MacGuyver_Season1
100 mW
Luna got back to me and said their BBSHDs come with the better mosfets and rotor now, so I went ahead and ordered a new “bare bones” motor/controller. Will report back progress. Thanks everyone.
How To, 50A BBSHD Controller Mod
- Thread starter Ebikes4Real
- Start date
MacGuyver_Season1
100 mW
Temporary mockup with amps set at 27 (x2 with shunt), new motor core w/ new mosfets and rotor. 61 mph bench test under no load…surprised battery could support. More testing to come…
MacGuyver_Season1
100 mW
What have I created?! Soldering job was less than perfect for sure, likely used more solder than needed. Perhaps increased resistance even more? This is with programming set at 30A (theoretical 60A). Display is only reading 0.5A and maxes out around 4A during acceleration? Again, under no load other than wheel/tire so I know not real world results. Heat sinks, PEEK gear, Mobil 28 re-lube next steps. Is there a high-current in-line volt/amp/watt meter to see what is actually being pulled from the battery? MPH seems way too high. Again I have 14sP5 Panasonic GA battery which I thought had a max BMS of 50A(?), 42T chainring, 11T highest gear on cassette, 29er. Raining here today, not about to test this thing on the road yet until better conditions. Yikes.
MacGuyver_Season1
100 mW
Efffff
That's the problem with shunt mods. The controller then has no idea how much current is flowing, and can no longer protect anything from overcurrent. 
MacGuyver_Season1
100 mW
Well, very surprised by that. Thought battery would be weakest link, not core. Did that with original nylon gear as well, which I thought would be somewhat of a buffer. Lessons learned I guess. Looking at heat sinks, fish tank temp probes, potentially a larger/wide-range cassette. But….it was awesome for about 10 minutes.
MacGuyver_Season1
100 mW
Also, FYI tested it on bike stand. Would run wide open for 20 minutes (or more) under no load with zero rise in temps of case (controller or stator side). I obviously bogged it down in too high of a gear under load to fry windings like that. So it can handle the current, just not heat/stress/low rpm. Argh, back to drawing board…
MacGuyver_Season1
100 mW
Hmm…42T chainring (lowest I can go for chain line offset) with a 50T 1st gear? Should spin up fast. Looks aluminum and not steel though. Of course rear derailed not spec’ed that large but see people using an extra long B screw. Starting to understand how this motor works (the hard/expensive way).
Whatever gets overloaded is going to be stressed, damaged, or destroyed. The first parts are the ones that generate the most heat and can't get rid of it as fast as it's generated. That is usually the motor windings and controller FETs.
BMS FETs are next, and when these fail they usually fail shorted, so the BMS is then stuck on, undetectably by the user, and the cells are no longer protected against any problems such as overcharge, overdischarge, overcurrent, overheating, etc., that can lead to a fire. One way to test this is to drain the pack (riding, etc) until it should turn off because it's empty. If it doesn't, the BMS has been damaged and can no longer protect the pack. Whether cell damage has occured as well it's tough to know.
When you do a typical (even partial) shuntmod, you are effectively removing the shunt and replacing it with a zero-resistance wire, so there is no longer any voltage across it during even the highest possible current flow before things explode, and the controller has no idea that it is even supplying any current, so it will never attempt to limit it.
That means that the FETs are handling far more current than they're designed for, as are the board traces, wiring, and the phase wires and motor windings, and so is everything that is supplying that current to the controller (battery connectors and wiring, BMS, cells, interconnects, etc).
If you have a very-low-resistance measuring device, better than things like the DE-5000, you can measure microohms to see what a modded shunt actually reads afterward, vs before. Most likely you'll see something in the 1-5milliohm range before a mod, and a few microohms or less afterward. That would be around a thousand times less resistance, so a thousand times less voltage across it for the same current, so a thousand times less feedback to the controller to tell it what is happening, so a thousand times more current allowed.
Thankfully the system wiring, connectors, motor windings, FETs in controller and BMS, etc., all have enough resistance in them so you don't actually get a thousand times the current, or most shunt-modded systems would just explode or catch fire as soon as the throttle was touched. :lol: But so much more current can flow that it's easy to damage or destroy parts in a shuntmodded system that aren't designed to handle that much current.
No load is just that--no load.
Without a load you won't draw any more current than it would have without the shunt mod, which is usually a couple of amps. The system will behave exactly the same with or without a shunt mod in this situation.
Retrorockit
1 kW
That looks like the ver.A rotor. Square cut gear right up to the laminations. I would say replace the whole motor assy. with the ver. B model. The shaft will be on the list of things to break if you don't. I broke mine at stock 1500W doing wheels up launched testing a chain guide. You need the better one anyway.
MacGuyver_Season1
100 mW
No it’s actually version B. I have an original version A too. I‘ve actually since swapped out version A windings into version B motor housing and reinstalled version B rotor (so bottom bearing size matches and still maintain strength of version B rotor). Working with no error codes so far, but want to put more thermal/current/load protection in place before I just go out and burn windings again. Looking at heat sinks, etc. Wondering if cpu thermal paste could be applied between rotor and coils? That seems to be where a ton of friction and heat is generated.
