New "TSDZ2 Torque Sensor Central Motor"

In reverse thinking with the motor off. The drive side pedals are connected directly through the torque sensor and then on to the non side by a shaft, which has 1 roller bearing on either side and is separated from the motor main gear by the sprag clutch.

The only drag we should get is the grease and roller bearing friction within the bearings and the two water seals along that main shaft. In practice though the springs holding the sprag clutch activating mechanism against the inner race, which is connected directly to the main shaft and the large volume of grease within that quite large diameter sprag clutch, is enough to ever slowly slightly rotate the main gear which can be felt at the pedals as a back ground notchiness ( main gear teeth rotating the blue gear ) the faster you pedal.
 
Doohickey said:
Maybe my TSDZ2 is somewhat noisy despite being a newer one with helical gears, but I've tested / cycled beside a few different BBS02s, and they are all much quieter.

A little noise is expected but this seems a bit extreme. At 10 feet away its in the 52dB range. Rider needs earplugs.

Here is a video; the noise you hear is coming from the TSDZ2 mounted on the trike ahead.

http://www.triketech.com/Drivetrain/PowerAssist/TSDZ2/TSDZ2_Bryce_Up.mp4
 
I got my TSDZ2 kit from Eco-ebike built up on my Santacruz Chameleon. I have a build thread at the following link in case you want to check out more details. https://endless-sphere.com/forums/viewtopic.php?f=6&t=111223

I have a few issues I could use people's help with:
#1 - Temp threshold and overheating - I am still figuring the motor out and how it best likes to be utilized, but I was quite surprised how quickly I can hit thermal throttling pedaling on flat ground... it's especially bad if I'm traveling on an 1-2% incline steadily. I rode the bike to work this morning, and within about 4 miles of 1% grade, I hit the thermal limit and the temp sensor was dialing the power back. By the time I got to work, I was easily into 180 degrees. I basically pedaled the last 1.2 mile with the assist level on 2. It seems that the motor persistently gets hotter and hotter even if I'm humming along letting it pull 6-10amps and will hit 170-180 degrees fast. When this happens, I'm averaging about 20mph and going back and forth between 18-25mph. Granted, this is with 29x2.6 tires, so there is some rolling resistance that has to factor in. I knew there would be some thermal limitations to keep an eye on, but I guess I didn't expect to be running into it maintaining 20mph on flat ground. Ironically, on my first ride out on this thing yesterday, I did a few climbs that were 8-10% grade for a half mile and ran into no thermal issues.

#2 - Battery voltage vs battery meter & Safe cut-off. I am using a hailong shark based battery from pswpower. The cutoff voltage per their site is listed at 36.4. I have it setup in the firmware to cut off at 39v. I did a ride this last weekend that was 22 miles, 1,500' of climbing and the battery started at 54.4 when I started and was showing 47.8v more or less when I finished. The graphical battery indicator was showing completely empty and red at the 47.X volt point. I assume this is just a display issue and it's fine to run it down to 39v? I don't have the resistance set yet for the battery, in case that factors into this.

I ordered a heatsink plate from the silent thunder ordinance and some arctic silver mx4 so that I can hopefully improve on the thermal throttling that I'm experiencing. Other then the thermal throttling, I am really, really happy with this thing.

Photos from my first ride out.
MjE7jMR.jpg


Motor mount using a bafang bracket
wW5NMA7.jpg


Pretty happy with how clean the setup looks
k2kLpDf.jpg
 
Alright... starting to transition from content with this motor to thinking I need to move it down the road quickly. Each ride, I keep trying to change how I ride the thing to keep the heat down, but it’s seemingly futile. On my ride home from work, 3 miles in, the motor was already 170+ and quickly approaching 180. Sometimes it seems to be like an out of control Diesel engine where it just keeps getting hotter and hotter no matter what I do... including turning the assist off and just pedaling. Then other times it will seem to lock in at 173-174 degrees and sit there no matter how much amperage it’s pulling or how much effort I’m putting in. I’ve tried riding it at assist level 3-20, and that doesn’t seem to play a huge factor.

So... what’s the most efficient way to ride this thing on flat ground? What amperage and engine RPM does this thing like to sit at to stay as cool as possible?

