so far with a few thousand km on low power, on paved roads). I think I just got a particularly noisy motor.

Isn't damping grease such as original NLGI1 high viscosity base oil Molykote em-50l meant for noise reduction? Strangely performancelinebearings.com recommends NLGI2 low viscosity, low friction Molykote pg-75 for some reason.

 

[sysrq]

Hi all,

I would like to share with the Forum what I did on the TSDZ2 engine to improve heat dissipation.

Part 3

3 - Lubricate the engine with a grease suitable for nylon Blue Gear.
I lubricated with a grease that is used to lubricate the nylon gears in the wind turbines. I used the Molykote PG-65 grease.

I applied Molykote both on Blue Gear and internally on the bearings as well as on all metal gears.
The engine was running much quieter.

Below are some pictures.

Nice work and pictures. Where did you buy Molykote PG-65 grease?
edit: I heard Molykote PG-54 plastilip grease could be fine.

Hi Dameri,

Thank you for your words.

I bought the Moly 65 in Portugal.

In Belgium, on the website, www.klium.com, you can buy, 1kg, for € 58.2 plus shipping costs (€ 20 for Portugal)).

They also have the PG-54 but it costs € 140.6 KG.

I sent the photos of the Blu gear and the engine to Molykote and the European technicians advised me on the PG 65.

As I said before, it is used in Nylon gears, in wind turbines.

I am not a grease specialist.

But what I found out the PG-65 has an NLGI Grade = 1 and it seems to me that it is better than the PG 54 which has an NLI Grade = 2-3.

If you check the grease that comes on the engine, on Blue Gear, when you buy it, it should have an NLGI = 1.

Regards

Hard to say why performancelinebearings.com recommends using Molykote pg-75 since the original one is Molykote em-50l NLGI1 with base oil viscosity over 1000 for noise reduction and damping.
The data sheet doesn't say what base oil viscosity is for Molykote pg-65. Normally for high speed applications it should be lower so that it has enough time to flow back where it should be.

 

[lafuente]

...Removing the thermal sensor (LM35) was the thing to do.......

:thumb: Glad it is solved now

I'm happy that the motor have no problem (still not tried it), but I need to figure what is the problem with thermal sensor as I mount it like explained on OSF github...

Today I measured voltages at my sensor connector and found 4,37V between +VCC and ground (white wire and black wire) where it should be 5V. I also found 3.5V between +VCC and throttle wire (orange wire).

Is it possible that using brake sensors can perturb the sensor ?

I tried brake sensors and they operate well because if I use them at power-up, the screen goes to "brake sensor error" as it should.

Thank you.

 

[lafuente]

I checked my sensor plug connexions like this :
- with an ohmmeter, I plug one lead at the sensor connector, the other at the 8 pin connector and check continuity between them to assure that I'm correctly connected. I need to check :
- 5V pin
- GND pin
- Throttle pin.

And they are correct. The 5v wire of the sensor goes to the 5v pin of the 8 pin connector pin (pin 4), the GND wire of the sensor goes to the GND pin of the 8 pin connector (center pin) and the Sensor Vout wire of the sensor connector goes to the Throttle pin of the 8 pin connector (pin6)




Does someone know what's wrong ?

Thank you.

 

[andrea_104kg]

I made one last try, a little desperate. The first controller, broken after less than an hour, (torque sensor not detected) had the pas sensor working, disassembled and tested. The second new controller, the torque sensor also did not work, but from tests made, the fault was of the pas that did not detect the cadence. I took the working pas sensor of the first control unit, I cut the wires and put it in place of the non-working one in the second control unit. I only did a few meters, with the engine open, but incredibly, IT WORKS !! I still have to do some tests and flash the right version of the software, let's hope something doesn't get stuck, but at least it's encouraging! In practice, due to an unfortunate series of circumstances, two new controllers did not work, luckily with different defects. I have to thank mbrusa who made me understand what was wrong, with the assistance of her and Elinx who prompted me to try the hall sensor, thanks!
I use 20.01b firmware, a little problem i must disable the odometer correction , or i do not see the speed, it's a bug?

 

[gfmoore]

Way to go andrea_104kg. I'm very happy for you. You didn't give up.

 

[gfmoore]

9.jpg

Does someone know what's wrong ?

Thank you.

Hi,
So what are you connecting to that gives 2.7 ohms?

 

[Elinx]

.... I took the working pas sensor of the first control unit, I cut the wires and put it in place of the non-working one in the second control unit......, IT WORKS !! I.....
I use 20.01b firmware, a little problem i must disable the odometer correction , or i do not see the speed, it's a bug?

congratulations for your perseverance. I'm glad something does work now for you.