Last edited:
MacGuyver_Season1
100 mW
*Technically not “Version A” with the Anderson phase connectors, but the older B version with the rotor which still had the issue of breaking. Circa mid-2016.
Last edited:
t3vio
1 µW
Did the 45A shunt mod. My controller had almost no thermal compound, fixed that. I am using 52V, a temp monitor and a peek gear. Controller is running surprisingly cool. Thank you for the ideas and infos. I also want to give something back to you. If you are using a eggrider display you can adjust the amp display in the calibration menu. Go to - Display advanced settings - CURRENT CALIBRATION. I use a 30A controller, so I calibrate the amps display with the factor of 1,5. This allows the eggrider display to show the correct amps.
MacGuyver_Season1
100 mW
Well, just jumped into the deep end and ordered a High Voltage kit w/ BAC855 controller.
Just make sure you get a full password/etc for using the setup software (BACdoor?).
Unless things at ASI have changed, if you don't get your own personal "forever" password from ASI, you may not be able to change any settings to tune things, correct problems, etc. People have had issues with some dealers giving them the deaier password, then ASI turns it off and they're stuck with a controller they cannot change settings on anymore.
Unless things at ASI have changed, if you don't get your own personal "forever" password from ASI, you may not be able to change any settings to tune things, correct problems, etc. People have had issues with some dealers giving them the deaier password, then ASI turns it off and they're stuck with a controller they cannot change settings on anymore.
MacGuyver_Season1
100 mW
Hmm, interesting. Will look out for that, thanks.
Retrorockit
1 kW
I've put CPU paste between the coils and the cover to get a wet joint there. Putting paste between the rotor and coils won't work because it's moving. There is a ferrro fluid mod that's done on hub motors that put oil with metal in it onto the magnets on the outside, to cool the rotor on the inside. But because of the higher motor speed the friction loss and heat gain is too much on mid drives. Luna has the stronger PEEK plastic gears in stock again.
BBSHD Silent Peek Gear (Nylon Gear Replacement)
Distributors and builders of life changing electric bikes.
lunacycle.com
You can look at the fluid mod at Grin Technology.
Statorade - Grin Products - Product Info
Statorade is a specially formulated ferrofluid that improves the thermal performance of direct drive hub motors. Just a small amount added to the motor magnets can double the thermal conductivity of the core, allowing the hubs to be driven harder and longer before risk of overheating.
ebikes.ca
Looks like mine. The magnets actually came loose and flung off the rotor!
tomjasz
1 GW
The BBSHD peanut butter hysteresis is for those overvolting. In typical use plastic gear failure is rare. In 6 years of parts sales the plastic gears were a poor seller. I have backup gears for all of my BBSxx series motors, but riding with sensible settings I've never replaced mine in any motor. Biggest failure with BBSHD was the first version rotor shaft breaking.
Bengy22
100 W
- Joined
- Dec 31, 2019
- Messages
- 164
I wouldn't say just for overvolting, more so over amping. Running higher voltage like 72v means you need less amps for the same wattage so less heat which is what causes the peanut butter failure. I would say the shunt mod controllers are the most likely to experience it, lower voltage with double the amps, stock inefficient bafang controller, no proper thermal throttling. But also from what I've heard easiest way to wreck the gear is to stick the bike in a really small cog in the rear and hammer the throttle from a stop.
I've been running my aftermarket controller at 52V and 65A on the stock gear for a year now still strong, and I believe it's because of the reasons I listed above
I've been running my aftermarket controller at 52V and 65A on the stock gear for a year now still strong, and I believe it's because of the reasons I listed above
How is that possible for an IPM rotor? Other people should have also reported this by now then.
Why has nobody tried to tap some threads in to heat spreader to bolt it thight against the housing like Toseven DM-01 does? Really weird inefficient design from Bafang in this regard. Wasn't paying enough attention to this before since everyone is in so much awe about this motor. Caring more about cooling of windings might be more rational since their resistance increases more with heat than mosfets though.
uhmwpolyethylene
100 W
What are peek gears actually made of? It just says it has some carbon mixed in , are they plastic or something else. Like some kind of heat resistant aramid or something? Or ceramic?
polyetheretherketone (PEEK)
industrial.vestakeep.com
3dprint.com
www.victrex.com
VESTAKEEP® PEEK - Evonik Industries
Gears move the world and have changed it. Plastic gears are superior to metal gears in many aspects and are conquering new applications thanks to temperature-resistant high-performance plastics such as VESTAKEEP® PEEK.
industrial.vestakeep.com
3D Printing Chemistry 101: The Molecular Makeup of PAEK, PEKK and PEEK Plastic - 3DPrint.com | The Voice of 3D Printing / Additive Manufacturing
If you’re anything like this author, you got a C- in high school chemistry and never looked back. With a newfound interest in the topic, I’m hoping to revisit the...
3dprint.com
PEEK Gear Solutions - Victrex
VICTREX™ Gear Solutions include gear design, prototyping, testing and state-of-the-art injection molding manufacturing capabilities using VICTREX HPG™ polymers and help our customers gain a competitive advantage.
Has anyone done this mod on a bbs02b controller? I am thinking about running 60v 40 amps. Is the resistor location and values the same as the bbshd? Also would like to hear where to place the thermistor. Do you just cut a hole in the pirring and put it in there?
