Hardware temperature control tsdz2

Discussions related to motors other than hub motors.
This includes R/C motors, botttom bracket, roller and geared drives.
andrea_104kg   100 W

100 W
Posts: 152
Joined: Mar 05 2018 5:51pm

Hardware temperature control tsdz2

Post by andrea_104kg » Aug 26 2019 3:33pm

I open this tread separate from the main one to ONLY post engine HARDWARE temperature control solutions.
This is one of the major flaws of tsdz2, please post ideas and solutions only on this topic. :D

If it can be useful to summarize my experiences made to check the engine temperature.
Sorry I can't put the photos in the right place :-(
First I put a lcd themometer capable of measuring up to 110 ° celsius.
tsdz2ter1.jpg
tsdz2ter1.jpg (46.12 KiB) Viewed 1056 times
tsdz2ter1-5.jpg
tsdz2ter1-5.jpg (43.87 KiB) Viewed 1056 times
tsdz2term2.jpg
tsdz2term2.jpg (41 KiB) Viewed 1056 times
I took a test climb.
it is a non asphalted country road. The maximum slope is 17%.
salita test.JPG
salita test.JPG (10.17 KiB) Viewed 1056 times
The outside temperature was around 16 ° celsius., The engine (tsdz2 48v 750w) was at 37 ° at the start. max power oscillating from 450w to 800w.
the result
prova1-7senza ventola.jpg
prova1-7senza ventola.jpg (57.8 KiB) Viewed 1056 times
then I tried putting a fan
prova1-1.jpg
prova1-1.jpg (24.44 KiB) Viewed 1056 times
prova1-2.jpg
prova1-2.jpg (26.64 KiB) Viewed 1056 times
prova1-3.jpg
prova1-3.jpg (28.93 KiB) Viewed 1056 times
prova1-4.jpg
prova1-4.jpg (27.38 KiB) Viewed 1056 times
prova1-4.jpg
prova1-4.jpg (27.38 KiB) Viewed 1056 times
prova1-5.jpg
prova1-5.jpg (27.44 KiB) Viewed 1056 times
prova1-6.jpg
prova1-6.jpg (48.7 KiB) Viewed 1056 times
The result has unfortunately been unsatisfactory.
prova1-7temp.jpg
prova1-7temp.jpg (54.78 KiB) Viewed 1056 times
I knew that in the summer I could no longer use tsdz2, so I resorted to drastic remedies.
201962921545_alet1.jpg
201962921545_alet1.jpg (53.72 KiB) Viewed 1056 times
2019629215436_alet2.jpg
2019629215436_alet2.jpg (35.23 KiB) Viewed 1056 times
2019629215456_alet5.jpg
2019629215456_alet5.jpg (52.52 KiB) Viewed 1056 times
I FILLED HOLES WITH HIGH TEMPERATURE SILICONE
20198262131_tempt1.jpg
20198262131_tempt1.jpg (81.31 KiB) Viewed 1056 times
This evening I did the test again, outside temperature 25 °, 37 ° engine at the start.
The result
201982621150_tempt3.jpg
201982621150_tempt3.jpg (41.36 KiB) Viewed 1056 times
the problem it's solved!
I also did a very tough off-shore tour and the maximum temperature reached was 81 °
Unfortunately, during the unmodified tests I reached 95 ° many times and sometimes even
201931417170_temp10.jpg
201931417170_temp10.jpg (46.5 KiB) Viewed 1056 times
The engine seems a little less powerful than a new one, but for now I won't change it!

If you have ideas, and above all test results, they are welcome !!

User avatar
Aquakitty   100 W

100 W
Posts: 138
Joined: Jun 10 2017 3:07pm

Re: Hardware temperature control tsdz2

Post by Aquakitty » Aug 26 2019 8:02pm

Hi, I am eager to follow this thread. What temperature sensor are you using?
Rans Enduro Sport w/Sturmey-Archer XFRD8, TSDZ2 open source mod 48v
GT Verb Comp f/s TSDZ2 open source 48v
Diadora Ampio bargain f/s TSDZ2 48v
2020 Diamant 247 belt drive 1000w 9C 48v
Hardtail made from random bits I had laying around, ginormous old Marzocchi fork, BBS02

andrea_104kg   100 W

100 W
Posts: 152
Joined: Mar 05 2018 5:51pm

Re: Hardware temperature control tsdz2

Post by andrea_104kg » Aug 27 2019 3:56am

It's a digital termometer, I found it on ebay..

casainho   1.21 GW

1.21 GW
Posts: 3577
Joined: Feb 14 2011 2:43pm

Re: Hardware temperature control tsdz2

Post by casainho » Aug 28 2019 4:30pm

Looking at the picture of your motor that it is brown, it is almost for sure already demagnetized/damaged and so has lower torque and pulling more current, that will result you pulling more current to have the desired assistance and it will get hot even faster.......

