Bafang BBS-02/HD vs M600 vs M620 (Ultra)

Ravi Kempaiah

Well-Known Member
Region
Canada
City
Halifax
I hope someone here with experience can share a detailed report with pictures on the differences between the legacy drive system from Bafang, like BBS-02/HD, with the latest M640 (Ultra) and M600 motors. This will help forum readers benefit from the collective knowledge of more experienced riders.

In the last 6 years, we have seen the evolution from BBS-HD --> M620 ---> M600 --> M820
This has not been without growing pains and supply chain issues, but forum readers would benefit from knowing more nuanced details, which can only be gained from experience.
  1. Quantum of power output and smoothness of torque sensing
  2. Reliability in all weather conditions
  3. Support and parts supply
  4. Controller programmability either from OEM or at the end-user level
  5. Communication with BMS, display, charge
I see @Gionnirocket has relevant experience and it would be good to hear from other riders as well. This would be quite valuable for a lot of readers.
 
Relative new comer to the M620 from the BBS... I can say this off the bat. Though the BBS was great for a dyi build, I wouldn't expect to see one on a production bike...nor would one interest me.
As far as the model progression I would have thought the M560 more desirable than M820. Though I consider it more of a side step then upgrade to the M620. The M560 lower weight sounds interesting but being only available in CANBUS with no end user programming leaves me hesitant.

Bafang has a good reputation as a solid motor.. But where it falls short is manufacture information, especially when it comes to firmware changes and programming support. Ravi you know we tried for months to get support and log_in credentials to use the official Bafang software to tune my UART M620 and we are still waiting. That said with the information available and some testing I was able to tweak it to my liking using the available hacked software. This is one of the biggest pluses for this motor. But there's still some parameters I haven't adjusted much as there's just no reliable data available. When the CANBUS version was released there were many complaints on how unrefined it performed. I think the stock programing is better now, but there's still no end user tweaking and as just mentioned above, even dealer programing is near impossible to obtain.
So to your asks...
1. After some end user tweaks I'm very happy with the torque sensor and motor output. Depending on PAS it's either unintrusively adding some assistance, to multiplying my input, to launching me on a power ride... but all predictable and well controlled.
2. Can't offer anything on all weather reliability in this short time... and the fact that I typically don't ride in bad weather. I'm delicate that way. 🙃
3. Support from vendors is getting better and for parts between places like California eBike, Greenbikekit and Aliexpress you usually you can find what you need. Again some manufacture support would be welcomed.
4.Programing, I think I've ranted enough... but a manufacture guide to ALL the parameters would be welcomed.
5. As far as I know BMS communication isn't available in this corner of the world and has only been available with an official Bafang battery sold only in China.
BMS communication would be welcomed.
 
Good morning. These are some really good questions and I will answer to the best of my experience since owning (rec'd 5/23) my m620 equipped bike.
  1. Quantum of power output and smoothness of torque sensing
Coming from Bafang hub drives- There is an obvious difference power wise between the hubs and middrives. The amount of power that the UART m620 puts out is awesome and another reason I got it. Out of the box the one thing I noticed though is the throttle power was too much from a start. Luckily I knew this going into the bike thanks to research. Easy fix with a quick programming change.
Torque sensing on the m620 has been a surprise favorite of mine. I like the predictable power application that is applied or taken away when riding. There is no perceived delay of on or off, making the bike feel more natural and controllable.

2. Reliability in all weather conditions
I really do not plan rides if the weather is going to be bad so my experience is minimal.
I have been caught in a typical rogue Florida rain shower and was able to get under cover for the extra 5 minutes it rained. No issues after the rain
Some rides after storms and through some puddles off main trail, without an issue.

3. Support and parts supply
The only things bought that are related to the m620 itself has been an eggrider (through Luna) and a set of new smaller cranks via Alieexpress.
From the looks of it, there are a couple of well known parts suppliers that help fill the void of parts that maybe Bafang could do better in offering.
Beyond that, I am still comfortable in owning the m620 and know that I can get parts, maybe not as fast as I may think I need them.

