Root cause insight into the common BMW blower motor resistorfailures

On 03/21/2013 10:25 PM, jim beam wrote:
> On 03/21/2013 07:23 PM, Nate Nagel wrote:
>> On 03/21/2013 10:07 PM, jim beam wrote:
>>> On 03/21/2013 09:00 AM, Bimmer Owner wrote:
>>>> On Thu, 21 Mar 2013 08:14:12 -0700, jim beam wrote:
>>>>
>>>>> it's cheaper to just buy a new one.
>>>>
>>>> To be clear, that's what 99.99999999% of the BMW owners do.
>>>> But that's not the point of this thread.
>>>>
>>>> The point of this thread is to get a handle on WHY they are all
>>>> failing.
>>>
>>> i already told you - it's overheating. semiconductors don't like heat.
>>>
>>
>> Yeah, we know that.
>>
>> *why* is it overheating?
>
> because it's linear, retard. if you don't know what they means, ****
> off until you find out.

so "because it's linear" it by nature overheats to the point of failure?
Odd, I'm pretty sure that that controller worked initially on, well,
all of the vehicles in which it was originally installed.

The question is, is it overheating to the point of failure because the
designer cut things too fine (in which case designing a better part
would be the right approach), or is it because there's another issue
somewhere else *causing* a part that would otherwise have acceptable
service life to fail (in which case replacing it with a stock
replacement and fixing the underlying issue would be the most economical
thing to do)?


>
>
>>
>>>
>>>>
>>>> Specifically, how to figure that out is the question.
>>>
>>> knowing how the light bulb blew doesn't fix it.
>>
>> But it may make the replacement last longer.
>
> putting facts in front you dumb ass all day long doesn't make you any
> smarter.

IKYABWAI.

It would really be just like you to spend all day designing a more
robust controller, building it, watching it too burn up, and then
realize that the problem was something else, like a chronic problem with
dry fan motor bushings, windings dragging on the case, something like that.

But sure, don't check the obvious stuff, just go into your long-winded
sometimes technically correct and sometimes not babble, I know that
anything I say won't stop you anyway.

nate


--
replace "roosters" with "cox" to reply.
http://members.cox.net/njnagel

In article >, jim beam > wrote:
>On 03/21/2013 07:23 PM, Nate Nagel wrote:
>>
>> *why* is it overheating?
>
>because it's linear, retard. if you don't know what they means, ****
>off until you find out.

Nothing wrong with linear motor control, it's just inefficient and
produces a lot of heat. I used to work in a place with a 1.2 MW DC
motor whose field coil voltage was controlled by a couple rooms full
of cast-iron resistors. The resistance array lasted nearly 80 years
before the whole facility was taken down.

As long as you keep within the safe operating area of the semiconductors,
you're fine. If you exceed them, bad things happen. But we don't know
if the semiconductors are failing on these things, or if it's just
ordinary RoHS solder failures; the RoHS crap doesn't like thermal cycling
so well.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Bimmer Owner > wrote:
>On Thu, 21 Mar 2013 19:15:16 -0700, jim beam wrote:
>
>> that's why pwm is the better solution
>
>Does pulse width modulation cause radio EMI?

It can if it's not properly filtered. And the reason why I think beam is
correct about this being a linear control is that there is no filtration
in there.

My guess is that the cost of proper filtering and shielding makes the pwm
controller cost more than the linear controller in this case, which is
probably why they went the linear route.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

On 03/21/2013 07:53 PM, Nate "anosognosic" Nagel bleated:
<snip drivel>
>
> It would really be just like you to spend all day designing a more
> robust controller, building it, watching it too burn up, and then
> realize that the problem was something else,

um, no, it would be me building something that worked first time. and
that has stayed working.

but you wouldn't know that because you're an unspeakable hamster brained
retard that hasn't got the slightest clue the **** they're talking
about, much less the ability to click on a link and READ it.


> like a chronic problem with
> dry fan motor bushings, windings dragging on the case, something like that.

which are entirely ****ing different, and don't ****ing disappear after
the controller has been changed.


>
> But sure, don't check the obvious stuff, just go into your long-winded
> sometimes technically correct and sometimes not babble, I know that
> anything I say won't stop you anyway.

i wish something would stop you. like self-awareness.


--
fact check required

On 03/21/2013 08:08 PM, Scott Dorsey wrote:
> In article >, jim beam > wrote:
>> On 03/21/2013 07:23 PM, Nate Nagel wrote:
>>>
>>> *why* is it overheating?
>>
>> because it's linear, retard. if you don't know what they means, ****
>> off until you find out.
>
> Nothing wrong with linear motor control, it's just inefficient and
> produces a lot of heat.

right, but what's the point in using 140W to run a 60W motor? and
/certainly/ not when the devices to do so are so cheap and abundant. i
can see doing it back in the day when there weren't any other options,
but today there are, and there have been for 20+ years.