Here is a photo I took of the display next to my garmin info screen once it tapped out while on my way home from work.
NmONRBH.jpg


Right now, I’m going to give this thing a few more tries and then remove it from the bike. Quite frankly, it’s slowing me down as I can actually maintain 20-22mph on a road bike for. 60 minutes straight without much issue, I just end up sweaty as hell. This thing overheats about 15 minutes or 3 miles into any 20-25mph sustained section and then I have to shut it down and for the next 15 minutes I’m riding at 15mph. Ultimately, right now, if the trip is longer then 5 miles, I’m faster without the motor.

Is this just an ignorance is bliss thing because most people don’t have the temp sensor wired in? What gives? If this is normal for this motor I would expect this thread to be absolutely full of complaints and people that take these things off immediately.

The irony is that I bought this thing to mostly help with commuting... which so far, it sucks hard at. It’s better at doing the thing I don’t need help with and that’s riding steep dirt climbs. :confused:
 
Weird, fwiw I don't have a temp sensor installed but have not had any issues with thermal throttling.
Granted, here its been low 60's on a nice day and i usually ride in the 3rd assist level. Plus, I am on 26x2.1's that weigh about 550g so less rotational mass. I weigh right around 160
With my setup being cadence limited to around 90 rpm cause i am on stock firmware and moderate pedaling, on the road, on a decent day it'll just crack 20mph overall for the ride. Yeah, i could pedal close to that but not on an MTB with risers for too long. Though a few years back when i was in better shape i did average 17.8 for a century ride on a MTB with 2.2" semi slicks. at 20mph avg the motor is barely warm to the touch. Flogging it off road, same thing. I am wondering if the rolling resistance of those tires is giving it a hard time. at the same cadence you wpuld probably be doing a good 5mph faster i would imagine.
IIrc ECO programs their controllers to 15a max whereas stock was 13 or something. Perhaps reducing the max current to the stock 13 would be a good place to start
 
ebbsocalMTB said:
Sometimes it seems to be like an out of control Diesel engine where it just keeps getting hotter and hotter no matter what I do... including turning the assist off and just pedaling.

Surely this means you have a fault in the temperature sensing?
 
You bring up good points. I have been watching the current draw on the display to get an idea of the load on the system and that may not be the end all sign. These tires are definitely not conducive to high speed. I assumed that since I could manage 12-15mph (dirt to pavement) on these same tires without a motor that I should be able to average 20ish with the motor. That may have been optimistic... I’m 175lbs and can hammer the pedals pretty hard, but I’m 6’2” So I make for some pretty crummy aerodynamics. This thing also has 780mm handlebars.

I think I’ll turn down the amperage and see if that lowers overall temps. I’ve got it set at 15 right now, and just keeping an eye on it so that it stays mostly a round 10amps. I think I’ll lower it to 13a and see if that helps.
 
Blacklite said:
ebbsocalMTB said:
Sometimes it seems to be like an out of control Diesel engine where it just keeps getting hotter and hotter no matter what I do... including turning the assist off and just pedaling.

Surely this means you have a fault in the temperature sensing?

Yeah, I honestly wasn’t sure what to think. My most recent applicable experience is with brushless motors on rc cars. Those seem to get to a point where if you are outside of the gearing window and causing current to spike, they start to just get out of control and need to be shutoff. Coasting and staying off the throttle isn’t enough to cool them down. From that experience I didn’t think much of the fact that after I shut it down, or coasted, it kept climbing in temp for a few degrees.

I guess it could be an issue with the temp sensor... but I don’t think shutting it off and turning it back on significantly changed the temp sensor reading.
 
ebbsocalMTB said:
Right now, I’m going to give this thing a few more tries and then remove it from the bike. Quite frankly, it’s slowing me down as I can actually maintain 20-22mph on a road bike for. 60 minutes straight without much issue, I just end up sweaty as hell. This thing overheats about 15 minutes or 3 miles into any 20-25mph sustained section and then I have to shut it down and for the next 15 minutes I’m riding at 15mph. Ultimately, right now, if the trip is longer then 5 miles, I’m faster without the motor.

IMO, this motor is made for 250w, 25kph limit, and was not designed to do more than that, despite being sold as 750w in some cases (in which case it's still the same hardware, just the stock firmware is programmed different).

In my experience, If I'd try to go 30 - 40kph on a flat, (which for me would mean about 300 - 400w of motor power), it will heat and reduce power (even more so on any kind of uphill). The various thermal mods (thermal paste and pads), help with this to some extent, but it will still overheat. It also heats faster when outside temperature is higher (30c in summer..)

the temperature increasing a few degrees after you stop pedaling is to be expected, there is some lag as the heat gets to the sensor. Obviously if the temperature increases significantly or doesn't go back down once you stop pedaling, yes, I'd think the sensor could be faulty.