4. Controller programmability either from OEM or at the end-user level

Having the UART m620 and being able to program it is the #1 reason I got what I did. I played around with the USB cable and Bafang program when I first got the bike. First thing I did was to program the throttle to be less dragster off the line and more manageable.
I did get the eggrider shortly after getting the bike. I would not change that decision. I like the smaller screen footprint on the handlebars plus the Bluetooth connection to my phone. I do rides with my cellphone mounted on the handlebars so having that display screen at the tip of my finger is fine as a main display.
Getting into the weeds and not trying to be an eggrider fanboi, but being able to reprogram at any point in a ride is great. No need to carry a laptop and cable. I have programmed something before a ride and found that it was not quite right, so all I had to do was pull off the trail, make a few changes and go. This scenario really is more of a ride and test situation and not so common once you get your settings dialed in on the bike.

5.Communication with BMS, display, charge
The Frey CC does not have a BMS according to Frey. So any of the models that uses the same battery as the CC Fat will not have a BMS.
I have had no issues when it comes to the motor talking to the display(s).

One thing that people looking into the higher power/ higher torque Bafang middrives should take into consideration is the wear and tear on the drive train if equipped with the chain and derailleur setup. This could go under Support and Parts supply, but being that this drive train setup is like a normal bike, parts from a local bike shop and support is possible.
 
Last edited:
I don't know wtf @Gionnirocket was rambling on .
I have two Canbus M620 motors 48v and 52v. By far compared to Brose TF motors are superior in performance and less downtime on my bike I can remove and replace motors very easily.
All other maintenance is simple
You don't know wtf I'm talking about because I adjust my motor controller with a laptop... and you adjust yours with a
smash.gif

But seriously I believe you don't have an early edition CANBUS, so that's good for you. I'd be interested in comparing a tweaked UART to it as far as refinement.
Still haven't heard of any user or vendor being able to make adjustments to a CANBUS M62O
 
Last edited:
You detuning your bike.? I have two
Factory tuned M620 Canbus .
I like it for the plug and play factor.
With uart you will never figure out where the sweet spot
Perhaps... I like having both a natural feel and an aggressive one. This I plugged in easily. Now it's play time.
Plug ń play would be great if it suited everyone... Glad your happy but you seem to be one of the few.
I've already found a pretty sweet spot... I just want manufacture information on the settings for a final fine tune.
 
You detuning your bike.? I have two
Factory tuned M620 Canbus .
I like it for the plug and play factor.
With uart you will never figure out where the sweet spot
With Canbus, you'll never know if there is a sweet spot. (until they crack it or allow tuning by the bike owner)
 
I'd ask @Tom@WashDC, @Deacon Blues , @Rome, @keithj69, @greeno as well as any others who undoubtedly have more experience with the m620, and coming from other sellers for their .02 cents.
Let's see if we can get this beast understood and supported as it should be.
I'm sure leaving my posted experiences out was an unintentional omission on your part.

What strikes me as non-ignorable to this day is that after several years (dating back to 2017) and several"programming guides" being posted on the interwebs and here on EBR's forums, it wasn't until last Spring that someone figured out what the columns in the M620's Torque Tab actually represented. Until then, everyone was not just assuming the inputs were things that were wrong, they were basing programming on those wrong assumptions - and yet claimed happiness with their changes?

Whether we've been reading placebo effects or differences too small to matter or riders tuning for their narrow specific use cases is an exercise I'll leave to the readers here.

Enjoy.
 
My two cents. I have several e-bikes including a Quietkat Jeep with a 750 watt rated Bafang 620 Ultra mid-drive motor. The torque output for the 750 watt version of the M620 is unclear - I've seen numbers from 120 Nm to 160 Nm. As far as power, the M620 has no more discernible power output than my 2019 e-bike Bosch CX Gen2 with 75 Nm of torque. The M620 motor has decent power (I'm a large heavy rider), but nothing earth shattering. There are also 1000 watts and 1500 watt versions of the M620. The M620 motor is by far the noisiest motor I've ever had in an e-bike. It's very noisy and after more than 1,000 miles, it is no quieter. The M620 is apparently very inefficient too. The range is very limited compared to other e-bikes, I have a 14 Ah 48V battery in the QK Jeep, and I am lucky to get 25-30 miles in warm weather and 7-10 miles in cold winter weather - and that's with a neoprene jacket over the battery. I am riding on snow in the winter with a fat tire bike at low tire pressure, so that does account for some of the reduced range.
 
