I have tried to repair the FSU for BMW e46 by replacing the two MOSFET transistors with equivalent: IRF3205 (and even two more powerful) but it seems they cannot do the job due to other electronic component which i cannot figure out. Even though there is a change on G -gate the transistor won't change the flow ..there must be something on the Base or Collector ... I'm stunned cause i tried repairing two of them

On Thursday, November 14, 2013 7:37:33 AM UTC-5, wrote:
I have tried to repair the FSU for BMW e46 by replacing the two MOSFET transistors with equivalent: IRF3205 (and even two more powerful) but it seems they cannot do the job due to other electronic component which i cannot figure out. Even though there is a change on G -gate the transistor won't change the flow ..there must be something on the Base or Collector ... I'm stunned cause i tried repairing two of them

Root cause is BMW designed/specd a piece of crap and
doesn't give a damn about resolving it, even thouh it's
been an obvious problem for a decade, guys like you
have bitched about it and their service people obviously
know about it.

They've achieved some other marvels of engineering, like
putting electronics inside the aux fan that sits in front
of the hot radiator, to vary the fan speed. Simple on/off
fan wasn't good enough. So, they put electronics in an
environment where it gets 130F, salt, water, God knows what.
Of course they fail. And you can't even diagnose it because
no one knows what kind of signal they send it to turn it on,
off, fast, slow, etc. BMW can, with their magic computer that
they can hook up to turn it on and off. So, every 2 years
you buy a new one for $275 aftermarket price, plus labor.
Go to BMW and get the OEM part, it's $750 by the time you're
done and they still fail.

Another favorite is how they manage to use a vast assortment
of cable connectors that even upon careful examination, you
can't figure out how they come apart. And that's assuming you
can see it well, right in front of you.

I've had basic American cars that went 100K+ miles with
no CV boot joint failures. A friend just got rid of a
Honda CRV with 200K+ miles and no CV boot failures.
Friend has an X5 and it went through 2 of them in less than 75K
miles. At 140K, they're shot again. The rubber boot on
the intake manifold also failed at 75K. Obviously their rubber
boot products are crap, but they don't care.

Or how about their window system in the X5, where they use
crappy cables to hold the windows up? I've seen those fail
with the car sitting in the driveway. Cable snaps, window
falls down and smashes to bits. Not on just one window, I
saw it happen on two different ones in a friends car.

But, you're already planning on buying your next $75K BMW wonder
car, right?

On 11/14/2013 11:48 AM, wrote:
On Thursday, November 14, 2013 7:37:33

AM UTC-5, wrote:
I have tried to repair the FSU for BMW e46
by replacing the two MOSFET transistors with
equivalent: IRF3205 (and even two more powerful)
but it seems they cannot do the job due to other
electronic component which i cannot figure out.
Even though there is a change on G -gate the
transistor won't change the flow ..there must
be something on the Base or Collector ... I'm
stunned cause i tried repairing two of them


But, you're already planning on buying your next $75K BMW wonder
car, right?


Do you get many on topic comments on a home
repair list?

--
..
Christopher A. Young
Learn about Jesus
www.lds.org
..

wrote:
On Thursday, November 14, 2013 7:37:33 AM UTC-5, wrote:
I have tried to repair the FSU for BMW e46 by replacing the two MOSFET transistors with equivalent: IRF3205 (and even two more powerful) but it seems they cannot do the job due to other electronic component which i cannot figure out. Even though there is a change on G -gate the transistor won't change the flow ..there must be something on the Base or Collector ... I'm stunned cause i tried repairing two of them



Hi,
Checked/set proper bias voltage for the replacement MOSFET?
If you don't know how or can't do it they won't work properly for sure.


Root cause is BMW designed/specd a piece of crap and
doesn't give a damn about resolving it, even thouh it's
been an obvious problem for a decade, guys like you
have bitched about it and their service people obviously
know about it.

They've achieved some other marvels of engineering, like
putting electronics inside the aux fan that sits in front
of the hot radiator, to vary the fan speed. Simple on/off
fan wasn't good enough. So, they put electronics in an
environment where it gets 130F, salt, water, God knows what.
Of course they fail. And you can't even diagnose it because
no one knows what kind of signal they send it to turn it on,
off, fast, slow, etc. BMW can, with their magic computer that
they can hook up to turn it on and off. So, every 2 years
you buy a new one for $275 aftermarket price, plus labor.
Go to BMW and get the OEM part, it's $750 by the time you're
done and they still fail.