On opensource firmware there is a way to limit the power (watts) that you can enable.
 
Doohickey said:
ebbsocalMTB said:
Right now, I’m going to give this thing a few more tries and then remove it from the bike. Quite frankly, it’s slowing me down as I can actually maintain 20-22mph on a road bike for. 60 minutes straight without much issue, I just end up sweaty as hell. This thing overheats about 15 minutes or 3 miles into any 20-25mph sustained section and then I have to shut it down and for the next 15 minutes I’m riding at 15mph. Ultimately, right now, if the trip is longer then 5 miles, I’m faster without the motor.

IMO, this motor is made for 250w, 25kph limit, and was not designed to do more than that, despite being sold as 750w in some cases (in which case it's still the same hardware, just the stock firmware is programmed different).

In my experience, If I'd try to go 30 - 40kph on a flat, (which for me would mean about 300 - 400w of motor power), it will heat and reduce power (even more so on any kind of uphill). The various thermal mods (thermal paste and pads), help with this to some extent, but it will still overheat. It also heats faster when outside temperature is higher (30c in summer..)

the temperature increasing a few degrees after you stop pedaling is to be expected, there is some lag as the heat gets to the sensor. Obviously if the temperature increases significantly or doesn't go back down once you stop pedaling, yes, I'd think the sensor could be faulty.

On opensource firmware there is a way to limit the power (watts) that you can enable.
I see my motor giving 1000W or even a bit more when my battery is fully charged. But yes, it will heat fast. But the idea is to have the Startup Boost, providing the max power like the 1000W at startup or other specific moments - you choose. And sure, this only with our OpenSource firmware :)
So, this is effectively at least a 750W motor, because it really provides that power and makes a huge difference on the torque it assist you on that moments: 1000W versus 250W at startup!!
 
Thanks for the feedback. I guess I thought I was being nicer on the motor then I was. I was watching the amperage draw and trying to keep it under 10-12amps of current, but the motor is vacillating between 300-400w at that point and that's when the temps start climbing. I guess I'll add the motor air gap plate, and a bit of thermal material and see where I'm at after that. I'm hoping that with a mixture of curbing my expectations and adding some thermal capacity, that I can meet in the middle and make this work.

I think I'll reduce the overall current limit to around 13amps, and play around with the startup boost feature to give it a bit of that exciting feeling as you get rolling.

A few other things things.
1) I believe that the graphical battery reporting was down to the watt/hours capacity setting and the battery resistance setting. I tweaked both of those in the firmware last night, not sure if it fixed it yet, but they were both off by a little bit.
2) Motor Voltage setting and higher cadence. I am running the system as a 48v system. Do I need to have "exp 48v" as the setting for this variable in order to enable the higher cadence? I guess the question is, can some one clarify the difference between 48v and exp 48v?

Just as a general point of interest. Here's the same ride (which is my current commute) on a standard bike vs an ebike. This is a ride that I do at about 70% effort on a pedal bike so that I don't arrive to work drenched. This is running with the motor rather un-corked and with a 15amp limit set. Note the graph is speed with elevation in the background.

Muscle bike with cross country racing tires. 500g, small block knob tire, Bontrager XR2 in back, XR3 in back in 29x2.4.
1GYko7O.jpg


Same exact route on the Cham-E-Leon, as pictured above in this thread. The motor thermally shut down at mile 3.99, 13 minutes into the ride.
R0bn8oe.jpg


NmONRBH.jpg
 
ebbsocalMTB said:
Blacklite said:
ebbsocalMTB said:
Sometimes it seems to be like an out of control Diesel engine where it just keeps getting hotter and hotter no matter what I do... including turning the assist off and just pedaling.

Surely this means you have a fault in the temperature sensing?

Yeah, I honestly wasn’t sure what to think. My most recent applicable experience is with brushless motors on rc cars. Those seem to get to a point where if you are outside of the gearing window and causing current to spike, they start to just get out of control and need to be shutoff. Coasting and staying off the throttle isn’t enough to cool them down. From that experience I didn’t think much of the fact that after I shut it down, or coasted, it kept climbing in temp for a few degrees.

I guess it could be an issue with the temp sensor... but I don’t think shutting it off and turning it back on significantly changed the temp sensor reading.