I suggest you to replace you motor asap and be careful with your tests of temperature or you may repeat the same demagnetization/damage to the new motor.

Also the results of your tests may not reflect what happens to a good working motor that other users have.
Developer of the Flexible OpenSource firmware for EBike motor controllers (TSDZ2 and KT) and LCDs (KT-LCD3 and Bafang 850C color LCD).

If you like my work, please consider making a donation. I am being using the donations to buy needed resources for my developments. My paypal: casainho AT gmail.com.

andrea_104kg   100 W

100 W
Posts: 152
Joined: Mar 05 2018 5:51pm

Re: Hardware temperature control tsdz2

Post by andrea_104kg » Aug 28 2019 6:08pm

casainho wrote:
Aug 28 2019 4:30pm
Looking at the picture of your motor that it is brown, it is almost for sure already demagnetized/damaged and so has lower torque and pulling more current, that will result you pulling more current to have the desired assistance and it will get hot even faster.......

I suggest you to replace you motor asap and be careful with your tests of temperature or you may repeat the same demagnetization/damage to the new motor.

Also the results of your tests may not reflect what happens to a good working motor that other users have.

Unfortunately, BEFORE making the change, at each tour the temperature reached a minimum of 90 ° and sometimes over 100 °.
Unfortunately they are my normal country paths, I can't avoid them.
This with a new engine (bought in January). I know it could be demagnetized, for now it continues to work even on the 25% climb so I don't replace it :-)
Currently it is very difficult to make him reach 70 °. I think the temperature reduction would also be in other engines.
Indeed it could be bigger, if mine is damaged it heats up more.

andrea_104kg   100 W

100 W
Posts: 152
Joined: Mar 05 2018 5:51pm

Re: Hardware temperature control tsdz2

Post by andrea_104kg » Aug 28 2019 6:13pm

I honestly don't understand how it is possible to go to the mountains without at least one temperature sensor. I believe that most users have a damaged engine and don't know it. Of course, if you use the bike 3-4 times a month, it will probably last a lot. I do those routes every day :-)

User avatar
Aquakitty   100 W

100 W
Posts: 138
Joined: Jun 10 2017 3:07pm

Re: Hardware temperature control tsdz2

Post by Aquakitty » Aug 29 2019 12:26am

I bought a few thermometers to test my units. It's kind of ridiculous this motor overheats so easily. Just riding to town and back and my spouse's motor was nearly too hot to touch.
Rans Enduro Sport w/Sturmey-Archer XFRD8, TSDZ2 open source mod 48v
GT Verb Comp f/s TSDZ2 open source 48v
Diadora Ampio bargain f/s TSDZ2 48v
2020 Diamant 247 belt drive 1000w 9C 48v
Hardtail made from random bits I had laying around, ginormous old Marzocchi fork, BBS02

andrea_104kg   100 W

100 W
Posts: 152
Joined: Mar 05 2018 5:51pm

Re: Hardware temperature control tsdz2

Post by andrea_104kg » Aug 29 2019 4:00am

I share. before the modification it was enough to climb 2 km to go over 90 °. On the road even worse if I tried to go to the maximum power on 6-8% climbs, starting from a 55 ° motor already after a km I was almost at 100 °.

