View Full Version : AC-20 - which way is reverse, and forward ?

18 November 2012, 0705
Hi guys,

testing the motor on the bike this afternoon, I'm using the AC-20 with a Curtis 1238-7601.

When connecting the controller as proposed by Curtis here (http://curtisinstruments.com/index.cfm?fuseaction=Datasheets.downloaddatasheet&prodid=109), the wheel spins in the wrong direction.

So i just disconnected the FORWARD wire (white, pin 22) from the battery + and connected the REVERSE instead (yellow, pin 33).

That does indeed reverse the spin, BUT:
that made me think: on the other side of the AC-20, there is a centrifugal fan. As far as i know, this fan is meant to suck air out of the motor, but only when it spins in the correct direction. If you spin the other way, it does not really work, cause it's trying to push air back into the motor and this is probably no working for a centrifugal fan.

If above is right, it means there is a FORWARD and REVERSE direction on this motor, because of the fan ? That seems strange. So it means for me, i would need to move my driven sprocket on the other side of the wheel ?!?!?! :confused:

Or maybe Curtis is selling the centrifugal fan for the other rotation ?

Surely i'm missing something here.

18 November 2012, 0829
ok, maybe i was a bit quick to post.

I found this thread (http://www.elmoto.net/showthread.php?2332-AC-20-cooling) on the forum, and a picture of the fan:

As far as i can see, the blades extend outward in a straight line along the radius of the fan. I thought they were shape in a spiral.
So i guess if they are straight then the fan can spin in both directions, doing the same effect.
A spiral would not work in both direction.

False alert then.


18 November 2012, 0939
MAN! That pic sure looked familiar...lol

but... every time i look at it, i think there has to be a better design than the high/low pressure venturi effect

18 November 2012, 0940
Thank you for the heads up on the wire....
as soon as i string some batteries together, i hope to be right behind you

18 November 2012, 0947
when you have a minute, and if the back of your motor is accessible, if you could double check that the blades of the fan are really "radial", and not spiral.
just to be sure. Unless you remember exactly from the last time you dismantled it ?
I cannot ope mine when it's mounted in place.

ARC EV Racing
18 November 2012, 1232
The fan design is typical of an industrial stationary motor. It's supposed to take the air that comes through the vent on the cover and force it to the outside of the cover where it's directed down the outside of the motor. It's supposed the be air cooling the case.

It's not very effective really. And it does not matter which way it spins.

18 November 2012, 1332
Curtis doesn't build these systems per se, it's HPEVS that builds the motors and pairs them with a Curtis controller.

The motor is bidirectional. Because it's an Induction motor, here's no need to make it run better in one direction or the other (commutation is done electrically). The fan is meant to go either direction and the blades are straight outward from the shaft, not spiral.

You should be using the HPEVS wiring schematic:
The Curtis is Generic, the HPEVS is specific to your system.

I've taken these motors apart quite a few times (Just serviced one last week). If you have questions let me know.

18 November 2012, 1333
yeah i will get you a better pic too, the cover is still off....i stare at it with angry eyes...
I swear its working too

18 November 2012, 1336

18 November 2012, 1339
You should be using the HPEVS wiring schematic:
Thanks for the link, I'm using the paper copy of this one.

So all fine then .

18 November 2012, 1340
My gallery when i took the old AC15 apart (click on the pic of the fan):

http://www.evfr.net/coppermine/albums/1-16-10/normal_IMG_2457.jpg (http://www.evfr.net/coppermine/thumbnails.php?album=38)

Here's a pic of the difference in size of the rotor on an AC15 and AC20 (the longer one):

18 November 2012, 1341

Definitely radial.

18 November 2012, 1346
Thanks for the link, I'm using the paper copy of this one.

So all fine then .

You could also give yourself forward AND reverse. If it rotates backwards when you run it, hook it up in reverse by hooking the wire going to pin 22, to pin 33 instead. Pin 22 and 33 are the FWD and REV direction pins. If you wire as shown below, you should be able to do reverse:

Pin 9 is an easy way to disable. Usually used as a throttle switch, but you should be able to wire it so that it disables the controller easily.

ARC EV Racing
18 November 2012, 1421
There's a parameter on the controller called something like 'swap encoder direction' which I think is what you're meant to use if the motor runs backward while the forward direction is selected. Sorry I'm being vague but I've never has to use it, sure it's in the manual though.

Sent from my AZ210A using Tapatalk 2

ARC EV Racing
18 November 2012, 1449
Just checked and it's 'swap encoder direction' and 'swap two phases'. Part of the initial setup section.

Sent from my AZ210A using Tapatalk 2

18 November 2012, 1458
Ooops, yeah, that too....

Keep FWD connected and do the swap 2 phases/swap encoder direction (but you need a programmer, forgot if you bought one).

Good remembering ARCEVRacing !

18 November 2012, 1935
I was going to say, don't you just swap two of the three phase wires to achieve the same thing?

18 November 2012, 2350
I was going to say, don't you just swap two of the three phase wires to achieve the same thing?

thanks guys for all your feed-back
i thought about that yesterday, swapping phases, cause i used to run small brushless motors, and the only thing you needed to do was to swap any two phases. But they did not have an encoder.

ANd just reading the Curtis manual now (could have done it yesterday you'll tell me):

Swap Encoder Direction
Changes the motor encoder’s effective direction of rotation. The encoder
provides data used to calculate motor position and speed. This parameter
must be set such that when the motor is turning forward, the controller
reports back a positive motor speed.
Positive motor speed must be in the forward direction in order
for the emergency reverse feature to operate properly.

Swap Two Phases
If, after Swap Encoder Direction has been set correctly, the vehicle drives
in the wrong direction (i.e., drives forward when in reverse, and vice
versa), try changing the setting of the Swap Two Phases parameter. This
parameter has the same effect as physically swapping the cables on any
two of the three motor phase connections.

So if i get it right, i should not have to physicaly swap any phases, i can just play with these 2 parameters, making sure that when forward wire is connected, i actually have forward movement.


ARC EV Racing
19 November 2012, 0623
Yep, that should do it.

Obvious one but is it worth checking you have your motor - controller phase cables connected correctly before you change any parameters?

Sent from my AZ210A using Tapatalk 2

19 November 2012, 1126

changed parameters for encoder direction AND "swap phases" (software, not hardware) and now it spins forward when forward wire is connected.

All is good.

Now i can concentrate on this noise problem.

ARC EV Racing
19 November 2012, 1132
Electrical noise?

Sent from my AZ210A using Tapatalk 2

Nuts & Volts
19 November 2012, 1301
Electrical noise?

Sent from my AZ210A using Tapatalk 2

Chain noise he talks about in this post http://www.elmoto.net/showthread.php?2141-e-Cruiser-version-1-0-build-thread&p=33643&viewfull=1#post33643

ARC EV Racing
19 November 2012, 1413
Ah I see. Thanks

22 November 2012, 1245
Sorry not to see this thread before but I was offline , anyway to change the + - rpm sigmg just use Swap encoder and Swap phase settin in software , no need to change wiring.

later if you have a test bench you can use the Curtis autotune routine to optimize the motor/controller match.

BTW if you need some VCL I can write it