Make: Fechters' Braking Regen Add-on for BLDC Motors

Joined
Jan 9, 2007
Messages
772
Location
Up where the air is fresh and clean
X-Large thanks to Fechter for conceptualizing and generating this circuit/diagram. It is located here..
http://endless-sphere.com/forums/files/brushless_regen_controller2_208.jpg

brushless_regen_controller2_208.jpg


I stared at this for a long while, struggling to understand how it might work. Having never tackled building a circuit quite this complicated -I tried looking at it from an electron's point of view.. then it began to make sense.

The really cool part that I visualized was the lucky electron that happened to be floating about a foot away from the ground -just as this machine is coming right at it. Then, being drawn through the wheel, this circuit, then landing on a bed of lithium. Fun.

I've already got the requisite "Cheap" CT-660B9 36V-40A brushed controller that should theoretically work with the 36V-20A C-lyte controller and 404 hub motor that I bought recently. Realizing that I might be half-way there, and have enormous respect for diagram's Author, why not take the leap (OK, broad step) of faith and build it?!

I printed this diagram and took it with me to work; which happens to be the local RadioShack that I manage, and filled-out the rest of a

PARTS LIST:

- RS# 276-1661 (Pkg. 4, bought 2) Rectifier Diodes. Rated @ 6A, 50V. There will be three in parallel on either side of the gates, so I think that these should work fine :roll:
:!: (Edit: It turns out that paralleling diodes will not parallel their outputs, at least not for this application. I decided not to wad-up this post because it is a good exercise in how NOT to make this circuit) :!:

pRS1C-2160501w345.jpg



- RS# 270-283 Project Enclosure with PC Board. 3.1875x2.125x1.375" is a good size for a Little Black Box to be.

pRS1C-2265051w345.jpg


- RS# 64-3025 Pkg. of 31 Assorted Grommets..
.. but Wallace was nowhere to be found :lol: (sorry)

I used a four conductor molded trailer connector, and a few black and red Anderson Power Pole's that had leads for patching into the motor/brushed controller/battery circuit.

Here is how it went together..

<a href="http://photobucket.com" target="_blank"><img src="http://i135.photobucket.com/albums/q133/Stevil_Knevil/BLDCRegenLBB/ConnectDiodes.jpg" border="0" alt="Photo Sharing and Video Hosting at Photobucket"></a>

<a href="http://photobucket.com" target="_blank"><img src="http://i135.photobucket.com/albums/q133/Stevil_Knevil/BLDCRegenLBB/DSCN0975.jpg" border="0" alt="Photo Sharing and Video Hosting at Photobucket"></a>

<a href="http://photobucket.com" target="_blank"><img src="http://i135.photobucket.com/albums/q133/Stevil_Knevil/BLDCRegenLBB/DSCN0982.jpg" border="0" alt="Photo Sharing and Video Hosting at Photobucket"></a>

<a href="http://photobucket.com" target="_blank"><img src="http://i135.photobucket.com/albums/q133/Stevil_Knevil/BLDCRegenLBB/DSCN0978.jpg" border="0" alt="Photo Sharing and Video Hosting at Photobucket"></a>

<a href="http://photobucket.com" target="_blank"><img src="http://i135.photobucket.com/albums/q133/Stevil_Knevil/BLDCRegenLBB/DSCN0974.jpg" border="0" alt="Photo Sharing and Video Hosting at Photobucket"></a>

<a href="http://photobucket.com" target="_blank"><img src="http://i135.photobucket.com/albums/q133/Stevil_Knevil/BLDCRegenLBB/DSCN0983.jpg" border="0" alt="Photo Sharing and Video Hosting at Photobucket"></a>

<a href="http://photobucket.com" target="_blank"><img src="http://i135.photobucket.com/albums/q133/Stevil_Knevil/BLDCRegenLBB/DSCN0984.jpg" border="0" alt="Photo Sharing and Video Hosting at Photobucket"></a>

<a href="http://photobucket.com" target="_blank"><img src="http://i135.photobucket.com/albums/q133/Stevil_Knevil/BLDCRegenLBB/DSCN0987.jpg" border="0" alt="Photo Sharing and Video Hosting at Photobucket"></a>

<a href="http://photobucket.com" target="_blank"><img src="http://i135.photobucket.com/albums/q133/Stevil_Knevil/BLDCRegenLBB/DSCN0988.jpg" border="0" alt="Photo Sharing and Video Hosting at Photobucket"></a>

<a href="http://photobucket.com" target="_blank"><img src="http://i135.photobucket.com/albums/q133/Stevil_Knevil/BLDCRegenLBB/DSCN0989.jpg" border="0" alt="Photo Sharing and Video Hosting at Photobucket"></a>

..I sure hope that it works as well as it looks. Does this look correct to y'all?


