Ebike Using ABB VFD

antsrealm

10 µW
Joined
May 12, 2016
Messages
5
Hi,

New post here. Just saw something today that got me thinking whether this is practical or not. Using an ABB ACS355 variable speed drive to run a 0.75kW 3 phase motor to run an ebike.

This drive normally taking single phase in and giving 3 phase out can be powered with DC directly to the DC bus via the Brake terminals and then I assume will regen back to the battery when required. The DC bus voltage range is 270VDC - 325VDC.

It has rs232 connection that support Modbus RTU that I was thinking about interfacing to a raspberrypiw zero running node red or something like that which in turn can present an interface to a mobile phone via http. This will be able to display all the information on the VFD for example DC bus voltage (essentially battery voltage), rpm, faults, etc.

Thinking something like 80 - 100 18650's in series.

I'm a sparky and work in industrial sector so the drive and motor is easy to find which is what led me down this path.

Can anyone advise me on whether this is worth exploring further or not? I have access to an engineering workshop and was thinking this could be a fun project for the next little while.

Ideally I can get 60km/hr top speed with a range of around 50km preferred.

One of my concerns is the best way to prevent against electric shock at that voltage in the event of a winding failure to the chassis.

Thoughts?

Thanks.
 
Possible? Perhaps. Practical? Probably not.

No need to try and reinvent the wheel. Plenty of kit exists to do more than you're asking with much less time and energy invested. Not to mention at more sane voltages.
 
Ideally I can get 6km/hr top speed with a range of around 50km preferred.

One of my concerns is the best way to prevent against electric shock at that voltage in the event of a winding failure to the chassis.
I assume the 6km/hr is a typo..?
If So , what top speed ?..60km/hr ?

The BEST way to prevent high voltage issues,..is not to use high voltage,... as it is unnecessary. !
 
antsrealm said:
Thinking something like 80 - 100 18650's in series.
Which means you will be routing how many watts through each Lithium-Cobalt cell?
 
I haven't seen any bikes done with an induction motor, but it certainly has potential. Tesla uses induction motors. You can get a lot more power out of a motor running it way higher than the power line frequency. This means you could probably use a smaller motor, like something out of a fan and get 750w out of it (at a much higher than rated RPM).
 
Induction is the future of ebikes. Quality magnets are soon to become rare and expansive. Induction motors are a natural evolution toward lighter weight and higher power. The pandemic had made a consistent growth of the ebike market in western countries, thus bigger investments coming in R&D.

Building one for yourself today might be expansive, especially if it is built for high performance. But sure we’d like to see it.
 
Hi, thanks for the feedback. Yeah the first motivation is that I have access to these VFD's and induction motors essentially for free. Also I use them in my line of work so it's value adding to work with them in this less common application.

I think it has potential. I was thinking maybe the 80 or so cells in series and then possibly 2 of those packs in parallel. I might try it on the bench as a proof of concept. The only items that really look like a challenge at the moment is the following.

- Determining a safe way to disconnect supply if the motor casing becomes live. Some kind of earthing system.
- Determining the ideal motor size, winding configuration (assuming delta due to the single phase to 3 phase situation).
- Determining ideal pack size considering operating voltage and DC bus tolerances.
- Understanding how to monitor these cells from a bms perspective.
- Understanding the best way to charge a pack at this voltage.

Thanks,
 
LewTwo said:
antsrealm said:
Thinking something like 80 - 100 18650's in series.
Which means you will be routing how many watts through each Lithium-Cobalt cell?

I think it's roughly 4A load through the battery. 3000ma LG brown 18650's maybe. Probably run a pack in parallel as well I guess to get a bit more run time. Plus should not be pulling FLA all the time. Guess some R&D is needed to confirm that.
 
Another option I've looked at is to modify the VFD to work on a lower voltage. All the control circuitry is going to be 5v and 12v stuff. I've also seen motors that were rewound to work on 72v systems that look like your standard industrial motor. Some are used in high performance golf cars. Something the physical size of a 1/2hp induction motor can probably do over 20hp at high rpm. Sort of big for a bicycle, but might be good for a small motorcycle or ATV.
 
antsrealm said:
LewTwo said:
antsrealm said:
Thinking something like 80 - 100 18650's in series.
Which means you will be routing how many watts through each Lithium-Cobalt cell?

I think it's roughly 4A load through the battery. 3000ma LG brown 18650's maybe. Probably run a pack in parallel as well I guess to get a bit more run time. Plus should not be pulling FLA all the time. Guess some R&D is needed to confirm that.

Might be good for a few rides, but since you won't be able to balance the battery pack, the lifespan of the pack may be pretty short, relative to a pack with a BMS that allows you to balance once in a while.
 
antsrealm said:
LewTwo said:
antsrealm said:
Thinking something like 80 - 100 18650's in series.
Which means you will be routing how many watts through each Lithium-Cobalt cell?

I think it's roughly 4A load through the battery. 3000ma LG brown 18650's maybe. Probably run a pack in parallel as well I guess to get a bit more run time. Plus should not be pulling FLA all the time. Guess some R&D is needed to confirm that.

I think that you missed the point. It was not how much power that you would be drawing from each cell but rather how much power you are passing through each cell.
 
LewTwo said:
I think that you missed the point. It was not how much power that you would be drawing from each cell but rather how much power you are passing through each cell.
It doesnt matter Lew, each series cell only “sees” the system current (4.0A ?) and its own voltage (4.0v ?)
.. so 16 W for each cell...no big deal !
.. but i dont want to be the guy hooking up 100 of them in series ! (400v)
 
LewTwo said:
I think that you missed the point. It was not how much power that you would be drawing from each cell but rather how much power you are passing through each cell.

4A passing through a single cell is the same as 4A passing through a 100S series of cells, from the standpoint of any single cell.
 
Yeah the cell balancing is one issue and the charging is another. I was wondering if it could have 2 charging ports that essentially charges to halves of the pack separately to make it easier. The BMS I'm not sure about yet but would love to be able to do cell balancing in a cost effective way. How does Tesla manage there packs from a cell balancing perspective?

I was also considering the idea of several smaller / safer voltage packs that then plug into a bus so to speak to make it safer. Maybe 40V packs. Almost like a big miklwaukee battery that clips into a backplane. Since the current draw is low due to the hgher voltage then the backplane doesn't need to be that big.

Hmm maybe a few of the milwaukee 72V batteries with a clip in bus system (MXFCP203) Do a tear down on one of these and reproduce something similar & cheaper.
 
Would it be practical to do something like this. Build 4 x 72V packs (20S2P) and then series them up.

Pack Config Image Link
https://drive.google.com/file/d/1OAa-kmUmuRHuNi7Azq8MIOasZLFWxr_Q/view?usp=sharing

I could have the pi zero monitor the BMS healthy status and if any pack goes out of spec it can shut off the run signal to the drive and advise complete power down. Also if the BMS itself shuts off 1 of the packs that could also work as the DC bus will go out of spec and the VFD will shutdown / fault anyway.

Could charge the individual packs or pairs of the packs whatever is easier.

Just getting my head around this.
 
The now defunct Alta Motors made motorcycles with 350v batteries. You might see how they did it. But it might be why they went out of business.
 
Back
Top