Motor Bearings DO make a difference in temps

[SirCrashALot]

I always replace worn bearings in my 4985 motors with SKF metal shielded 626Z. They are rated for 80k rpm. And i choose the ones with the tighter tolerance. (C3 i believe).
It'd be nice to see you test those.
Recently i put regular rubber shielded SKF bearings in my 1520 motor. SKF didn't have the metal shielded C3's in the correct size. Given that the regular rubber shielded SKF's are rated for 50k rpm, i thought I'd give it a try.

 

[BashingBrian]

I run stock bearings in all motors, bash hard, don’t clean them, all’s good..!! and I don’t run fans either, just gear correctly..

 

[Logan's_Rcs]

I run stock bearings in all motors, bash hard, don’t clean them, all’s good..!! and I don’t run fans either, just gear correctly..

Interesting. I think Castle and Hobbywing use better quality NMB bearings than TP does. I mean, maybe that's why I've heard a lot of people say they run hot? From what I've seen personally, the TP bearings are filled with a lot of grease, they are smooth, but they don't spin "freely" if that makes sense.

 

[Logan's_Rcs]

Recently, I asked TP Power to send an extra set of end caps for their 4060 line so I could test out some different bearings in my motor.. I tested the bearings by spinning my motor no load for 45 seconds at full throttle (Do not do this at max voltage of your motor, it will destroy it). After 45 seconds no load, I took the temperature.

First, I tested Acer Ceramics at 45 seconds. After 45 seconds, no load, my motor came to around 91 degrees Fahrenheit.
Next, I tested the stock NMB 688 bearings. After 45 (actually slightly less) seconds, the motor was at 101 degrees Fahrenheit!

Just by going to better bearings, my motor dropped 10 degrees in the same amount of time. There was no air flow around the motor, and all the ESC settings were the same. Only change was my bearings. I took the tests on different days inside and the ambient temp was the same. The motor started at the same temp both times.

I'm going to be looking for more bearings to test out in this regard, are there any other brands you guys would like to see me try? I was considering VXB and pro-motion rc ceramics next.

VXB bearings arrived today!

 

[Chapinb]

Wooo!

 

[Logan's_Rcs]

Wooo!

Quick update: I ordered the VXB ceramics as dry ceramics (dry lube like graphite, not oil or grease). I've heard that motor bearings are supposed to have some oil in them, but that is mainly for the friction between the steel balls and the outer bearing. Because these are ceramics, I'm not sure they'll need grease. Anyway, without them even broken in, spinning by hand they spin at least 10x longer than the stock NMB and Acer bearings I picked up. I'll try installing them in the motor and doing the 45 second test again!

I'll post back with results when I do the test

 

[Logan's_Rcs]

Quick update: I ordered the VXB ceramics as dry ceramics (dry lube like graphite, not oil or grease). I've heard that motor bearings are supposed to have some oil in them, but that is mainly for the friction between the steel balls and the outer bearing. Because these are ceramics, I'm not sure they'll need grease. Anyway, without them even broken in, spinning by hand they spin at least 10x longer than the stock NMB and Acer bearings I picked up. I'll try installing them in the motor and doing the 45 second test again!

I'll post back with results when I do the test

Just tested the VXB bearings. I wouldn't recommend anybody buy from VXB especially for a motor bearing. While the bearings spin very long and are quite efficient, their ABEC7 rating isn't even close to true. The motor shaft wiggles side to side like the bearing is bad even though it is new! (And yes, the inner diameter of the bearing is the same as the motor shaft). However, with that said, the VXB's still performed better than the stock NMB's that came with the car. The motor got to around 87 degrees after a full 45 second run.

I think I may try and find some higher quality ceramics from Boca or other bearing companies. For now though, the Acer bearings still seem to be the best for motor replacements. No play in the Acer bearings, and they run the motor as cool as the VXB's even with grease!

 

[ih8nixon]

You're running tests on the far left side of this plot where the slope of the efficiency is almost vertical. The slope there means the faster it spins the worse the motor efficiency. I can't remember if it's the motor fighting itself or something with the back EMF causing reduced current that drops the efficiency in this region. This is what was reffering to when I said the faster running on could heat up more as its efficiency is lower.

The other problem I mentioned is you're running the bearing in basically an unloaded state. Just spinning up a bearing doesn't really put it under much load. You need some sort of load pressing on the shaft. For instance with XJeepguy's example at his work if the bearings were unloaded before testing a good number of the shot bearings wouldn't run hot and appear to function fine. Or as you pointed out the wobbly shaft on the new bearings is bad, but without a load on the shaft that poor tolerance won't manifest as the loses it should. WIth your current testing the loosest dryest bearing will always win.

If you want to get good data you will need an ESC with a governor mode (esc tries to run at constant rpm heli guys use them). Various versions of BLHeli escs have governors and can be pretty cheap. You'll also need to load the motor shaft. Gearing it to a prop would be a good way to get a consistent load at and the gearing would ensure the bearing is loaded like it is in the cars. A consitent voltage source and a watt meter would let you measure the difference in losses from the bearings.