Thanks, -Stevil
 
I'm glad somebody is going to test the idea.

The workmanship on the diode bridge is excellent. Too bad they're only rated for 6A, which means the current limit in the regen controller better be 6A or they might not last long. If you don't give the regen controller much throttle, you can keep the amps down.

I was thinking of using one of these:
http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?Ref=66290&Row=157872&Site=US
which is rated for 35 amps / 100v and everything is in a nice hefty package. This would need to bolt onto a heatsink of some sort. A spot on the frame might work as a heatsink.

d63.jpg


I vaguely remember ordering a couple of them, but I haven't seen them show up, so I'll have to check.

If your diodes blow up, let me know and I'll see if I can find you a good bridge.

There's some on eBay sometimes. search on "bridge rectifier" and look for a 3 phase one.
 
you can get a 3 phase bridge that will handle 200V and 25A continuous/360A peak for under $17 (digikey 26mt20-nd) or use 3 dual common anode diodes, that way you could get Schottky parts with less drop for better regen efficiency and less heating in the diodes, but unless you get a deal on them they would be more expensive, since the common anode is not very popular

-bob
 
You could also use two very common single phase full wave bridges and tie them together. Half of one of them would be unused, but they are nicely packaged and have isolated heatsink mountings.
 
Doh! :oops: Thanks for the clarification guys.
Note to self: check-in with the pros BEFORE buying stuff and soldering it together.

Q: If I used a 35A / 100V 3 phase bridge rectifier @ 50V, would the rectify-able amperage be 70A?

Alternately, if it were a 25A / 200V bridge running @ 100V, could it handle 50A?

Am I thinking about this too hard, or not hard enough?
(Edit: My monster headache and inability to sleep would indicate the former)

Thanks again, -S
 
I don’t understand what this is about, but looks like you guys are on to something interesting here.

Really into those special gadgets that generates juice.
Not how much juice I can carry rather, how much juice can I produce, is key.

Keep it up!
 
being cheap (and a former RS mgr)
i'd buy two
FWB-352 $2.50 35a 200v
allelectronics.com
.
i got fired for making RS too much money. made everyone else look bad. they still owe me $550, gain from a 2nd inventory(they didn't believe mine) :roll:
sales declined 45% under new management LOL
 
Stevil_Knevil said:
Q: If I used a 35A / 100V 3 phase bridge rectifier @ 50V, would the rectify-able amperage be 70A?

Alternately, if it were a 25A / 200V bridge running @ 100V, could it handle 50A?

Am I thinking about this too hard, or not hard enough?
(Edit: My monster headache and inability to sleep would indicate the former)

Thanks again, -S

Since the current is shared by the diodes, the total current will be higher than the individual element rating. Since it's 3 phase, there's some formula you have to use, and I can't remember it.

If you use a pair of 35A bridges, the total might be around 50A. It won't be double.

You will need to bolt the rectifiers to some kind of heatsink, but I'm not sure how big it will need to be. Just a chunk of aluminum might be enough.

Here's how to hook up a pair of single phase bridges:
 

Attachments

  • bridge rectifier.jpg
    bridge rectifier.jpg
    46.5 KB · Views: 8,705
RatoN said:
I don’t understand what this is about, but looks like you guys are on to something interesting here.

Really into those special gadgets that generates juice.
Not how much juice I can carry rather, how much juice can I produce, is key.

Keep it up!

If you hit the regen while pedaling, the pedal resistance would go up and you would be generating back into the batteries. You could pedal with resistance going downhill and really get some amps. Sort of a human hybrid.



Stevil-
I forgot to ask: Does your cheapie brushed controller have current limiting? It might be good to lower the current limit to something less than the diode rating. This would also limit the amount of torque on the motor during braking. Too much and it might rip out the axle.
 
