The efficiency of the BLDC motor Tongsheng TSDZ2 and Astro 3205 compared

avandalen

100 W
Joined
Oct 2, 2010
Messages
175
Location
Maastricht, The Netherlands
How good is the brushless DC motor that is being used inside the Tongsheng TSDZ2 ?

To answer this question, I have compared the TSDZ2 motor with one of the highest quality electric motors available: the Astro 3205 from Astro Flight.

This is what’s Astro Flight tells: The goal with the 32XX has been to create a superior quality motor. Starting at the core of the motor, the 32XX uses the highest grade of sintered samarium-cobalt magnet. This is a very high-quality magnet that is rarely found on motors of this price point.

Here are the corresponding efficiency curves:

Motor-graph-TSDZ2_.jpg

Motor-graph-Astro-3205-14T.jpg

You can see at the graphs that the TSDZ2 motor is almost as good as the expensive Astro 3205, which cost over $400. The motor inside the TSDZ2 is really good.
See the article
https://www.avdweb.nl/solar-bike/hub-motor/tongsheng/efficiency-bldc-motor-tongsheng-tsdz2-and-astro-3205-compared
 
What about higher torques? The motors don't seem bear saturation in your testing?

And hos about iron losses. They might be dramatically different in the two motors :)
 
If you factor in the gear reduction required for the toy airplane motor, you'll find it's always at a disadvantage unless you're building a toy airplane.
 
Just for a reality check, motors that don't reach 90% aren't something called efficient IMHO.

The last few motors I worked with dyno'd peaking 97%, and stays over 94% for a great portion of the efficiency curve. I would not consider either of those motors to be efficient, but like Chalo mentioned, you would certainly lose a lot more efficiency in the power transfer stages to convert the high RPM Astro to drive bicycle wheel speeds.
 
liveforphysics said:
Just for a reality check, motors that don't reach 90% aren't something called efficient IMHO.

The last few motors I worked with dyno'd peaking 97%, and stays over 94% for a great portion of the efficiency curve. I would not consider either of those motors to be efficient, but like Chalo mentioned, you would certainly lose a lot more efficiency in the power transfer stages to convert the high RPM Astro to drive bicycle wheel speeds.

True, but we need to note that the tzdz2 motor setup is also geared down, possibly even more than needed for the astro.
 
Universalus said:
Your graph about Astro 3205 has incorrect X axis values.

Motor-graph-Astro-3205-14T.jpg

I will repair that soon at my website
https://www.avdweb.nl/solar-bike/hub-motor/tongsheng/efficiency-bldc-motor-tongsheng-tsdz2-and-astro-3205-compared
 
It's strange that the Astro motor doesn't peak higher than it does in your measurement. I wonder if it could be an artefact from testing at too low speed or too low torque. Since the efficiency (losses vs output) is a balance between mainly copper losses at higher currents and increasing iron losses at higher rpm a dyno plot of only efficiency vs torque won't tell the whole picture, right?

Also the Astro blab about cobalt samarium magnets is not really true, these magnets perform better at elevated temperature than neodym but neodym has a higher field strength. This would mean that you can probably push the Astro motor further into heating but that there is no peak efficiency gain from the magnet choice.

I've seen pics where the Astro seemed to have distributed winding, not good for efficiency. Have you opened the astro, can you confirm this?
 
Vbruun said:
What about higher torques? The motors don't seem bear saturation in your testing?

And hos about iron losses. They might be dramatically different in the two motors :)
Iron losses are included in Tf
I made new graps to 1000W :
https://www.avdweb.nl/solar-bike/hub-motor/tongsheng/efficiency-bldc-motor-tongsheng-tsdz2-and-astro-3205-compared
 
Chalo said:
If you factor in the gear reduction required for the toy airplane motor, you'll find it's always at a disadvantage unless you're building a toy airplane.
The speed difference is just 8000rpm - 4000rpm
Both motors can use 2-stage gear transmissions.
 
liveforphysics said:
Just for a reality check, motors that don't reach 90% aren't something called efficient IMHO.