Elinx   100 mW

100 mW
Posts: 43
Joined: Aug 09 2019 6:58am

Re: Hardware temperature control tsdz2

Post by Elinx » Sep 01 2019 5:08pm

casainho wrote:
Aug 28 2019 4:30pm
Looking at the picture of your motor that it is brown, ...........
On the first picture (with the steel cables) the laminated steelcore isn't brown.
Isn't this brown color a copper plate placed on the laminated steelcore? (the airfins are also copper)
Because this looks almost as flat as the end caps. Normally it is about a few mm difference

edit:
andrea_104kg said here it is 0,8mm bended copperplate
Last edited by Elinx on Sep 13 2019 5:40pm, edited 1 time in total.

sysrq   100 W

100 W
Posts: 113
Joined: May 07 2018 6:05pm

Re: Hardware temperature control tsdz2

Post by sysrq » Sep 01 2019 6:08pm

Probably even something like this
https://www.fastrax-rc.com/index.php?_r ... tsink-unit
https://www.ebay.co.uk/i/382967894510?c ... UJEALw_wcB
won't be able to help much.
When I opened the black external cover for the first time it smelled like a stale outgassing, not like burnt enamel, so it might have been overheated slightly.
BLDC PM motors are known for performance loss due to overheating despite having high efficiency for their size, so nothing out of ordinary.

Theoretically anyone can get and wrap some pipe around any size motor.
https://www.lazada.com.ph/products/orig ... 88434.html
Otherwise there is no way to cool it evenly in such a constricted space if one wants to preserve the original looks.
Since everyone knows that for applications with unpredictable loads it's always better to use oversized motors or appropriate gearing instead of overengineered cooling solutions, then there's nothing else to say.
search.php?keywords=Objective+&t=19359&sf=msgonly

sysrq   100 W

100 W
Posts: 113
Joined: May 07 2018 6:05pm

Re: Hardware temperature control tsdz2

Post by sysrq » Sep 02 2019 9:21am

Could heat pipes be used in general to transfer the heat passively while relatively simplifying whole installation process compared to copper plates on laminated core?
Vapor chamber might be too expensive.

andrea_104kg   100 W

100 W
Posts: 152
Joined: Mar 05 2018 5:51pm

Re: Hardware temperature control tsdz2

Post by andrea_104kg » Sep 02 2019 12:24pm

Some information based on my experiences:
1) there is a way to see if an engine is burned. The current absorption increases excessively as there is a short circuit. If you can't get away with just a few watts (visible on the LCD03) it is likely that the engine is damaged. Obviously a damaged engine breaks down completely after a while, but for a while it can still work, especially when only one of the three phases is burned.
2) the motor of the tsdz2 absorbs on average 500w. The weight of the engine is about 3kg. Imagine having to heat a body exposed to the air by 3kg with a hair dryer of 500w. Obviously you would be in trouble. this means that it is enough to ensure contact with the case to drastically lower the engine temperature. I don't even need cooling fins. however it is not easy to ensure good motor-case contact.
An alternative way to mine can be to pierce the case and to force inside the copper straw hat with patience. the holes can then be closed with epoxy paste.
https://www.alambiccorame.it/it/pagliet ... empimento/

OttawaEbiker   10 µW

10 µW
Posts: 6
Joined: Sep 03 2019 11:44am

Re: Hardware temperature control tsdz2

Post by OttawaEbiker » Sep 03 2019 11:49am

I've got a TSDZ2 on the way... andrea_104kg, could you provide some installation details for the final copper fin solution?

I assume:
- buy a copper sheet at home depot (gauge?)
- cut to fit in motor slots
- any glue/silicone between the layers to keep in place?
- copper fins are simple long sheets bent as needed? how are they glued to the copper layer underneath?
- is it still worth installing the temp sensor with this solution?

Thanks!

sysrq   100 W

100 W
Posts: 113
Joined: May 07 2018 6:05pm

Re: Hardware temperature control tsdz2

Post by sysrq » Sep 03 2019 2:43pm

Another low key solution I can think of would be tying all around the motor for even some 2-5mm thick OFHC copper 90 degree L bends depending on measurements in between the screws around in conjunction with thermal pads using a jubilee clip. The other ends would press against the round protrusion of the external cover above the motor terminal plate making it possible to put some pressure on the 2 mm thick thermal pads while tightening the cower screws. This approach might provide more uniform heat removal from the motor, bigger effective heat transfer surface area, more flexible tolerances and thinner thermal transfer path and somewhat adjustable compression pressure.
Last edited by sysrq on Sep 03 2019 5:44pm, edited 1 time in total.