Another favorite is how they manage to use a vast assortment
of cable connectors that even upon careful examination, you
can't figure out how they come apart. And that's assuming you
can see it well, right in front of you.

I've had basic American cars that went 100K+ miles with
no CV boot joint failures. A friend just got rid of a
Honda CRV with 200K+ miles and no CV boot failures.
Friend has an X5 and it went through 2 of them in less than 75K
miles. At 140K, they're shot again. The rubber boot on
the intake manifold also failed at 75K. Obviously their rubber
boot products are crap, but they don't care.

Or how about their window system in the X5, where they use
crappy cables to hold the windows up? I've seen those fail
with the car sitting in the driveway. Cable snaps, window
falls down and smashes to bits. Not on just one window, I
saw it happen on two different ones in a friends car.

But, you're already planning on buying your next $75K BMW wonder
car, right?

Hi i just bought an fsu from behr for my 2002 e46 and i opened it . Inside was only one transistor 2n0605 from infineon and an elmos 10901d.The 2n0605 is 80 amps 55v mos.I am now working to figure out the exact schematics

In the past i replaced the transistor with some irf mosfet .they were powerfull but didnt work properly .now i know you have to put low voltage and hi amps mos.

On Tuesday, February 4, 2014 11:11:36 AM UTC-5, wrote:
Hi i just bought an fsu from behr for my 2002 e46 and i opened it . Inside was only one transistor 2n0605 from infineon and an elmos 10901d.The 2n0605 is 80 amps 55v mos.I am now working to figure out the exact schematics

Out of curiousity, how did you open it up? The ones I've seen
are potted in 1/2" deep epoxy, with no obvious way to get that out
without destroying what's inside....

Also, Ive been helping a friend with the FSU problem. His blower
went from working fine to not working at all. Being concerned about
all the apparently crap ones out there, bought one from a BMW dealer
that is marked Sitronic. It appeared to work, but..... With the
blower on high it will run all day, no problem. You can back the blower
off to about 80% and as you do so, the heat sink starts to get very
hot, but the blower still runs at 80%. If you back the speed down
anymore, it gets even hotter and then starts pulsating the blower,
finally stops all together. I tried dipping the end of the heatsink
rods into a container of water when it does that and within a couple
seconds the blower starts working again fine. So, it's definitely
heat related.

Any ideas on what's wrong? First impression would be that it's the new
FSU because the blower works on full or near full speed. But I guess
the other possibility is that the motor has a problem? Seems odd
though that a single winding type DC motor could run fine at full
speed, but then cause the FSU to overheat at lower speeds.

wrote in message
...

Also, Ive been helping a friend with the FSU problem. His blower
went from working fine to not working at all. Being concerned about
all the apparently crap ones out there, bought one from a BMW dealer
that is marked Sitronic. It appeared to work, but..... With the
blower on high it will run all day, no problem. You can back the blower
off to about 80% and as you do so, the heat sink starts to get very
hot, but the blower still runs at 80%. If you back the speed down
anymore, it gets even hotter and then starts pulsating the blower,
finally stops all together. I tried dipping the end of the heatsink
rods into a container of water when it does that and within a couple
seconds the blower starts working again fine. So, it's definitely
heat related.

Any ideas on what's wrong? First impression would be that it's the new
FSU because the blower works on full or near full speed. But I guess
the other possibility is that the motor has a problem? Seems odd
though that a single winding type DC motor could run fine at full
speed, but then cause the FSU to overheat at lower speeds.

Could be the motor. While not exectally correct, the motor requires a
cretain ammount of power. When running at full speed the transistor is
acting more like a swithc. At slower speeds it is acting more like a
resistor. Any power that is not used to get the motor to speed is converted
to heat in the FSU. The slower the motor is ran, the hotter the FSU will
get. If the bearings are going bad or there is some other reason the motor
is drawing more than the rated current the FSU will heat up more than
normal, especially at slower speeds.