The last few motors I worked with dyno'd peaking 97%, and stays over 94% for a great portion of the efficiency curve. I would not consider either of those motors to be efficient, but like Chalo mentioned, you would certainly lose a lot more efficiency in the power transfer stages to convert the high RPM Astro to drive bicycle wheel speeds.

The speed difference is just 8000rpm - 4000rpm.
Both motors can use 2-stage gear transmissions, so with the same gear losses.
What is the power of the motors you are talking about? Higher power motors have higher efficiency.
 
larsb said:
It's strange that the Astro motor doesn't peak higher than it does in your measurement. I wonder if it could be an artefact from testing at too low speed or too low torque. Since the efficiency (losses vs output) is a balance between mainly copper losses at higher currents and increasing iron losses at higher rpm a dyno plot of only efficiency vs torque won't tell the whole picture, right?
The parameters are measured in this way:
https://www.avdweb.nl/solar-bike/hub-motor/measuring-motor-parameters
 
larsb said:
Also the Astro blab about cobalt samarium magnets is not really true, these magnets perform better at elevated temperature than neodym but neodym has a higher field strength. This would mean that you can probably push the Astro motor further into heating but that there is no peak efficiency gain from the magnet choice.

Yes that's interesting, if the Astro doesn't have stronger magnets then the motor doesn't have a higher efficiency either
 
larsb said:
I've seen pics where the Astro seemed to have distributed winding, not good for efficiency. Have you opened the astro, can you confirm this?
that might explain this:
The motor resistance of the 8-turn motor is 0.05Ohm according to the datasheet.
At 14 turns, the resistance should be 0.05 Ohm (14/8)^2 = 0.153 Ohm.
But I measured 0.195 Ohm, which is 27% larger.
 
avandalen said:
Vbruun said:
What about higher torques? The motors don't seem bear saturation in your testing?

And hos about iron losses. They might be dramatically different in the two motors :)
Iron losses are included in Tf
I made new graps to 1000W :
https://www.avdweb.nl/solar-bike/hub-motor/tongsheng/efficiency-bldc-motor-tongsheng-tsdz2-and-astro-3205-compared

How can they be included? The iron losses are dependent on rpm and your graph only maps efficiency as a function og torque.
 
It’s included in the Tf parameter (friction torque) in Avandalens’ calculations, however these two motors have different rpms and iron losses scale by frequency^2 so there could very well lie an issue with the efficiency calc there.

Generally they are both tested at the same power but one of them could have a better performance at another load point. Noload losses at the same rpm would be interesting to see.

The astro has surprisingly high phase resistance and this really lowers the comparable efficiency, right Avandalen? Twice the kV but still higher resistance…not good.
 
avandalen said:
liveforphysics said:
Just for a reality check, motors that don't reach 90% aren't something called efficient IMHO.
The last few motors I worked with dyno'd peaking 97%, and stays over 94% for a great portion of the efficiency curve.
What is the power of the motors you are talking about? Higher power motors have higher efficiency.
Yeah, I'm curious to know if any sub 1kw motor exists that is over 90% efficient at bicycle speeds.
 
larsb said:
It’s included in the Tf parameter (friction torque) in Avandalens’ calculations, however these two motors have different rpms and iron losses scale by frequency^2 so there could very well lie an issue with the efficiency calc there.

Generally they are both tested at the same power but one of them could have a better performance at another load point. Noload losses at the same rpm would be interesting to see.

The astro has surprisingly high phase resistance and this really lowers the comparable efficiency, right Avandalen? Twice the kV but still higher resistance…not good.

Ah I see.

So the graph shows efficiency at different torques @ a predefined (preferably well chosen) RPM.

Now I get it. Good job avandalen.
 
Back
Top