andrea_104kg   100 W

100 W
Posts: 152
Joined: Mar 05 2018 5:51pm

Re: Hardware temperature control tsdz2

Post by andrea_104kg » Sep 03 2019 3:06pm

the annealed copper sheet (softer than the raw copper) that I had was 0.8 mm. To curve it I beat it lightly with a hammer on a curved surface (you can see the marks on the photo). it took two layers to bring the engine to the desired thickness. I don't remember what glue I used to hold them together :-(. But I think a common thermal paste is enough. When the job is done, the fins hold them in place.
The fins are obtained from a single folded sheet.
They are held in place by a thin stainless steel wire which, as seen in the photo above, passes over a motor screw and underneath I put a screw on the crankcase where I tied the wire. However after the silicone sealant has hardened I believe that the wires are no longer needed.
Also this evening I did the usual climb where the engine arrived at 90 °, the engine stopped at 67 ° so I think the temperature sensor is not needed.

sysrq   100 W

100 W
Posts: 113
Joined: May 07 2018 6:05pm

Re: Hardware temperature control tsdz2

Post by sysrq » Sep 03 2019 5:50pm

andrea_104kg wrote:
Sep 03 2019 3:06pm
the annealed copper sheet (softer than the raw copper) that I had was 0.8 mm. To curve it I beat it lightly with a hammer on a curved surface (you can see the marks on the photo). it took two layers to bring the engine to the desired thickness. I don't remember what glue I used to hold them together :-(. But I think a common thermal paste is enough. When the job is done, the fins hold them in place.
The fins are obtained from a single folded sheet.
They are held in place by a thin stainless steel wire which, as seen in the photo above, passes over a motor screw and underneath I put a screw on the crankcase where I tied the wire. However after the silicone sealant has hardened I believe that the wires are no longer needed.
Also this evening I did the usual climb where the engine arrived at 90 °, the engine stopped at 67 ° so I think the temperature sensor is not needed.
If one wants to keep the outer cover intact then transferring the heat to the outer cover directly is an other option.

OttawaEbiker   10 µW

10 µW
Posts: 6
Joined: Sep 03 2019 11:44am

Re: Hardware temperature control tsdz2

Post by OttawaEbiker » Sep 03 2019 8:18pm

andrea_104kg wrote:
Sep 03 2019 3:06pm
the annealed copper sheet (softer than the raw copper) that I had was 0.8 mm. To curve it I beat it lightly with a hammer on a curved surface (you can see the marks on the photo). it took two layers to bring the engine to the desired thickness. I don't remember what glue I used to hold them together :-(. But I think a common thermal paste is enough. When the job is done, the fins hold them in place.
The fins are obtained from a single folded sheet.
They are held in place by a thin stainless steel wire which, as seen in the photo above, passes over a motor screw and underneath I put a screw on the crankcase where I tied the wire. However after the silicone sealant has hardened I believe that the wires are no longer needed.
Also this evening I did the usual climb where the engine arrived at 90 °, the engine stopped at 67 ° so I think the temperature sensor is not needed.
Thanks for the information. I'm sure I have several weeks of waiting before receiving the motor so I'll be monitoring all the temperature fix threads closely.

maximusdm   10 W

10 W
Posts: 69
Joined: Sep 08 2018 12:51pm

Re: Hardware temperature control tsdz2

Post by maximusdm » Sep 07 2019 7:10am

Nice solution, Andrea104!

I made a similar one from aluminium. The complicated part is how to get good thermal contact with the exterior shield so you can draw the heat to the outside.
How did you did that?

andrea_104kg   100 W

100 W
Posts: 152
Joined: Mar 05 2018 5:51pm

Re: Hardware temperature control tsdz2

Post by andrea_104kg » Sep 07 2019 5:07pm

honestly I do not know! my idea was to make fins but then the copper is so conductive that where it touches in the casing it transmits a lot of heat .. the casing is always very hot

sysrq   100 W

100 W
Posts: 113
Joined: May 07 2018 6:05pm

Re: Hardware temperature control tsdz2

Post by sysrq » Sep 07 2019 6:12pm

5 or 6 adhesive pyrolytic graphite sheet strips around the motor and stacked 3mm thermal pads on top of the motor end cap pressing against the cover lined with adhesive pyrolytic graphite sheet to spread the heat better, since the covev looks lika a casted aluminium alloy which a are known to for their poorer thermal conductivity. I ok
EYGA091202A was no more expensive than a copper plate. Hard to find straightforward solution in such a small and relatively complex shaped space while preserving the original IP rating.

