Author Topic: RE: Modifying the GM controller without touching the shunt  (Read 7116 times)

Offline Phil O Dendron

  • Confirmed
  • New Member
  • *
  • Posts: 11
Hi,

I got a Magic Pie II kit with internal controller and I`m quite happy with it now.
Did some modifications to the internal controller that I found here in the forums:

-added thermal grease
-replaced the capacitors with better&bigger ones
-improved the pcb where high currents flow

Now I`m stuck with this thread:
http://goldenmotor.com/SMF/index.php?topic=4072.0

I can see that it is an external controller, print circiut board 'mx318'.
I got an internal controller, print circuit board `mx328'.

Can / should I do this modification on my internal controller before bypassing the shunt?
I cannot find the resistor R54 with 100Ohm on my pcb.
Was the design of the circuit modified that this mod is no longer necessary and helpful?

Thank you for your help.

Kind regards, phil.

Offline Bikemad

  • Global Moderator
  • Professor
  • PhD. Magic
  • ******
  • Posts: 5,523
Re: RE: Modifying the GM controller without touching the shunt
« Reply #1 on: May 06, 2013, 01:49:42 AM »
Hi Phil andto the forum.

Unfortunately GM don't disclose what the differences are between the circuit board revisions so it's difficult to say what changes have been made.

I have only ever modified the shunt on my internal controllers and I can confirm that reducing its resistance too much does kill the controller.  ::)

My original MKI Pie ran on a modified controller delivering ~60 Amps maximum for almost 12 months before I killed it while doing some stall torque testing. The repaired controller was subsequently tweaked a little bit more, but finally died whilst delivering a mere 97.75 Amps.
My current MPII had the shunt slightly modified before fitting the wheel 12 months ago and has been working fine ever since and regularly delivers ~50 Amps on a 7 cell LiPo pack and up to 58.75 Amps on a 14 Cell LiPo pack.
The only problem I have had was a blown 40A fuse when I used my Smart Pie's battery pack without thinking. ::)

I have found that modifying the shunt makes a big difference to the power and works very well for me provided I don't try to overdo it.

Alan
 
« Last Edit: July 03, 2017, 10:56:31 PM by Bikemad »

Offline Morgen 3Eman

  • Confirmed
  • PhD. Magic
  • ******
  • Posts: 620
Re: RE: Modifying the GM controller without touching the shunt
« Reply #2 on: May 06, 2013, 04:12:57 AM »
Nearly 100 Amps at 48V!!!!  Good Grief, that has to be fun:)

TTFN,
Dennis

Offline Phil O Dendron

  • Confirmed
  • New Member
  • *
  • Posts: 11
Re: RE: Modifying the GM controller without touching the shunt
« Reply #3 on: May 07, 2013, 12:06:39 PM »
Hi Alan,

thanks for your help and accepting me posting in the forums.

Nice performance on your Magic Pie builds!  8)

I bridged about one third of the shunt, results were about 40 amps peak and, at lower speeds, about 25 amps continous
with software settings 55 amps peak and 30 amps continous.

with my diy limn battery made of 14s7p samsung icr 18650 cells it made about 2100 watts peak.
I'm very happy with that!   :)

The bad things:
The bearing lost a bit of the grease because of the temperature inside.
I can live with that, maybe get some different "heavier" grease for it

I managed to burn the wiring when climbing a hill (10 minutes about 1500 watts).
It burnt outside the motor inside the black rubber box where it splits from one cable
into the signal and the battery cable.

Maybe the connection was not good from start, now I'm wiring a little thicker battery cables
(2,5mm² instead of 1,5mm²) and have to redo the signal cables, too.

Will post pictures when finished, hope that the controller survived.
The pie was working till I stopped because of cut-outs and a little smoke from the cable.   ;D

Cheers, phil.

Offline Bikemad

  • Global Moderator
  • Professor
  • PhD. Magic
  • ******
  • Posts: 5,523
Re: RE: Modifying the GM controller without touching the shunt
« Reply #4 on: May 07, 2013, 03:02:38 PM »
Phil, I live in what I consider to be a very hilly area, but I don't have any hills that would take as long as 10 minutes to climb on any of my typical rides.

I imagine the hill that cooked your wiring must be either very steep, or at least 2.5 miles long, to take 10 minutes to climb at 1.5kW.
Perhaps it's just the combined weight of those 98 LiMn cells! ;D

I too hope that the controller has been able to survive the ordeal.



It is said that "all good things must come to an end", perhaps that's why these controllers keep dying on us just as we've finally managed to make them "good"?

Nearly 100 Amps at 48V!!!!  Good Grief, that has to be fun:)

Dennis,

"Fun" is a bit of an understatement, I'm now convinced that both the adrenalin rush and the "EV grin" are directly related to the number of Amps being consumed.
One exception being a dead short across the power connector of an unfused LiPo pack, which also consumes many Amps (and produces a rather large adrenalin rush) but fails to come up with any sign of an "EV grin" whatsoever!


Alan
 
« Last Edit: July 03, 2017, 10:54:28 PM by Bikemad »

Offline Phil O Dendron

  • Confirmed
  • New Member
  • *
  • Posts: 11
Re: RE: Modifying the GM controller without touching the shunt
« Reply #5 on: May 09, 2013, 06:58:44 PM »
Thanks for crossing fingers, the pie is on the road again and pushing as hard as before  ;D

The trail I was riding is about four kilometers long, the top is about 1000 meters above sea level.
I think that the start is about 500 meters.
Yes, there are lots of hills & mountains here in Austria.

I used an audio cable (2x 2.5mm², for speakers) for the battery wire and a computer network cable for the signal wire.
Could not fit a bigger battery cable through the axle and bearing.

Having the pie open I cut a thicker aluminium plate in the shape of the original heatsink
and mounted it over the standard plate. Should help a bit for short term excesses  8)

The 30 amp fuse of the battery pack melted a hole into the plastic housing, but is still working.

Will be more careful with my pie now, I`m happy with the performance now and don't want to break something again.

Cheers, phil