How hot is the motor casing? Looks to me like an issue with the temp sensor - from what you describe - and the not too hot ambient temps (17c from the garmin image) - i can't see why it would be overheating so much... I'm 105kg riding a 25kg bike up hills and such - admittedly in the UK but 17c isn't unheard of and I rarely get the motor that warm...
 
I'm planning to fit max temperature strips https://www.testo.com/en-UK/products/temperature-measuring-strips to all the motors I have apart in future just to see what max temps these motors sustain.

We push our motors much harder than ebbsocalMTB using a lot more current and do a lot more climbing with in my case nearly 100kg on board and to date we haven't yet had any over heated motors.

I started to do some research into what electric motors are rated at as far as temperature goes and there are appears to be 3 different ratings ranging from 135 degrees C to the higher level of 185 degrees C on some. Does anybody have a contact at Tongsheng we could pass the question to as I think 100 degrees C is fairly conservative for a working electric motor.

A good read.
http://www.drivesandautomation.co.uk/useful-information/nema-insulation-classes/
 
ebbsocalMTB said:
Thanks for the feedback. I guess I thought I was being nicer on the motor then I was. I was watching the amperage draw and trying to keep it under 10-12amps of current, but the motor is vacillating between 300-400w

2) Motor Voltage setting and higher cadence. I am running the system as a 48v system. Do I need to have "exp 48v" as the setting for this variable in order to enable the higher cadence? I guess the question is, can some one clarify the difference between 48v and exp 48v?

Umm at 13amps x 52 volts ( you could have 54 volts with a fully charged battery ) = 676 Watts

I thought the Exp 48 V settings had been removed in the 20.1C version and with the use of Field Weakening that setting was only in the original 20.1 version ?
 
Waynemarlow said:
I'm planning to fit max temperature strips https://www.testo.com/en-UK/products/temperature-measuring-strips to all the motors I have apart in future just to see what max temps these motors sustain.

I started to do some research into what electric motors are rated at as far as temperature goes and there are appears to be 3 different ratings ranging from 135 degrees C to the higher level of 185 degrees C on some. Does anybody have a contact at Tongsheng we could pass the question to as I think 100 degrees C is fairly conservative for a working electric motor.

A good read.
http://www.drivesandautomation.co.uk/useful-information/nema-insulation-classes/

Figuring out the maximun temperature a magnet can withstand in a motor is complex and depends on more parameters than just their maximum working temperature : their shape and positioning are important factors as well.
Another good read :
https://www.supermagnete.de/eng/faq/What-temperatures-can-magnets-withstand

Given the complexity of the matter (since just knowing the max working temperature of the magnets in the motor would not be enough), going the empirical way with max temperature strips as you plan to do seems a good option.
I feel it will be difficult to figure out exactly when the magnets begin to be damaged with temperature, but I'm looking forward to hearing about your results !
 
Hi, interested in mounting a TSDZ2 to my kids 20" to help extend range on our cycling adventures. I've seen some differences in TSDZ2 versions - helical gears, different axles, etc

Is there a specific way to check if a motor has the latest changes prior to purchase? Or have those changes been available for quite some time and all stock in the market should be current design?

FYI it is 36v 250w that I'll be purchasing (only powering a 10kg bike with 28kg rider)
 
Major Clod said:
Hi, interested in mounting a TSDZ2 to my kids 20" to help extend range on our cycling adventures. I've seen some differences in TSDZ2 versions - helical gears, different axles, etc

Is there a specific way to check if a motor has the latest changes prior to purchase? Or have those changes been available for quite some time and all stock in the market should be current design?

FYI it is 36v 250w that I'll be purchasing (only powering a 10kg bike with 28kg rider)
See the wiki, with pictures of my son 24 inches wheel EBike with TSDZ2.

You should install our OpenSource firmware where you can limit the motor current ramp / acceleration, motor max current, motor max power and max wheel speed. With this you can make good limits to you son EBike!!
 
Ok... I started to get to the bottom of the temperature issues I was having. I think it was an interesting confluence of issues at the same time that lead to it. Death by 10000 cuts so to speak.

