[EVDL List]

I expect that in a *stalled* motor, there is no back EMF
and thus the output *voltage* of the (AC) controller is
very low, because it will run either artificially limited
or at max motor current - which will command a low voltage
into the motor without back EMF. At high RPM the back EMF
will increase, so the required motor voltage will increase,
but if I am not mistaken I remember that a "full throttle"
(max motor amps at 250A from my AC controller) would draw
only 25A from the pack, so apparently the AC voltage that
could increase to about 160V RMS (312V nominal pack) was
around 16V RMS at that time.
This means that even if it would produce DC (it did not,
as I have ACIM and you could hear the swish-swish from
the motor due to the rotating -slipping- field) then
at a *stalled* motor and field the DC peak would be about
57% duty cycle (the companion transistor delivering 43%)

Regards,

Cor van de Water
Chief Scientist
Proxim Wireless Corporation http://www.proxim.com
Email: xxx@xxx.xxx Private: http://www.cvandewater.com
Skype: cor_van_de_water XoIP: +31877841130
Tel: +1 408 383 7626 Tel: +91 (040)23117400 x203

-----Original Message-----
From: xxx@xxx.xxx.edu [mailto:xxx@xxx.xxx.edu] On
Behalf Of Morgan LaMoore
Sent: Thursday, April 26, 2012 6:27 PM
To: Electric Vehicle Discussion List
Subject: Re: [EVDL] How To Burn Up Your DC Motor

On Wed, Apr 25, 2012 at 7:11 PM, Lee Hart <xxx@xxx.xxx>
[quote]wrote:
>> A minor point of correction is that all of the switches in an
>> inverter are pulse-width modulated, even if tasked with delivering an

>> effective DC current to just one phase, so they are never 100% on. In

>> some ways this is worse, though, as it means there are switching
>> losses on top of the conduction losses.
>
> That's true, though the duty cycle will be essentially 100% for one
> transistor, and 1% for its twin.

That enormously depends on the method of pulse width modulation.

A classic trapezoidal PWM scheme will have a duty cycle near 100% like
you describe.

In a simple sinusoidal implementation, 50% duty cycle will likely be
used as an artificial neutral point. With a low output voltage, all duty
cycles are near 50%; duty cycles only approach 0% or 100% when
outputting near maximum AC voltage, which would only happen at high
frequency.

Some space vector modulation routines also keep the artificial neutral
near 50%, so they also only approach 100% duty cycle at high frequency.

A more clever space vector modulation scheme may always keep one of the
six transistors at 100% duty cycle to eliminate switching losses in that
phase; in that case, they would have to consider the effect of stall on
the transistor current ratings more. Hopefully if they are optimizing
their algorithm to reduce switching losses, they will also think about
these types of problems.

-Morgan LaMoore

_______________________________________________
| Moratorium on drag racing discussion is in effect.
| Please take those discussions elsewhere. Thanks.
|
| REPLYING: address your message to xxx@xxx.xxx.edu only.
| Multiple-address or CCed messages may be rejected.
| UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
| OTHER HELP: http://evdl.org/help/
| CONFIGURE: http://lists.sjsu.edu/mailman/listinfo/ev

_______________________________________________
| Moratorium on drag racing discussion is in effect.
| Please take those discussions elsewhere. Thanks.
|
| REPLYING: address your message to xxx@xxx.xxx.edu only.
| Multiple-address or CCed messages may be rejected.
| UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
| OTHER HELP: http://evdl.org/help/
| CONFIGURE: http://lists.sjsu.edu/mailman/listinfo/ev

[EVDL List]

[quote]Cor van de Water wrote:
> I expect that in a *stalled* motor, there is no back EMF
> and thus the output *voltage* of the (AC) controller is
> very low

Agreed. The situation is analogous to a DC motor at stall. If the driver
commands maximum torque, the controller operates at a very low duty
cycle to delivers a very low voltage but at maximum current. The PWM
duty cycle might be a few percent, with the transistor on 3% of the time
at full current, and the freewheel diode on 97% of the time, also at
full current.

With the brushed DC motor, one commutator bar is "parked" under each
brush. The length of time you can do this is limited by how long it
takes a commutator bar to overheat. It's not long! Just a few seconds.
The bar is pretty small, and heats up fast!

With a PM AC (synchronous) motor, worst case is also at stall. Again,
you have to maximum DC current to one phase winding. Again, the inverter
transistors will operate at a very low duty cycle. The "on" transistor
will only be on maybe 3% of the time, but the "freewheel" transistor
will be on 97% of the time. It's the one that can overheat, because it
never sees anything close to this much current on a normal basis.

With an AC induction motor, stall is *not* the worst case. At stall, the
slip means that normal 3-phase AC waveforms are being applied to the
stator windings. Just at a very low frequency (the slip frequency), and
thus a low voltage and high current. The heat isn't concentrated in any
one transistor; it is distributed between them, same as it would be at
any other speed.