So, just to be sure, the other PAS sensor is really defect.
Now it is cutted away, you can measure that too, with a multimeter.
You understand that the low sensitivity of the (build in) torquesensor isn't ideal.
It could be that something was wrong with the discs when mounting them.
So you have to do some more homework

The odometer compensation for stock display like vlcd5, is a count down. Because your display with startup, monitors the capacity and voltage of the battery.
This is done with the speedvalue for a certain time.
Full capacity is 99.9 and Voltage 41,5 for 5 sec does mean for the display, you have driven with 99.9 km/h for 5 sec and 41,5 km/h for 5 sec. This has influence on the distance value (odo).
For compensation you must riding a certain time till that wrong distance is compensated.
The same thing happens if you monitor your data values (power/temperature/cadence etc) when biking

At odo compensation time you see 0.00 km/h on display
If your speed is 25 km/h : you need 99.9/25*5 = 20sec and 41,5/25*5= 8sec is a total of 28 seconds with no speed visible
If your speed is only 10 km/h: 99.9/10*5= 50sec and 41,5/10*5= 21sec is a total of 71 seconds no speed visible

NB.This calculation is for you to imagine what is happening.
In reality, the time will be slightly shorter, because the distance you travel while compensating ODO also counts.

So if you want to monitor the max. of 6 values when biking for 5sec. each, you have to compensate a lot of the ODO distance.
All that time you see 0.00km/h

 

[blowhole]

I made one last try, a little desperate. The first controller, broken after less than an hour, (torque sensor not detected) had the pas sensor working, disassembled and tested. The second new controller, the torque sensor also did not work, but from tests made, the fault was of the pas that did not detect the cadence. I took the working pas sensor of the first control unit, I cut the wires and put it in place of the non-working one in the second control unit. I only did a few meters, with the engine open, but incredibly, IT WORKS !!

Great news! That's encouraging! May I ask when you say the "torque sensor not detected" and "torque sensor also did not work" what were the numeric values shown on the display? Zero, like in my case?

 

[Elinx]

.... May I ask when you say the "torque sensor not detected" and "torque sensor also did not work" what were the numeric values shown on the display? Zero, like in my case?

It is not complete clear that the torque sensor or controller or PAS sensor is the fault for the non working tsdz2.
I say this because andrea_104kg has changed the torque sensor and controller the same time and both configurations didn't work. That's why he thought it was the same defect

But if one of the sensors (speed/torque/pas) isn't working right, the tsdz2 has the same behaviour >>> it does not respond

Why? That is the real question. It could be the PAS sensors, it could be the controller. It could be the torquesensor, it could be everything.
That is why andrea_104kg was desparate.

andrea_104kg has installed OSF. In that case he can't use the hidden menu anymore.
But must measure the torquevalues another way.
With help of mbrusa he did this (and other values too) with the last torquesensor, but replaced the first one right away.

 

[blowhole]

andrea_104kg has installed OSF. In that case he can't use the hidden menu anymore.
But must measure the torquevalues another way.
With help of mbrusa he did this (and other values too) with the last torquesensor, but replaced the first one right away.

Ah I found his original post where he says with the replaced controller and torque sensor his readings were 110/140. Well it might help me if he remembers checking the readings for the original controller and/or torque sensor too.

Back to original firmware, walk works, throttle works, torsiometer doesn't work! In the vlcd5 menu, where it shows the force on the torsiometer the value is 110, if I press it reaches 140, so the torsometer works but the engine does not start! what can it be?

 

[Elinx]

..... torque sensor his readings were 110/140. Well it might help me if he remembers checking the readings for the original controller and/or torque sensor too.....

imho he never had measured this the right way,
Because he had OSF flashed, the hidden menu didn't work and gave zeros ofcourse
He replaced the components right away.
With the second set he had some assistence of mbrusa.

 

[Manbeer]

My TSDZ2 was out of commission for about a week due to me wanting to switch from a 6 wire to 8 wire controller and a few hiccups so i have been messing around with the bbs02

the first thing i thought when i hopped back on the tsdz2 was how weak it felt, but the next thing i thought was what a great BIKE it felt like. Just smooth, seamless assist.

The other thing i noticed is for a similar level of speed and effort, i get about 20% better efficiency from the tsdz2 which really adds up- roughly 8.6 vs 10.6 wh/mi at around a 17mph casual pace and none of that odd bafang pas behavior that makes you scratch your head

 

[lafuente]

9.jpg

Does someone know what's wrong ?