1) Mechanical Issues. When I converted the freehub of this wheelset from XD to shimano, I forgot to install the .5mm spacer that goes between the freehub bearing and the hub shell bearing. This was causing binding that was actually intermittent. When on the bike stand, everything felt pretty much fine. There was some additional resistance that I could feel, but I just assumed it was the motor. Well, once I put the bike on the ground and weighted the chassis, with the minimal axle flex in the rear hub, it would cause the free hub to bind on the hub shell which would cause split second spikes in resistance as my weight in the saddle would weight and unweight with bumps and my cadence. I also noticed that the adjustable dropouts were slightly out of spec from one side to the other, which would cause additional lateral loading, further compounding the above issue... especially when leaning the bike over to one particular side. Also... the tires I had on the bike seem to be good for about 50-75watts of resistance/effort at 20 mph. All in all, I'm guessing there was 200watts or so of just garbage resistance and impact to the system, with some spikes due to the hub binding.

2) Firmware setting issues that may have been playing off of the mechanical issues, that were then magnified by stupid issues. The ramp rate was set to 8amps per second in the firmware. I think the combination of the above split second increases in drag, combined with the ramp rate was causing additional load to the motor.

3) The final nail in the coffin... stupid. I was just generally approaching the entire thing wrong. I was basically riding this thing in such a way that was causing the issue to show up, almost on purpose. Once this issued first showed up, I was trying to find the break points of rider watt output vs motor output and draw/load on the system. I tried that with low assist levels and high rider output with both low and high cadences, and then with high assist so that I could isolate the rider input vs motor input. Well, to cut it short... I was being an idiot and not controlling variables and not paying attention to the right things. My tests were being done with different grades, headwinds and then I wasn't paying total attention to motor watts and was at times, too focused on amps. Voltage from the battery wasn't consistent... etc etc. Basically, idiot mode.

So... Where I'm at now. I've got the bike in a good place where at assist level 8, it balances out to 20mph +/-, while needing 100-200w of human power and 150-300w of motor power. Assist level 2 doesn't feel draggy any more, now that the resistance is out of the system. I curbed the overall max amp draw to 13amps and turned the amp per second gain rate down to 5amps. I'm going to get into a rhythm with the build/bike and then likely start playing around a bit with the boost feature to get a little bit of the zip back.

Regarding the hardware of the bike. I've got the following things on my radar:
1) Build in some more thermal capacity into the motor. I've got the silent thunder ordinance air gap plate en route. I've got some thermal interface material in route as well, including pads, thermal grease/paste and I've got some heatsink material to play with as well.
2) Change up the tire situation. I'm going to run a 27.5 x 2.8 rear tire with cushcore to get a bit better bump absorption. I'll also change out to a more trail bike rear tire like a maxxis rekon or bontrager xr3. I'll pair that with a 29x2.6 front tire with cushcore XC. That tire I will change out to be situational (I have extra front wheel/brake rotor setups sitting around) so that I can use a lower rolling resistance tire for commuting and less serious trail riding and then have the meaty stuff available for more serious off roading.
3) I need a new bike seat... ebikes certainly have different seat requirements, and this thing as a hardtail isn't helping.
4) Grips, I ordered up a set of ergon grips with a wing for some more hand/wrist support.

And some more tweaks and adjustments along the way. Thanks for the help and patience folks. More to come.
 

Here is a photo of the TSDZ2 heatsink mod ready to be assembled.
3"ODx1/8"wall tubing with a gap cut and beveled to clear the capacitor on the controller.
3mm thermal pads behind it. Thermal grease. Slip fit w/o the pads. Tight with them so they're compressed.
A block of wood and a mallet to press it together. Note thermal paste at bottom of ring also.
Bevel the top outer edge to help assembly.
This is a large image. if you open it in another tab you can zoom in.
 