With these AC motors, the one transistor that gets hit with high current
during these situations is the one that will fail. Transistors are also
small, and can fail quickly when overloaded like this.

Worst case for the induction motor is when it is rotating *backwards* at
the slip speed, and you are commanding maximum forward torque. For this,
you need to apply the same DC to one phase winding that you would for
the PM AC motor at stall.

> Hopefully if they are optimizing their algorithm to reduce switching losses, they will also think about
> these types of problems.

Yes. A novice building something from a hasty manufacturer's appnote is
likely to miss this worst case. An experienced professional design for
it, as he's probably been burned before by a drive that had unexpected
failures when the motor was commanded to produce high torque near stall.

--
*BE* the change that you wish to see in the world.
-- Mahatma Gandhi
--
Lee A. Hart, 814 8th Ave N, Sartell MN 56377, leeahart at earthlink.net

_______________________________________________
| Moratorium on drag racing discussion is in effect.
| Please take those discussions elsewhere. Thanks.
|
| REPLYING: address your message to xxx@xxx.xxx.edu only.
| Multiple-address or CCed messages may be rejected.
| UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
| OTHER HELP: http://evdl.org/help/
| CONFIGURE: http://lists.sjsu.edu/mailman/listinfo/ev

[EVDL List]

[quote] Lee Hart <xxx@xxx.xxx> wrote:
> ...
> With a PM AC (synchronous) motor, worst case is also at stall. Again,
> you have to maximum DC current to one phase winding. Again, the inverter
> transistors will operate at a very low duty cycle. The "on" transistor
> will only be on maybe 3% of the time, but the "freewheel" transistor
> will be on 97% of the time. It's the one that can overheat, because it
> never sees anything close to this much current on a normal basis.
> ...

Lee,

This usually isn't how inverters modulate. Since both positive and
negative voltages are required (line-to-line), outputting "0V" is done
by setting all 6 transistors to 50% duty cycle (minus dead time).

To put out 10% voltage from A to B, the A phase is modulated 55%/45%
and the B phase is modulated 45%/55%. To put out -10% voltage from A
to B, the duty cycles are switched.

Any given transistor (or parallel freewheel diode) only approaches
100% when the inverter approaches maximum AC amplitude. Since that
only happens at high frequency, it isn't applicable here.

There are other modulation algorithms that use a different method to
output three duty cycles from three phase voltages, but most of the
simple sinusoidal algorithms keep the 0V/neutral duty cycles near 50%.

-Morgan LaMoore

_______________________________________________
| Moratorium on drag racing discussion is in effect.
| Please take those discussions elsewhere. Thanks.
|
| REPLYING: address your message to xxx@xxx.xxx.edu only.
| Multiple-address or CCed messages may be rejected.
| UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
| OTHER HELP: http://evdl.org/help/
| CONFIGURE: http://lists.sjsu.edu/mailman/listinfo/ev

[EVDL List]

Partial retraction of my statement:

The Leaf does, indeed, creep forward (or backwards) while pressing the brake
and accelerator at the same time. I tried it a few times on some hills.
Very nice behavior. I don't know if they have any sort of limiting to
prevent overheating and I don't plan on experimenting to find that out

I have not tried parallel parking on a steep hill yet, though. There was
something happening that made it tricky but apparently it isn't the
brake-accel interoperation. More info if I discover something.

Peri

-----Original Message-----
From: Peri Hartman [mailto:xxx@xxx.xxx]
Sent: 23 April, 2012 7:49 AM
To: 'Electric Vehicle Discussion List'
Subject: RE: [EVDL] How To Burn Up Your DC Motor

The Leaf seems to have a bit of this problem. If parallel parking on a
hill, and you're using the brake, I think it's ignoring the accelerator
pedal till you take you foot off the brake. This results in a small surge
causing a small bump into the vehicle in front (or back). Not sure of the
best way to solve this, but something to think about.

Peri

-----Original Message-----
From: xxx@xxx.xxx.edu [mailto:xxx@xxx.xxx.edu] On Behalf
Of Mark Warner
Sent: 23 April, 2012 7:02 AM
To: Electric Vehicle Discussion List
Subject: Re: [EVDL] How To Burn Up Your DC Motor

I would suggest an "override" feature that could be triggered by a momentary
switch that the operator has to push if he wants to do something out of the
norm, like drive the EV up onto a trailer. During normal operation, there
would be normal stall protection. Voila, the best of both worlds.