Have you checked the current the motor is drawing and made sure it is
getting the voltage it is suspose to at full speed ?



---
This email is free from viruses and malware because avast! Antivirus protection is active.
http://www.avast.com

wrote:
In the past i replaced the transistor with some irf mosfet .they were powerfull but didnt work properly .now i know you have to put low voltage and hi amps mos.

Hi,
If bias voltage does not match close enough heavier one won't function
properly. Is there a way to improve cooling by heat sinking or applying
proper thermal paste?

The fsu from behr doesnt have that resin and it was easy to open.the new module has the transistor on a copper plate to make the heat transfer fast but the contact with the aluminium its not so good.I did lost the waranty when i did this,but i'm not giving up in building an fsu from scratch.I have some pictures of the fsu.

miercuri, 5 februarie 2014, 01:43:49 UTC+2, a scris:
The fsu from behr doesnt have that resin and it was easy to open.the new module has the transistor on a copper plate to make the heat transfer fast but the contact with the aluminium its not so good.I did lost the waranty when i did this,but i'm not giving up in building an fsu from scratch.I have some pictures of the fsu.

https://www.dropbox.com/sc/lfsinwz9sh48jid/W--GuYxNWN
with this new fsu i can see the circuit bord and i can make an ideea about elmos 10901d and find a replacement

https://www.dropbox.com/sc/a9bmsmsoagh1bzt/6KAIZ_tgiJ

And this is the heatsink i want to use and maybe add a little fan which will run when hvac comes on but i need to find a +12v from hvac control
https://www.dropbox.com/sc/h14pr26wq0xgd63/-Zrd9vszho

On Tuesday, February 4, 2014 12:02:11 PM UTC-5, Ralph Mowery wrote:
wrote in message

...



Also, Ive been helping a friend with the FSU problem. His blower

went from working fine to not working at all. Being concerned about

all the apparently crap ones out there, bought one from a BMW dealer

that is marked Sitronic. It appeared to work, but..... With the

blower on high it will run all day, no problem. You can back the blower

off to about 80% and as you do so, the heat sink starts to get very

hot, but the blower still runs at 80%. If you back the speed down

anymore, it gets even hotter and then starts pulsating the blower,

finally stops all together. I tried dipping the end of the heatsink

rods into a container of water when it does that and within a couple

seconds the blower starts working again fine. So, it's definitely

heat related.



Any ideas on what's wrong? First impression would be that it's the new

FSU because the blower works on full or near full speed. But I guess

the other possibility is that the motor has a problem? Seems odd

though that a single winding type DC motor could run fine at full

speed, but then cause the FSU to overheat at lower speeds.



Could be the motor. While not exectally correct, the motor requires a

cretain ammount of power. When running at full speed the transistor is

acting more like a swithc. At slower speeds it is acting more like a

resistor. Any power that is not used to get the motor to speed is converted

to heat in the FSU. The slower the motor is ran, the hotter the FSU will

get.

Yes, I hear what you're saying. I assume it's being used like a switch,
all the time being switched on and off in a PWM fashion, like a
light dimmer. But while going from off to on, in the middle, it
will generate more heat. That's why electronic light dimmers get warm too.
And I guess if something is wrong with the motor, like your bearing
example, it could generate more heat in the FSU. I guess a way to look at
it is like what would happen if you put a 1500W load on a triac based
AC light dimmer rated for 700W. At full power, it would probably
be fine. At half power, it could get so hot it fails.





If the bearings are going bad or there is some other reason the motor

is drawing more than the rated current the FSU will heat up more than

normal, especially at slower speeds.



Have you checked the current the motor is drawing and made sure it is

getting the voltage it is suspose to at full speed ?


Not so far because there isn't an easy way to do so.
I don't have a clamp on DC ammeter and with the connector
they use, no way to get something into/onto the circuit. It
hasn't blown the blower fuse. I guess I could stick the
wire with a pointed probe, but even that isn't easy because
there isn't much access to the cable.