Thank you.

Hi,
So what are you connecting to that gives 2.7 ohms?

My poor ohmmeter

Anyway I found my fault. I miss-wire the sensor part. I used this one :

https://www.digikey.fr/product-detail/fr/texas-instruments/LM35DZ-LFT1/296-35151-1-ND/3640792

Is this correct ?
Thank you.
Manu

 

[sysrq]

Does anyone know why some motors have deeper alloy enclosure around the secondary gear while others are flat. Is there a a new version? I guess flat ones make easier time when it comes to degreasing and regreasing the secondary gear without removing it.

 

[feketehegyi]

Does anyone know why some motors have deeper alloy enclosure around the secondary gear while others are flat. Is there a a new version? I guess flat ones make easier time when it comes to degreasing and regreasing the secondary gear without removing it.

I'm not sure if it is related but I needed to replace the controller recently. I've had the 6 pin one as old but received the 8 pin variant as new. It was really hard to let the big 8 pin connector through the hole between the two sides. I was thinking on if the 8 pin variant has maybe a slightly different house then.

 

[Waynemarlow]

The 8 Pin connector has always been tight to get through, you have to sort of wiggle and push it part way then angle wiggle and push the rest of the way.

 

[sysrq]

Hi all,

I would like to share with the Forum what I did on the TSDZ2 engine to improve heat dissipation.

Part 3

3 - Lubricate the engine with a grease suitable for nylon Blue Gear.
I lubricated with a grease that is used to lubricate the nylon gears in the wind turbines. I used the Molykote PG-65 grease.

I applied Molykote both on Blue Gear and internally on the bearings as well as on all metal gears.
The engine was running much quieter.

Below are some pictures.

Nice work and pictures. Where did you buy Molykote PG-65 grease?
edit: I heard Molykote PG-54 plastilip grease could be fine.

Hi Dameri,

Thank you for your words.

I bought the Moly 65 in Portugal.

In Belgium, on the website, www.klium.com, you can buy, 1kg, for € 58.2 plus shipping costs (€ 20 for Portugal)).

They also have the PG-54 but it costs € 140.6 KG.

I sent the photos of the Blu gear and the engine to Molykote and the European technicians advised me on the PG 65.

As I said before, it is used in Nylon gears, in wind turbines.

I am not a grease specialist.

But what I found out the PG-65 has an NLGI Grade = 1 and it seems to me that it is better than the PG 54 which has an NLI Grade = 2-3.

If you check the grease that comes on the engine, on Blue Gear, when you buy it, it should have an NLGI = 1.

Regards

This page says Molykote pg-65 is NLGI2 with base oil viscosity of 17 which might seem like it wouldn't reduce any noise since the original Molykote Em-50l with base oil viscosity of 1050 works as a damping grease. While for some reason data sheet says Molykote pg-65 is NLGI1.

 

[sysrq]

How do people normally remove the internal brushless motor? I have seen three examples. With a wooden block and a rubber mallet against the edge of the motor, with a rubber mallet against the outer casing and pulling it with hands. Pulling with hands looks like the best option but mine seems to be stuck.
Couldn't find any other examples on the internet.

 

[sysrq]

Another problem I got today while riding as usual, small trip of 5 kms around home. Suddenly, I could hear the motor working and see on display the motor ERPS speed at max values of 525 although no help assisting to pull the bicycle...

The motor shaft broke!!! Does anyone knows why did this happen?? -- then I exchanged for a new motor and seems to work fine again on a short ride but tomorrow I will do a long trip during the full day with my wife... I hope the motor shaft will not brake again.......

yes, I broke an engine in a similar way :-( Simply, in one of my 2 tsdz2 the engine did not want to disassemble and I beat it lightly with a hammer. I didn't notice anything but once reassembled, after 1km the engine broke exactly as in your photo.
I think it is a fragile alloy that does not allow the slightest decline.

Same here, might have hit it with a hammer and a wooden block a bit too hard compared to jblat video.
Couldn't notice any vibrations only the current in walk assist mode is slightly lower around 0.12Amps (was steady 0.14Amps previously) and has some fluctuations (0.11-0.12Amps). There are also some periodic fluctuations in noise profile. The vibrations seem to be less after regreasing than with the motor intact. Will see if it brakes at some point.
Shame there aren't any recommendations for motor puller in the original service manual.

 

[sysrq]

Another problem I got today while riding as usual, small trip of 5 kms around home. Suddenly, I could hear the motor working and see on display the motor ERPS speed at max values of 525 although no help assisting to pull the bicycle...