What would happen with this circuit and the bike if:

A FET in the brushed regen controller blew (would the driving wheel lockup?)

A diode in the bridge fried-n-died (could it short the motor wires, locking the wheel and blowing the primary controller?)
 
xyster said:
What would happen with this circuit and the bike if:

A FET in the brushed regen controller blew (would the driving wheel lockup?)

Yes. This could be ugly. It wouldn't exactly lock, the wheel would still move, but slower than the speed of the bike. With a hub motor, this might tend to over torque the axle.

xyster said:
A diode in the bridge fried-n-died (could it short the motor wires, locking the wheel and blowing the primary controller?)

Maybe. The resistance of the wiring would probably protect the main controller and the current limiter would kick in to limit the output.

Due to both of these possibilities, it would be a good idea to put fuses between the motor and the regen bridge diode.

I think normally you would want to limit the regen current to about 1/2 of the main controller's current limit.
 
fechter said:
Does your cheapie brushed controller have current limiting? It might be good to lower the current limit to something less than the diode rating. This would also limit the amount of torque on the motor during braking. Too much and it might rip out the axle.

It has a weak ramp-up. I would say that it is "efficient", in the sense that the current is very limiting. I don't think that there is much potential for this controller to rip anything apart.. but I have been wrong once in a row in building this circuit.

The controller itself is potted in that nasty epoxy, so it is not going to be very much fun getting to the PCB.

Should I start soaking it in acetone now?

:twisted: :D


-Stevil
 
No acetone, it sounds like it will work fine. You can always limit the throttle on the regen.

It would be a good idea to think about an interlock that will prevent the regen and main controller from being on at the same time.

I think the simplest way would be to tie into the brake lever switch. The regen controller has to "float" electrically, so you can't tie the grounds together. You could use a small relay, a double pole switch, or an optical coupler to kill one controller or the other. The relay or switch could short the throttle wiper on one or the other.

For experimentation, I was thinking you could have an actual throttle mounted backwards on the left handlebar to control the braking. In practice, it might be more user friendly to just have a fixed setting for the regen and make it on / off with the brake switch.
 
looks good Steve. Glad to see someone trying to put the regen system together. I have been really busy but hope to post more frequently once I get a bit of time.
 
You can just use a 12 volt auto regulator for the 3 phase bridge rectfier but you need to by pass the controller with a good switch.This way you can use regen at any speed while charging a un used bank of batteries in 12 volts parallel.Otherwise just get a Sevcon programable BLDC controller with regen.
 
Is there a controler for a Crystalyte 5304 motor that has regen? You mention "Sevcon"

I ike this project but it is a bit over my head and some have brought up the possibilities of catastrophic failures.

I love regen. I know all about the 10% return figures but I ride over a mountain and besides the braking from from my TF regen I watch my SOC increase coming downhill and in up and down mild hilly sections I watch my SOC remain steady.

So I use my Tidalforce regen effectively and need regen on my Crystalyte 5304 before I can commute over the mountain with it.

Are there controlers for 36, 48 or 72v that have regen?

Thanks,
Al
 
I'm not sure if Curtis has a brushless controller. The make an induction motor controller.

In industrial lingo, it's a PMAC controller you'd be looking for.

Sevcon has one, but it is 48v / 300amps.
http://www.sevcon.com/pages/millipakPMAC.html

I think you could program it for a lower current limit and it could work with a X5. You'd still be limited to 48v. It has regen.
 
Sevcon has one, but it is 48v / 300amps.
http://www.sevcon.com/pages/millipakPMAC.html

I think you could program it for a lower current limit and it could work with a X5. You'd still be limited to 48v. It has regen.

To convert some of those extra amps to volts, could you put a DC-DC voltage converter between the motor and the controller? Would it work in reverse such that regen was unaffected?
 