There is one more avenue being pursued. This car uses a cabin
air filter and it's fairly dirty and needs to be changed.
Thinking that might have some effect on it, but on the other
hand AFAIK, it only filters the outside incoming air, which is
like 20% max and the dash controls allow that to be shut off,
or automatically controlled,
etc. So I would think the blower should run fine with just the
cabin air, especially at lower speeds, but who knows. So one
of those is on the way.

After that, I think next is going to be to get another
FSU, from another manufacturer and try it. The BMW one was
$125, there are Behr ones available for $45. As you can see
from the thread, these things are poorly designed and have
a high failure rate. They've been around since 2000 and
they still all suck and fail. Folks can't even agree if
any of the various manufacturers are better. They have
changed the design. The original ones had less heat sink
pins, but they were larger. New ones have at least 2X the
pins, like 40 of them, so clearly they've changed the
design because of heat issues. Because they all suck is why
the guys here have been trying to reverse engineer it and
see if they can do something to make it work. I think just
a larger heat sink would probably solve it, but no easy way to do
that I can see.

Still you think a new one from BMW would work out of
the box, but who knows. The other sad news is that to get to
the blower requires removal of the whole dashboard

If you want to see a pic of what one looks like, just
google BMW FSU.

On Tuesday, February 4, 2014 7:27:52 PM UTC-5, wrote:
And this is the heatsink i want to use and maybe add a little fan which will run when hvac comes on but i need to find a +12v from hvac control

https://www.dropbox.com/sc/h14pr26wq0xgd63/-Zrd9vszho

Good pictures.

It looks like the chip, the transistor and a bunch of
caps, resistors, diodes. Anyone know what that funny
looking copper strip is with glue or whatever under
it is?


I would think that adding a little fan to the heatsink
might be the most productive approach. If you can find
a power source that's on when the HVAC on that would be
great. But you could also power it from anything that has
12V when the ignition is ON, which could be a lot easier
to find. Any chance of tapping into any of the blower
airflow to cool it somehow? And is there even room in
there for the additinal fan? Where the part goes under
the dash, I can't tell what room there is behind it, but
I would doubt there is enough for a fan.

One difference I see with your new part and Sitronic one
from BMW is that your Behr one uses an aluminum housing
all around the part, while the BMW part uses plastic.
I noticed that difference in some of the pics before,
but couldn't tell which ones were all aluminum and
which just had aluminum for the heatsink. The all
aluminum should be better because it can help conduct
more of the heat away.

Another perhaps minor point. IDK how the thing is fastened
into your car, but on the X5, it kind of snaps in, but
from the removal instructions, there are supposed to be
two small bolts that hold it down. The one here had
been replaced previously and had no bolts. I can see
why, given where it goes it would be hard to get to the
bolts and the clip seemed to hold it well enough.
But.... If it's bolted to metal and it's all aluminum
like yours, that would help take some more of the heat
away. But then it looks to me like most of the ones
out there have plastic frames anyway, so in that case
I don't think bolting it down is going to make much
difference in transfering heat, as plastic doesn't
conduct that well.

I also don't know how productive figuring out what's
hooked up to what on that board is going to be. No
one has been able to find a datasheet on that Elmos chip.
No one is even sure what it really is.
If the intention is to build a new design, my approach
would be this. There are only 5 wires to this FSU.

12V
GND
Signal ( very small wire)
Motor +
Motor -

The missing piece here is what's on that control
signal wire that sets the speed? If I was going to
make a new design, I'd put a scope on it and find out
if it's an analog signal, digital, etc. If it's just
analog and varies with the set speed, then you can
design a new circuit. Of course getting it into the
right form factor, etc is going to take a lot of work.
IMO, it's not worth it.

Oh, and regarding that single wire control, they use
a similar single wire to control the variable speed
auxilliary fan for the radiator. That fan has electronics
in it too that can vary the aux fan speed, based on a
control input. Simnple on/off like a billion other cars
have was too simple. And it's in the worst place possible
right next to the radiator.

The reason I bring that up, that fan fails frequently
and can drain the battery. The car here, the fan stopped
working, but the AC still works fine except when stopped
in traffic. I was worried about it draining the battery,
which they are known to do when they fail. So, I unplugged
it. It's just 3 wires, 12V, gnd, control.
And..... the AC no longer worked at all. The only
way I could explain what was happening was that somehow
the HVAC control knew the fan was disconnected, leaving
me to wonder if that single wire control to it was
bi-directional. IDK why anyone would design it that
way, not 100% sure it's done that way, but thought I'd
mention it in the sense that the single wire into the
FSU might not be so simple.