Attachments

  • 001.JPG
    001.JPG
    261.4 KB · Views: 1,896
I'll just copy the original instructions for this here. I hope I don't get yelled at for double posting.
TSDZ2 heatsink part 1
I've taken tools in hand and started to make an aluminum heatsink for the motor.
It's actually turned out to be pretty simple to do.
I went to my local metal supplier and got a few pieces of 3"x 1/8" Aluminum tubing. For rough cuts they only guarantee
0 to+1/8" tolerance.They could be up to 1 3/16" but mine are 1 1/16" long.
Using a sabre saw I cut a slot in the tube, and then another about 1/4" away.This makes the part flexible, and allows for thermal expansion. If you use a square and mark your tube you can get a pretty straight cut. Decide which end of the tubing is cut the closet to being square and file or sand it flat. this will be the end that goes against the main casing. You need the cover off the motor for the next part.
Next the ends you just cut need to be beveled to clear the capacitor.Using a Dremel tool and a sanding drum cut one at a 30* angle, then cut the other at 60* to form a 90* notch for the capacitor. Don't try for a close fit. Everything will be working to close that gap when you assemble, and run the motor. Put a small bevel on all the top and bottom corners of the tube.Put a larger bevel on the top outer edge of the ring to help assembly when the thick thermal pads are installed. Sand any rough spots smooth.
Remove the wires from the motor, and passing the small wires through the slot slip the ring all the way down over the motor Test fit by slipping the cover over the aluminum ring. It should bolt down fairly easily.
I'm waiting for some 3mm thermal pads to arrive to go between the bolts on the motor and fill the gap between the armature and the aluminum ring. This will also require thermal grease between the cover and the ring for a wet joint, also around the bottom of the ring to send heat into the main casting( I said bottom because the motor is off the bike and on the workbench). This will help cool the part of the motor that is away from the cover.
I don't have, or plan to install a thermal sensor on my motor. So I won't have any numbers for you.
In metric this would be 76.2x3.15mm tubing. I'm guessing 75x3mm is what would be available. Since it's smaller it should work with just a little more thermal grease needed in places.
I've found that 76x3mm tubing is listed by suppliers, so that is what I would try to use.

TSDZ2 heatsink continued.
I got my 3mm thermal pads and installed the heatsink ring today. 100x100x3mm gives enough pad to do the job twice in case of a mistake, or blue gear repair later. I removed the motor to install thermal pads on the armature between the motor screws, and put thermal grease all the way around the base of the motor for heat transfer in that direction. Reinstalled the motor. The only special tool needed was pair of large external snap ring pliers to spread the ring to go over the thermal pads. This was expected. Compression helps with the heat transfer of the pads. Once the ring was over the pads, and seated around the bottom of the motor, the wiring was connected, and the cover could be installed. The ring will be slightly larger than before due to the thermal pads. A more generous bevel at the top outside edge of the ring would have helped. I coated about 1/2" down from the top of the ring, and about 1/2" inside the cover at the edge with thermal grease. You only need to do this 1/2 way around where the cover and the ring will be touching. Squeezing the ring in I got it started going into the cover. It helps to have the crank tube bracket attached as a guide.The cover was a much tighter fit than before and needed to be tapped down into position using a piece of wood to protect the cover.But I knew it was just the compressible pads making it tight.
I didn't use expensive computer overclocking thermal compound like Arctic Silver. I got a 20g. tube of Thermalgrease 20 which is just a generic cooler paste. Again I think there is enough for a rework if needed.
There's nothing here that couldn't be done with a hacksaw, and a rough file. Or even the razor saws they sell in craft stores.

This got started because I added a motor controller to get a throttle, and I had it flashed with the Open Source Firmware. But I did some measuring and saw the possibility of a decent heatsink solution and tried it. This has thermal pads all the way around the armature. The heatsink is full circle also, and the heat goes not only to the cover but down into the base of the motor housing.
I hope someone with a thermal sensor gives this a try and reports back.
 
casainho said:
See the wiki, with pictures of my son 24 inches wheel EBike with TSDZ2.

Thanks. Is there a way to tell the difference between newer TSDZ2 (helical gear, narrow axle) and old TSDZ2 without taking them apart?
 
Back in January there was some discussion on the sleeving of the motor for heat disappation on the thread https://endless-sphere.com/forums/viewtopic.php?f=28&t=102046&start=150 where we discussed the use of 76mm pipe.

Can I add a few further developments to this. If you take a 76mmx 3.2mm Ali pipe and cut it to 80mm in length. Mount it in the lathe and reduce the thickness to 2.2mm. At the outer end make a 8 degree chamfer cut back 15mm. Cut to a length of 38mm and now slit lengthways. Thats your outer sleeve as in the picture.

Now the remainder in the lathe, reduce to 2mm thick and cut it to 21mm in width. Now cut 6 sections 30mm long. These sections fit against the windings of motor parted by the long screws which hold the motor together. These don't quite follow the radius so I've been simply compressing them lightly in a vice to fit correctly.

These measurements are for a 36 volt motor, I'm not sure exactly on the 48 volt unit and you may have to ammend them slightly.

Now all you have to do is put heat transfer paste on the outer of the windings, press the small pieces on, recoat with grease on the top of pieces and around the top and bottom sections, spread the sleeve with a wide screwdriver blade turned slightly and the sleeve will now slide down over the motor.