 

[Logan's_Rcs]

View attachment 421793
You're running tests on the far left side of this plot where the slope of the efficiency is almost vertical. The slope there means the faster it spins the worse the motor efficiency. I can't remember if it's the motor fighting itself or something with the back EMF causing reduced current that drops the efficiency in this region. This is what was reffering to when I said the faster running on could heat up more as its efficiency is lower.

The other problem I mentioned is you're running the bearing in basically an unloaded state. Just spinning up a bearing doesn't really put it under much load. You need some sort of load pressing on the shaft. For instance with XJeepguy's example at his work if the bearings were unloaded before testing a good number of the shot bearings wouldn't run hot and appear to function fine. Or as you pointed out the wobbly shaft on the new bearings is bad, but without a load on the shaft that poor tolerance won't manifest as the loses it should. WIth your current testing the loosest dryest bearing will always win.

If you want to get good data you will need an ESC with a governor mode (esc tries to run at constant rpm heli guys use them). Various versions of BLHeli escs have governors and can be pretty cheap. You'll also need to load the motor shaft. Gearing it to a prop would be a good way to get a consistent load at and the gearing would ensure the bearing is loaded like it is in the cars. A consitent voltage source and a watt meter would let you measure the difference in losses from the bearings.

I was considering getting a motor analyzer with a power supply to do that, I have to wait till the next pay check though!

 

[Chapinb]

When my bearings come in, your hitting on a few things I'll be testing. Will be affixed to a prop. Find a set rpm, measure watts in and temp.
Will be powered from a bench top power supply so can get watt accuracy into the hundredths.
Will be ran for a time to allow break in of bearings as well.

 

[Logan's_Rcs]

When my bearings come in, your hitting on a few things I'll be testing. Will be affixed to a prop. Find a set rpm, measure watts in and temp.
Will be powered from a bench top power supply so can get watt accuracy into the hundredths.
Will be ran for a time to allow break in of bearings as well.

What bearings did you order?

 

[Chapinb]

What bearings did you order?

From trb. Not positive the oem. Replacing stock bearings with new steel bearings. All bearings show similar markings and construction, so hoping for the only variable to be the ceramic. Ceramics only have ceramic bearings, race/cage is steel. Bearings are sheilded.
Going to degrease both then add a light bearing oil.

 

[Logan's_Rcs]

From trb. Not positive the oem. Replacing stock bearings with new steel bearings. All bearings show similar markings and construction, so hoping for the only variable to be the ceramic. Ceramics only have ceramic bearings, race/cage is steel. Bearings are sheilded.
Going to degrease both then add a light bearing oil.

I think it would be interesting if you tested it with the grease in and then with the oil, just to see if that makes a difference as well

 

[Logan's_Rcs]

I think it would be interesting if you tested it with the grease in and then with the oil, just to see if that makes a difference as well

Found a video from Tekin where they replaced the original bearings in one of their motors with their ceramic bearings. No load current dropped by 0.7 amps and the motor gained 500rpm.

Check it out:

 

[Barmaleus]

Have you tried Fushi bearings?

 

[Logan's_Rcs]

Have you tried Fushi bearings?

I have not, I looked at them though. Do they make ceramics?

 

[Barmaleus]

Can't say for sure, sorry.
I got a bunch of their rubber sealed ABEC 3/ABEC 7 bearings for my Kraton. So far feel solid.

 

[Logan's_Rcs]

Can't say for sure, sorry.
I got a bunch of their rubber sealed ABEC 3/ABEC 7 bearings for my Kraton. So far feel solid.

Nice! Where did you order them from?

 

[Barmaleus]

I got these from their official store at Aliexpress.
They seem to have a dedicated section of ceramics indeed: https://fushi.aliexpress.com/store/...e=bestmatch_sort&shop_sortType=bestmatch_sort

 

[Chapinb]

Testing has started. Decided against the prop, as it loads the bearings axialy rather than in the car where it loads radially. Dunno if it makes a difference, but figured I'd get as close to our use case as possible. Using a 23/50t combo with the motor as far over, screw against the stop. The mesh is reasonable, but looser than I'd like. But this exact position is easily repeatable due to it being at the far extreme limit of the mesh movement. Diffs were recently rebuilt with used stock, so there will be some break in for them as well. Ran a different motor for ~1hr till the watts settled, telling me the diffs are more or less stable.
Currently breaking in stainless balls, 15v, started at about 27w, not positive because I was playing a little first. Not sure how long it will take to break the bearings in. Not sure if I'll see the watts settle or just give it a specific time. Hw app says I'm sitting at 15% throttle and 11.2k rpm - don't know if it's actually 11.2k or 5.6k since it's 4 poles/not sensored. Current can temp 100, bearing 90, ambient 66, no fan.