Hi.I managed to fit the fsu right behind the glove compartiment.Added a little fan ,but the hvac doesn't have a +12 so i made a circuit with an 10k ntc that allows the fan to start only when it gets hot
The copper strip is just a bypass for the +12 of the blower .
The radiator that i used is all copper and a had to isolate it because when it touches the ground(-) the blower comes on at full speed.
I made the test on low speed half hour and it stays warm.
https://www.dropbox.com/sc/bkpt1joem68b4u6/kv5qGP_CAL

The fsu gets really hot at low speeds so the air flow its not enough to cool it down.At hi speed its cold .So that why i decide to add the fan .A cpu cooler will do just fine.
I did drill a hole were the transistor has that hole to be placed on a heatsink and added a screw so that the heat transfer is better in that area.

On Wednesday, February 5, 2014 11:47:31 PM UTC-5, wrote:
The fsu gets really hot at low speeds so the air flow its not enough to cool it down.At hi speed its cold .So that why i decide to add the fan .A cpu cooler will do just fine.

I did drill a hole were the transistor has that hole to be placed on a heatsink and added a screw so that the heat transfer is better in that area.

Saw your pics and sounds like you have a good design there that
should work. The idea of just making the fan go on when it gets
hot was a good one, so you could then just tie it to the 12V that
is already there. If you tied the fan to the 12V for the blower,
you might want to put an inline fuse in. That blower circuit is
30A? I've never seen a CPU fan like that cause a problem, but
just a thought. The fan may have some kind of protection of it's
own though, so manybe it's not an issue.

Good idea to screw the transistor down. I noticed that in the
photo, it wasn't clear exactly how they had it secured down.
And it's also near one edge of the heatsink, instead of being in
the middle. Before you modified it, was there thermal paste
between the back of the board and the heatsink?

This is just a really bad design. If they had correctly bonded
the heatsink directly to that transistor where all the heat is
being generated, I bet it would work fine.

The original fsu board was tied down only on the edges by that black plastic which was clipped by some small fins of the aluminium radiator .It did have a thermal paste i also used a paste .
The +12 i took from a main +12 and used a 2.5 amps fuse .
The transistor i think disipates about 50w to 80w of heat at low speeds that makes the heatsink to small to dissipate that heat and the airflow of the blower at that speed is too little.Its true the transistor is made for 250 degrees but lets be true they dont make like they used.

On Thursday, February 6, 2014 11:04:17 AM UTC-5, wrote:
The original fsu board was tied down only on the edges by that black plastic which was clipped by some small fins of the aluminium radiator .It did have a thermal paste i also used a paste .

The +12 i took from a main +12 and used a 2.5 amps fuse .

The transistor i think disipates about 50w to 80w of heat at low speeds that makes the heatsink to small to dissipate that heat and the airflow of the blower at that speed is too little.Its true the transistor is made for 250 degrees but lets be true they dont make like they used.

I think it's probably generating more heat than that. The transistor
is rated at 300W. In any case, I don't think it's the power transistor's
power handling capability that is
the problem. It's that the heat causes other parts to fail. I'm sure
you've seen pics of ones where the solder joints came undone. IDK
what causes the one here to basically shutdown once it gets too hot,
but suspect it's something to do with the Elmos chip, either by
intentional design or that it just shuts dowm by malfunctioning from
the heat. Whatever the max operating temp is for that Elmos ASIC,
I'll bet it's being exceeded when these things fail.

The overall design is probably just enough if everthing works like
it should. If the blower winds
up pulling some more current than normal, for whatever reason,
then it pushes it over the edge. I'm thinking as Ralph suggested,
it's probably the blower that's the root cause here. And to fix
that requires pulling the dashboard in the X5. Another great
design. Not something I wanted to do 30 years ago, but today
with all the air bags and God knows what else crammed in there,
it's got to be a lot worse. Wish I could get some oil to those
motor bearings and see what happens.