The motor shaft broke!!! Does anyone knows why did this happen?? -- then I exchanged for a new motor and seems to work fine again on a short ride but tomorrow I will do a long trip during the full day with my wife... I hope the motor shaft will not brake again.......

yes, I broke an engine in a similar way :-( Simply, in one of my 2 tsdz2 the engine did not want to disassemble and I beat it lightly with a hammer. I didn't notice anything but once reassembled, after 1km the engine broke exactly as in your photo.
I think it is a fragile alloy that does not allow the slightest decline.

Same here, might have hit it with a hammer and a wooden block a bit too hard compared to jblat video.
Couldn't notice any vibrations only the current in walk assist mode is slightly lower around 0.12Amps (was steady 0.14Amps previously) and has some fluctuations (0.11-0.12Amps). There are also some periodic fluctuations in noise profile. The vibrations seem to be less after regreasing than with the motor intact. Will see if it brakes at some point.
Shame there aren't any recommendations for motor puller in the original service manual.

After 40 mile ride still no signs of broken motor axle. The only difference is slightly more pronounced oscillating whirring sound. Hard to avoid any cognitive biases but before regreasing it was kinda "dry" grainy sounding.

 

[ebbsocalMTB]

hey folks. I have been MIA in this thread for a bit, but wanted to stop in and give an update on my santacruz chameleon tsdz2 build. I did all of the temp mods that one can do to this motor. STO shim plate, thermal paste on the motor flanges, eco-ebike putty to fill up the spaces between the motor and the casing.

I'm happy to say that this motor is humming along perfectly with the temp mods in place. It easily chugs along at assist level 6 @ 48-52v when the battery is fresh, averaging 200-250w. The temps hold at around 125F - 135F in 60 degree weather and up to 145F in hotter ambient temps. Only when I hammer it hard at 350-400w for extended periods does it start to approach the temp limit at 165F. Another thing to note, is that the motor is running much more efficiently and consuming way less power, which is obvious as not as much is getting sucked away by the heat. One last expected, but very welcome outcome is that the temp quickly comes back down when sitting at stoplights or when turning the assist level down.

Overall, I'm happy with what this build has achieved at this point. It hasn't been without some ups and downs with both the motor and the bike hardware. I blew up a rear hub when riding this bike in Breckenridge Colorado. Something happened with the rear 6 bolt flange and the rear rotor ripped right off in the middle of a ride. That left me with 2,000' of elevation to lose to get back to my condo, and only a front brake to do it with. That made that ride interesting for sure.

The high point of this build so far, was riding this thing in mammoth lakes ca at the mammoth bike park with my wife on her SC bullit. i didn't take it down anything gnarly, I had my DH bike on hand for that stuff... but it was able to easily cruise buffed green trails and was a hoot to ride.

Since every forum post is worthless without pics... here's a little photo dump.

I finally drilled the frame and removed the massive zip ties I was using. This setup is good enough to commit to... If I tire of it, I'll just sell the bike/setup as a whole from this point forward.

Here's two pictures of the hub destruction when the 6 bolt flanges failed and the rear rotor tore off the bike mid ride. Definitely a first for me and a pretty startling moment. Friends don't let friends use Novatech hubs. I knew this before, and had to re-learn it I guess.

Had a great weekend in mammoth riding my enduro rig, my wife on her SC bullit and me on my converted Cham-E-Leon. It's a giant Santacruz Fanboi party at my house.

Me making bad life choices.

This drop ends up at about 10' out and 15' down from take off to landing. You end up landing just about at the end of that first patch of shade on the runout. By the time you land, your forward momentum is 100% converted to downward... and you accelerate out of the runout at mach chicken... shooting for a 6' gap between two trees on either side. Sketchy, but exhilarating.

 

[lafuente]

Hello,

I finally got my motor work. The power (vcc, gnd) wire of thermal sensor was simply inverted :-(
On the datasheet of the LM35, the TO92 package pins are show from "bottom", which I didn't notice at the first read.

Anyway, even if the motor is a good thing, it lack of power (500W/48V version). I'd like to know if :

• anyone tried the difference between 500W and 750W ?
• is this difference is mechanical, or only software ?

Thank you,
Manu

 

[Elinx]

....difference between 500W and 750W ?
... mechanical, or only software ?

They are equal.

The lack of power could have to do with an unsensitive torque sensor.
Software calibration could help a bit. But if it is very unsensitive, a hardware calibration does a better job.