As far as power, the M620 has no more discernible power output than my 2019 e-bike Bosch CX Gen2 with 75 Nm of torque.
Are you sure about this?
The motor armature of the M620 is 2x the size of the Bosch CX, and the Bafang M620, in my experience, is a lot more powerful than Bosch CX Gen2.
I have ridden both of them extensively and I agree that the torque sensing is sub-par. I did have a discussion with the R&D department of Bafang, and I was told that older M620 worked on the UART system, and their controller technology was quite dated. The newer systems they are offering, like M560, are far better, and this can be attributed to the newer controller technology and motor design.

There are some reports here of their older M510 motor: https://www.emtbforums.com/threads/bafang-m560.33293/page-10#post-547760

I did ride this exact bike shown in the picture below and I can say it is as good as Bosch in terms of smoothness but a bit more powerful.

1718157636323.png
 
The torque assist on an M620 running the UART takes a bit of tuning from stock via a Windows computer and an accessory cable with the correct software. Along with a fair amount of hunt and pecking. Once achieved I find it to be on par with any pedal assist system I have experienced. Doesn't surprise me that their more modern CanBus systems have finally gotten there but the first ones with no way to tune them were said to be pretty bad.

As far as the noise goes a factory motor is annoyingly loud but dropping the motor and splitting the case to insert some John Deere Cornhead Grease quiets it to no more than any of the other popular motors, including the Brose. As tested riding side by side with Levo's. The factory only puts a small amount of white lithium grease in which isn't sufficient.

My biggest complaint with the M620 is its weight. At least it is distributed in the best possible spot down low and centered in the frame. Can be hard to get over logs though!

Also I found switching to 150mm cranks to allow for a higher cadence helps to keep the motor spinning more efficiently.

Not sure what to say about the above posters experience with his in relation to his Bosch but that hasn't been my experience at all.
 
Totally agree on the UART requiring tuning and some grease... Though I went with Mobile SHC100 and it didn't seem to quite as much as I would have liked. But I also didn't over_pack as I see many did with the Cornhead so that might be the difference.
I'm very happy with the tune and it's very efficient after tweaking... It does exactly what I want.
Have to agree with you on the weight as well. . .
 
Ravi:
"4) Controller programmability either from OEM or at the end-user level "
999/1000 or more riders do not attempt the HIT OR MISS " Programming controllers "
I toiled 32 years on computers ( Main frames to SUN to APPLE ETC ... ) as a Mechanical Engineer and when I needed a Programmer
I CALL THE HELP DESK.
Thanks for asking
Mike
 
Ravi:
"4) Controller programmability either from OEM or at the end-user level "
999/1000 or more riders do not attempt the HIT OR MISS " Programming controllers "
I toiled 32 years on computers ( Main frames to SUN to APPLE ETC ... ) as a Mechanical Engineer and when I needed a Programmer
I CALL THE HELP DESK.
Thanks for asking
Mike
But it sure is nice to be the 1/1000 😊
 
Though I went with Mobile SHC100 and it didn't seem to quite as much as I would have liked. But I also didn't over_pack as I see many did with the Cornhead so that might be the difference.
I only put in a heaping tbsp of Cornhead in mine and the others I have done that replaced about a half a tsp. Those that fill the area with grease are in for problems I feel as there is no real gasket between the gearbox and the controller?
 
I've recently changed mine from Mobile SHC100 to Cornhead. The SCH100 grease has a greater viscosity to it. On checking my motor the SHC100 had held its consistency well. No signs of any getting to the controller. Everything looked pretty good to be honest.

In comparison, the Cornhead grease is clearly less viscous than the SCH100. I do have some fears that if you packed out it could get through to the controller side, and so I have used much less grease than I did before. Mind you in testing I do think it probably is quieter than the SCH100 grease! But we wait and see how it hold up long term.
 
Well there's definitely some differences between the two. Mobil SHC100 is NLGI #2 where Cornhead is NLGI #0. The yellow grease that was in there from Bafang was at least as thick as the Mobil.
Grease-Consistancy-Chart.jpg
Additionally Cornhead grease states that it liquifies in use and that's what gave me some pause since there is no seal between the gearbox and controller.
This all said the thinner NLGI 0 greases are typically used in gearboxes whilst the thicker NLGI 2 are typically a bearing grease. But then most gearboxes are sealed and don't also encase electronics.
Now I'll add this from Mobil...
• Mobilith SHC 100 is an antiwear and extreme pressure grease primarily recommended for higher speed applications such as electric motors, where reduced friction, low wear and long service life are required. It is an NLGI 2 Grade / ISO VG 100 grease with a synthetic base fluid. Its operating temperature range is -40º C* to 150º C.
 
Back