Going back to what Ralph said about bearings, I'm not convinced
that the blower is putting out full RPM even when it's connected
to 12V directly. There ia a lot of air coming out, but I think
it may have been even more previously, hard to tell for sure.
But if it is putting out less at full, then that could be a sign
of the bad motor bearing thing. Another sign I think is that
it seems even when you first start it up, with a cold FSU, it
seems to me that the airflow at say mid speed is more like
airflow at 1/3 speed, etc.

Good luck with your project at let us know how it works after
you've had it running for awhile.

wrote in message
...
The original fsu board was tied down only on the edges by that black
plastic which was clipped by some small fins of the aluminium radiator .It
did have a thermal paste .i also used a paste .
The +12 i took from a main +12 and used a 2.5 amps fuse .
The transistor i think disipates about 50w to 80w of heat at low speeds
that makes the heatsink to small to dissipate that heat and the airflow of
the blower at that speed is too little.Its true the transistor is made for
250 degrees but lets be true they dont make like they used.

The heatsink of the unit needs to be where air can circulate past it. The
fan is a good idea. I am not sure how the transistor is fastened to the
heat sink. It needs to make good mechanical contact with it.

When you use the thermal paste just use the smallest ammount you can. The
paste is not really that conductive of heat. It is just beter than air.
You want the transistor to make good mechanical contact with the metal of
the heat sink. There are very small inperfections in the metal that has air
voids in it. The object is to just fill the air voids. Too much paste is
worse than none at all.





---
This email is free from viruses and malware because avast! Antivirus protection is active.
http://www.avast.com

I know that it needs a space with more air thats why im thinking of rellocating the fsu in the engine compartiment,but im waiting to get warmer outside.
The transistor is now fastened very good not like the original were it made contact only on the margins of the copper plate

On Thursday, February 6, 2014 2:30:05 PM UTC-5, Ralph Mowery wrote:
wrote in message

...

The original fsu board was tied down only on the edges by that black

plastic which was clipped by some small fins of the aluminium radiator .It

did have a thermal paste .i also used a paste .

The +12 i took from a main +12 and used a 2.5 amps fuse .

The transistor i think disipates about 50w to 80w of heat at low speeds

that makes the heatsink to small to dissipate that heat and the airflow of

the blower at that speed is too little.Its true the transistor is made for

250 degrees but lets be true they dont make like they used.



The heatsink of the unit needs to be where air can circulate past it. The

fan is a good idea. I am not sure how the transistor is fastened to the

heat sink. It needs to make good mechanical contact with it.


The power transistor is not fastend to the heat sink at all.
If you look at the pics, it's on the top side of the PC board.
The bottom of the PC board has copper over all of it. Then the
bottom of the PC board contacts the heatsink. So, you have the
PC board material in between, which is not a good heat conductor.
And the transistor is closer to one side of the heat sink and no
bolt through the usual bolt location provided.

And to top it off, these things have been failing for 10+ years
and not one of the many suppliers has seen fit to make some of
the simple changes that would fix it.

Its directly on the copper plate.The pcb has a rectangular hole .its tinned

Unlike the older models when the transistor heated up it transmitted the heat to its pins ,now the heat transmits to the whole pcb thanks to genius who decided to glue the pcb to the copper

On Thursday, February 6, 2014 4:29:23 PM UTC-5, wrote:
Unlike the older models when the transistor heated up it transmitted the heat to its pins ,now the heat transmits to the whole pcb thanks to genius who decided to glue the pcb to the copper

I just discovered something important that I haven't seen mentioned
here or in any other threads on the BMW blower motor resistor (aka final
stage unit) problems. This module is mounted in a location where the
heat sink fins are directly in the blower motor airstream. It relies on
the blower air to keep it cool. From where it's located and how you
remove/install it, it's not obvious that the heat sink fins wind up in
the blower air, because you can't see where they actually are.

When I replaced it, to make sure it worked before putting it together,
I just left it hanging for testing. It overheats as soon as you slow
the blower to about 80% of full speed and stops working. Once I stumbled
on the fact that if it's mounted it gets air flow to cool it and
I mounted it, it's apparently working. Just thought I'd pass that along,
as it seems like something that could trip up a lot of people.