> I used to work in a place with a 1.2 MW DC
> motor whose field coil voltage was controlled by a couple rooms full
> of cast-iron resistors. The resistance array lasted nearly 80 years
> before the whole facility was taken down.
>
> As long as you keep within the safe operating area of the semiconductors,
> you're fine. If you exceed them, bad things happen.

right. but again, we're dealing with mba's here. as an engineer,
you're going to design with reliability and a safety margin built in.
as an mba, you're going to cut and keep cutting until it meets "business
objectives".


> But we don't know
> if the semiconductors are failing on these things, or if it's just
> ordinary RoHS solder failures; the RoHS crap doesn't like thermal cycling
> so well.

very true. but at the end of the day, that's still heat. and a linear
semiconductor controller is just stooopid when a wire coil will do the
same job more reliably and at a fraction of the cost.


--
fact check required

On 03/21/2013 08:10 PM, Scott Dorsey wrote:
> Bimmer Owner > wrote:
>> On Thu, 21 Mar 2013 19:15:16 -0700, jim beam wrote:
>>
>>> that's why pwm is the better solution
>>
>> Does pulse width modulation cause radio EMI?
>
> It can if it's not properly filtered. And the reason why I think beam is
> correct about this being a linear control is that there is no filtration
> in there.

it's the heat sink that's the dead give-away. a pwm controller heatsink
would be 1/10th the capacity. or less.


>
> My guess is that the cost of proper filtering and shielding makes the pwm
> controller cost more than the linear controller in this case, which is
> probably why they went the linear route.

i think the cost of that honking great heatsink significantly exceeds
the cost of a couple of extra inductors and caps.


--
fact check required

On Thu, 21 Mar 2013 20:01:20 -0400, clare wrote:

> Just put your ammeter into the heater blower fuse connector and you
> get the current of the blower motor.

That's an interesting idea.
http://www3.picturepush.com/photo/a/...g/12475041.jpg

The fuse for the blower motor is called the "infamous F76" for a reason.
http://bimmerfest.com/forums/showthread.php?t=674612

It's a 40 amp fuse under the glovebox but it's in a really inaccessible
spot; however, it's right side up, so, the wires going INTO it are
visible from the tips of your feet under the glove box.

So that's a possibility; but you'd have to cut the wires.
http://www5.picturepush.com/photo/a/...g/12475043.jpg

On Thu, 21 Mar 2013 22:54:59 +0000 (UTC), Bimmer Owner
> wrote:

>On Thu, 21 Mar 2013 15:03:45 -0700, Ashton Crusher wrote:
>
>> If it's being blown by intermittent high current draws the
>> fuse could protect the $100 FSU.
>
>That's an interesting idea. The FSU supposedly consumes the
>most power when the blower motor is set to the LOW settings
>(simply because it has to dissipate the power as heat), so,
>we could prevent excess current by fusing... say with a 10A
>fuse, the blower motor (which is said to consume 5 to 6 A).

Hmm, it did look like it had quite a heat sink. I had assumed it used
PWM to change speed, which should not generate much heat but my
assumption might be wrong (or my understanding of PWM...) A solid
state design that gets hot on purpose seems like a poor design to me.

On Mar 21, 10:25*pm, (Scott Dorsey) wrote:
> Bimmer Owner > wrote:
>
> >On Thu, 21 Mar 2013 15:16:46 -0400, Scott Dorsey wrote:
> >> has anyone got docs on that mystery IC there? *It's from Elmos
> >> Semiconductor, but it's not a standard Elmos part number on it
>
> >Focusing just on that Elmos Semiconductor AG IC from this thread:
> >http://www.bimmerfest.com/forums/sho...d.php?t=309399
> >It looks like the PN is ELMOS, 10901D, 667A 1302A
>
> >It might be a generic or a special chip; I can't find it on the web:
> >http://www.elmos.com/produkte/automo...motor-ics.html
>
> All the standard Elmos part numbers begin with an E.
>
> My guess is that the number on the chip is 10901D and that the other
> two numbers are date and batch codes.
> --scott
> --
> "C'est un Nagra. C'est suisse, et tres, tres precis."

If you google for 10901D it comes back with hits to
Chinese chip brokers that show it as an Elmos 16
pin surface mount chip. Which is consistent with what's
in the picture of the failed module, it has 16 pins.
But I could not find any data sheet on the part either.

On Fri, 22 Mar 2013 05:20:37 -0700, wrote:

> If you google for 10901D it comes back with hits to Chinese chip brokers
> that show it as an Elmos 16 pin surface mount chip.

I found the same. The chip is listed he
http://www.jotrin.com/product/parts/10901D

But there is no datasheet there.