Elinx   100 mW

100 mW
Posts: 43
Joined: Aug 09 2019 6:58am

Re: Hardware temperature control tsdz2

Post by Elinx » Sep 08 2019 2:17am

sysrq wrote:
Sep 07 2019 6:12pm
.........
EYGA091202A was no more expensive than a copper plate. .........
The sheets I found are only 17µm, meaning you need 60 layers to get a 1mm. :?

I think the cheapest way is to use metal for reasonable thermal conductivity
Massive, braided, pressed sponge or pressed mesh, where (massive) copper is preferred and second aluminium.

But I think, for better thermal improvement you can use every material that has a bigger thermal conductivity than dry immovable air.
Filling the airgap between the laminated core and cover case at the front, has the highest effectivity.
Also you have there the biggest contact area with the hottest part and coolest (airflo) part of the motor.
Last edited by Elinx on Sep 08 2019 5:04am, edited 1 time in total.

andrea_104kg   100 W

100 W
Posts: 152
Joined: Mar 05 2018 5:51pm

Re: Hardware temperature control tsdz2

Post by andrea_104kg » Sep 08 2019 4:08am

I don't think that the thermal conductivity of aluminum is so bad ... with my system I lowered the temperature by 20 degrees. I don't think complex and expensive solutions are necessary.

Elinx   100 mW

100 mW
Posts: 43
Joined: Aug 09 2019 6:58am

Re: Hardware temperature control tsdz2

Post by Elinx » Sep 08 2019 4:41am

andrea_104kg wrote:
Sep 08 2019 4:08am
I don't think that the thermal conductivity of aluminum is so bad ..................
No it is not bad at all. It is one of the better cheap metals for thermal conductivity, but copper is preferable.

Siver 430W/m.k
Copper 380W/m.k
Gold 320W/m.k
Aluminium 210W/m.k
Brass 150W/m.k
Stainless Steel 100W/m.k
Iron 80W/m.k
Air 0.03W/m.k
Last edited by Elinx on Sep 09 2019 1:21am, edited 1 time in total.

sysrq   100 W

100 W
Posts: 113
Joined: May 07 2018 6:05pm

Re: Hardware temperature control tsdz2

Post by sysrq » Sep 08 2019 5:32am

Elinx wrote:
Sep 08 2019 2:17am
sysrq wrote:
Sep 07 2019 6:12pm
.........
EYGA091202A was no more expensive than a copper plate. .........
The sheets I found are only 17µm, meaning you need 60 layers to get a 1mm. :?

I think the cheapest way is to use metal for reasonable thermal conductivity
Massive, braided, pressed sponge or pressed mesh, where (massive) copper is preferred and second aluminium.

But I think, for better thermal improvement you can use every material that has a bigger thermal conductivity than dry immovable air.
Filling the airgap between the laminated core and cover case at the front, has the highest effectivity.
Also you have there the biggest contact area with the hottest part and coolest (airflo) part of the motor.
So there might be insufficient heat transfer capacity below 1mm thickness?

https://www.electronicdesign.com/circui ... al-demands

Cooling only front part of the motor itself could create internal hotspots.

Elinx   100 mW

100 mW
Posts: 43
Joined: Aug 09 2019 6:58am

Re: Hardware temperature control tsdz2

Post by Elinx » Sep 08 2019 6:36am

sysrq wrote:
Sep 08 2019 5:32am
...............
So there might be insufficient heat transfer capacity below 1mm thickness?

https://www.electronicdesign.com/circui ... al-demands

Cooling only front part of the motor itself could create internal hotspots.
We are not talking about chips with nano dimensions for the electronical parts into the chip, but about a relative big laminated iron core.
The sheets you are talking about are so thin for the best contact with the aluminium heatsinck above a chip. The conductivity is higher than grease or glue, that is what they mean with preventing hotspots.
Als for 1mm you need 60 layers of these sheets. Between every layer you got conductivity loss.

The iron core can conduct the heat relative easy, but the air gap is not, so the heat stays in the motorcasing.
So fill the air gap with copper or aluminium is the easiest way.

Post Reply