[quote] Cor van de Water <xxx@xxx.xxx> wrote:

> Can the stall protection be configurable, so you have it default "off"
> but if someone needs it, it can be configured "on" and that person
> must take care when loading the car on a trailer that he may instantly
> lose traction and catapult backwards due to this feature?
> (the safe way would be to use a winch or at least a protection that
> can be ratcheted so the car does not uncontrolledly leave the
> trailer...)
>
> I deal with products being deployed in a wide variety of environments
> with conflicting needs on a daily basis, so this is one of the escapes
> - keep it compatible (default = old behavior) and allow it to be
> "upgraded" to the new feature...
>
> Regards,
>
> Cor van de Water
> Chief Scientist
> Proxim Wireless Corporation http://www.proxim.com
> Email: xxx@xxx.xxx Private: http://www.cvandewater.com
> Skype: cor_van_de_water XoIP: +31877841130
> Tel: +1 408 383 7626 Tel: +91 (040)23117400 x203
>
> -----Original Message-----
> From: xxx@xxx.xxx.edu [mailto:xxx@xxx.xxx.edu] On
> Behalf Of Jeffrey Jenkins
> Sent: Monday, April 23, 2012 5:23 PM
> To: xxx@xxx.xxx.edu
> Subject: Re: [EVDL] How To Burn Up Your DC Motor
>
>
> Bruce Lawton wrote
> > ... I parked the car in the garage, set the brake and pressed the
> > accelerator for at least 2 seconds, probably 4 or 5 seconds. The
> > battery delivered 25A during this. The brake won, I turned
> > everything off and went inside. The next day, as the car began to
> > roll, there was
>
> > a tick-tick sound. I'd overheated the commutators and lifted 4 bars.
> > A
>
> > LOT of motor amps had heated it when it wasn't turning.
> >
> > The Zilla has stall protection; my Soliton apparently does not....
> >
>
> As of version 1.5.1 of the Soliton code that is correct - we do not
> prevent you from pushing current through the motor even when stalled.
>
> We did contemplate adding this function at one time but did not
> because it has a high annoyance potential (ie - stall detect
> triggering while trying to drive up onto a trailer would infuriate me
> - I certainly wouldn't be "thanking my controller" for its
> overzealousness), and, frankly, we feel it is the motor manufacturer's
> responsibility to provide adequate specifications and warnings for
> their products, rather than it be entirely up to use to protect their
products from abuse.
>
> I should also point out that you are the only person I am aware of
> that has damaged a motor by stalling it in the 3.5 years we've been
> selling Soliton controllers, so I'd say this is a very rare problem.
> In contrast, we have destroyed 7 motors ourselves by pushing too much
> current through them for too long but I rather suspect that no one
> wants us to lower our current ratings or otherwise thermally-cripple
> our controllers to protect against that kind of abuse...
>
> But if protecting against a stall seems really important to people -
> particularly people that already own our controllers - then we'll add
> it to the 1.6 release of the code.
>
> We are about to roll out an intermediate version, 1.5.2, to correct
> the bug in the start button function (failing to idle the motor) and
> add a couple of minor functions: block throttle when signal is above
> calibrated range and a somewhat convoluted reverse switch function to
> protect against reversing the field polarity when the motor is already
> spinning. Coincidentally, both functions are being added because
> someone did something dumb with their controllers.
>
>
>
> --
> View this message in context:
> http://electric-vehicle-discussion-list.413529.n4.nabble.com/How-To-Bu
> rn -Up-Your-DC-Motor-tp4577309p4580199.html
> Sent from the Electric Vehicle Discussion List mailing list archive at
> Nabble.com.
>
> _______________________________________________
> | Moratorium on drag racing discussion is in effect.
> | Please take those discussions elsewhere. Thanks.
> |
> | REPLYING: address your message to xxx@xxx.xxx.edu only.
> | Multiple-address or CCed messages may be rejected.
> | UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
> | OTHER HELP: http://evdl.org/help/
> | CONFIGURE: http://lists.sjsu.edu/mailman/listinfo/ev
>
> _______________________________________________
> | Moratorium on drag racing discussion is in effect.
> | Please take those discussions elsewhere. Thanks.
> |
> | REPLYING: address your message to xxx@xxx.xxx.edu only.
> | Multiple-address or CCed messages may be rejected.
> | UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
> | OTHER HELP: http://evdl.org/help/
> | CONFIGURE: http://lists.sjsu.edu/mailman/listinfo/ev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.sjsu.edu/mailman/private/ev/attachments/20120423/8e37ffdc/attac
hment.html
_______________________________________________
| Moratorium on drag racing discussion is in effect.
| Please take those discussions elsewhere. Thanks.
|
| REPLYING: address your message to xxx@xxx.xxx.edu only.
| Multiple-address or CCed messages may be rejected.
| UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
| OTHER HELP: http://evdl.org/help/
| CONFIGURE: http://lists.sjsu.edu/mailman/listinfo/ev



_______________________________________________
| Moratorium on drag racing discussion is in effect.
| Please take those discussions elsewhere. Thanks.
|
| REPLYING: address your message to xxx@xxx.xxx.edu only.
| Multiple-address or CCed messages may be rejected.
| UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
| OTHER HELP: http://evdl.org/help/
| CONFIGURE: http://lists.sjsu.edu/mailman/listinfo/ev