hi mark,
i built a regen braking system on a brushed hub motor, I started out same place as you, at the basics. The actual circuit turned out to be very complex. I landed up using relays and car alternator diodes, took me a few weeks to build it but it worked a treat.
The bike is being sent back to me soon so I should be able to do a good circuit diagram of it.
It turned out to be quite complex to do for a brushed motor, will be even harder for a brushless. I only got up to the stage of making the regen work, not actually puting it back into the batteries. That would add even more complexities. I can slow he bike down on very steep hills with the regen as its just going into heating the windings, and the amps are very large: 10km/hr 10amps, 20km/hr : 20amps etc, puting into the batteries will reduce is alot. I'm not actually keen to try and put the regen back into the batteries, the final circuit will be very complex indeed.
I have put some info on this page:
http://www.users.bigpond.com/solarbbq/regen/regen.htm
I never got around to writing in up properly, the last circuit diagram on that page is not the correct one but gives an idea of whats involved.
When I get the bike back ( a uni student was using for some research ) which will be in the next few weeks, I'll draw a diagram of exactly what i've got set up.
regen controllers for brushless motors are quite cheap these days, no need to build one, I would avoid all the headaches, pretty sure you'll find its going to land up more complex than first envisaged, if you work out a simple method would great though.
I've got a controller now that is programmable ( off the shelf and cheap ) but I haven't got the info on how to program it as yet, it would be possible to program it to be fully variable regen but once again a very big job.
At the moment it is just full regen when regen is applied via brake handle switch. Amount of current depends on rim size and speed of course, in a small rim (16-20") can get 20amps max going back into batts, on larger rims its reduced down to 10amps max or so, it also cuts off regen when voltage reaches 60volt to protect batts. designed for sla agm batts.
 
Ok, I've been reading this thread and a few like it on here.... and I think I've learnt something....possibly even enough to set up my own "plug brake" on my 5304.

I particularly like Richards concept of the brushed controller acting as the variable resistor for the plug brake.

My background is in the mechanical and aero disciplines so while I can mostly follow the logic, I sometimes get caught out.

I'm just wondering if the heat generated from the brushed controller set up is generating heat in the windings, or is it the controller that dissipates the heat...... is there a way of using this and feeding the heat to a heat sink?? I wan't the plug brake to be variable in it's intensity, but I don't want to charge the batteries in any way....

It might be a dumb question, but I've kinda mixed myself up with all of the different threads that I've read. I know that the actual mechanism for applying the plug brake is a hurdle, but I'll think of some way of doing that.

I've had an X5 controller blow in the past and somewhere along the way one of the FETs must have shorted because when the controller was plugged in, the wheel was very hard to turn... even harder as I tried to roll faster down a hill. While it was very inconvenient, I could see how it could also be useful... I just don't have the smarts or the time to experiment with it too much myself.

Any input would be greatly appreciated :)
 
on another thread here they are starting to play with a Kelly controller that has regen and is sized more for the X5. depending on options it can go up to 72V and a couple of hundred amps.

maybe just wait and see how regen works out for them.

the price is also competitive.

rick
 
? you don't want any current to go back to the batteries? Why?

If so, then this circuit will not really work as is. You would need to re-arrange it a bit and add a humongous resistor, which could just be a bunch of wire wrapped around the frame like Gary did here: http://endless-sphere.com/forums/viewtopic.php?f=2&t=2803

The resistor would need to go between the brushed controller and the output of the bridge. Even then, I think you would get some regen into the batteries, but not much.

EDIT:
I think this won't work either, you'll still get quite a bit of juice back to the batteries.

There might be another way, let me digest the thought.
 

Attachments

  • Brushless Regen adapter2.jpg
    Brushless Regen adapter2.jpg
    22.8 KB · Views: 5,843
? you don't want any current to go back to the batteries? Why?

Firstly I don't plan on braking for long periods of time (ie down long mountainous roads) but rather lots of short stop/starts. From what I've read you need to be going relatively fast on order to get a high enough voltage to charge the batteries.
Secondly, from what I can gather LiFePo4's have a bit of a charging regime that they must follow so I'd rather not mess with them and just keep them out of the circuit. I don't want to waste a set of batteries dor the sake of the brakes.

From my experience with the shorted FETs I know that the plug brake will work at relatively low speeds and I've felt its effects first hand. It's kinda like using the engine as a brake on an ICE. It's smooth, quiet and because there are no physical contacts, it means that the friction pads can be spared as an emergency back up or something.

I also like the look of the set up using the bridge rectifier and the resistor. This would be even better due to its simplicity but I'd have to find a variable resistor that's capable. Does anyone know of one that I could use? I can make a heat sink with the cooling fins and all at work if need be, but my limited vocabulary of electronic components means that I can't think of a variable resistor which would be suitable.
 
Back
Top