Coat the section where the sleeve is closest to the case and also in the case itself and the outer case should now slide over the lot without any tapping with a mallet and also when you come to replace the blue gear or do any maintenance on the motor, the cover will come back off easily.

As a nicety, drill and tap a 3mm thread into the sleeve and you can mount your heat sensor directly to it.

We've been running a couple of motors this way all winter without problem and I had intended this weekend to fit a temp sensor and check to see what exactly is going n temp wise with one of them.
 

Attachments

  • motor sleeve.jpg
    motor sleeve.jpg
    125.7 KB · Views: 1,855
Major Clod said:
casainho said:
See the wiki, with pictures of my son 24 inches wheel EBike with TSDZ2.

Thanks. Is there a way to tell the difference between newer TSDZ2 (helical gear, narrow axle) and old TSDZ2 without taking them apart?

No and maybe. But the old motors are now probably 4 years old at least, the chance of getting one of them is probably remote.

Maybe also the later motors have 4 screws on the cover on the drive side and the earlier 3, can anyone confirm that.
 
Waynemarlow said:
Back in January there was some discussion on the sleeving of the motor for heat disappation on the thread https://endless-sphere.com/forums/viewtopic.php?f=28&t=102046&start=150 where we discussed the use of 76mm pipe.

Can I add a few further developments to this. If you take a 76mmx 3.2mm Ali pipe and cut it to 80mm in length. Mount it in the lathe and reduce the thickness to 2.2mm. At the outer end make a 8 degree chamfer cut back 15mm. Cut to a length of 38mm and now slit lengthways. Thats your outer sleeve as in the picture.

Now the remainder in the lathe, reduce to 2mm thick and cut it to 21mm in width. Now cut 6 sections 30mm long. These sections fit against the windings of motor parted by the long screws which hold the motor together. These don't quite follow the radius so I've been simply compressing them lightly in a vice to fit correctly.

These measurements are for a 36 volt motor, I'm not sure exactly on the 48 volt unit and you may have to ammend them slightly.

Now all you have to do is put heat transfer paste on the outer of the windings, press the small pieces on, recoat with grease on the top of pieces and around the top and bottom sections, spread the sleeve with a wide screwdriver blade turned slightly and the sleeve will now slide down over the motor.

Coat the section where the sleeve is closest to the case and also in the case itself and the outer case should now slide over the lot without any tapping with a mallet and also when you come to replace the blue gear or do any maintenance on the motor, the cover will come back off easily.

As a nicety, drill and tap a 3mm thread into the sleeve and you can mount your heat sensor directly to it.

We've been running a couple of motors this way all winter without problem and I had intended this weekend to fit a temp sensor and check to see what exactly is going n temp wise with one of them.

I'm glad to see some testing being done on motor sleeving. If you look at the date on the photo above you will see that method was posted last summer. But the computer containing the photo quit working so I was unable to add that until now.
Based on the success of just stuffing thermal pads 1/2 way around the motor i think sleeving should be much better.
There is no machine shop work in the one shown. Just the rough cut from the metal vendor, and some saw and file work. The slot cut in the tube, and the thermal pads makes it flexible enough to conform closely to the taper and diameter of the motor cover. It also allows room for thermal expansion. The only tool an amateur mechanic might not have are the large snap ring pliers to avoid damaging the soft thermal pads when installing the ring. It took longer to write the instructions than it did to make the ring.
The ring installs fairly easily without the thermal pds in place. it's just the compression of the pads that makes it necessary to tap it together. Since this improves their thermal performance I feel it's worthwhile. You could use 2.5mm pads for a looser fit. But you lose an advantage right where it's needed most.
Here is a chart showing thermal resistance of thermal pads under different pressures.It will be much higher when the pad is the same thickness as the gap it's filling. Clamping pressure is also a factor in the performance of thermal grease
http://www.sg-thermal.com/uploads/allimg/170601/1-1F601163016144.png
Thermal resistance is the opposite of thermal conductivity. Lower is better here.
A simple improvement to my design would be to use better quality thermal pads. I used 6W conductivity pads and 12-14W are available. The flexible pads allow a simpler and less precision method, but the design requires some pressure to work properly.
One option would be the better pads at the thinner size.
One way to assure easy removal would be to thread the 2 bolt holes in the cover at the motor end, and use jacking screws to press it back off.
 
Back
Top