Fan Controller and Fan causes floppy drive failure

My thesis, which will sound nutty, is the following: The use of my Vantec
Fan Controller to turn down the speed of my two intake 80mm case fans causes
my floppy drive to fail. I will defend this below but now for the setup,
pathology, and history.

Setup:
I recently installed a Vantec Silver Nexus NXP-205 Rheobus (with 4 fan
controllers) in my Kingwin K11 case. This case comes with 4 80mm fans two
of which are intake fans that sit at the bottom front of the machine. Not
much above these fans is the 3.5 inch drive external bays. The bottom bay
contains my Vantec fan controller and the one immediately above that has my
floppy drive. These two intake fans, that are close to the floppy drive,
are attached to the first two of the 4 fan speed controllers. The other two
controllers are attached to the other two distant case fans but apparently
have no effect adverse on my floppy drive (regardless of how I set their
speed controllers).

History/pathology:
The two intake fans are the loudest. I experimented with turning their
speed down by setting their fan controllers to the lowest setting and
noticing that the system temperature was not adversely effected (even though
it was clear from the noise attenuation that the fans turned much much
slower). I don't use my floppy drive much but eventually, over a few weeks,
when I did I noticed it behaved erratically. I did not think for a moment
that these two things were related. Specifically, some boot floppies (3
different ones) wouldn't boot almost all of the time. I thought the
floppies were bad but they booted in other machines.

I decided to quickly bypass this problem by making a boot CD from the
floppy, one that same computer because it had the burner, using Nero CD
burner, but when I tried that it couldn't read the files from the floppy or
of that floppy's copy. I tried this multiple times.

To me, from my experience, this indicated that the floppy drive was probably
bad (versus the motherboard or the cable).

However, when I swapped the floppy with that from another computer the
supposedly bad floppy drive worked great.

So, I figured it must be the floppy cable but when I tried the original
floppy cable with the original floppy drive on the second computer it still
worked great!

Maybe a bad MB, hard to believe, so now I was confused so I reattached that
cable and floppy to my original machine but I decided, really out of
laziness, to leave the floppy drive outside the case on a mat. This time
everything worked. I booted it to the floppy no less than 10 times with no
problem (taking this floppy out and putting it back in - rebooting by power
off, reset and CTRL-ALT-DEL).

I booted up into Win2k and using Nero I was now able to make my boot CD from
this floppy. I did this multiple times.

I then repeated this whole process a few times removing the floppy from the
drive bay (but still having it attached) and then putting it back in. I
even removed it from the case out the front still leaving the cables
attached so I wouldn't have to remove and reattach the floppy cables as that
could theoretically be another source of error. Every time when the floppy
drive was in the case it wouldn't work but when it was outside the case it
would.

Couldn't believe it!

Finally, with the floppy drive installed in the drive bay, I turned up the
fan controllers of the two intake fans all the way and NOW I was able to use
Nero to successfully create the boot CD from the boot floppy.

I did this test multiple times. When I turned the two (front) fan
controllers all the way up the floppy always worked but when I turned it all
the way down it always failed. It seems to fail even if the fan controllers
half way but does work if its about 2/3 way.

Then I tried turning the other two fan controllers all the way down with the
intake fan controllers all the way up. This worked fine. But when I turned
one of the intake fan controllers all the way down the floppy failed. Here's
a table:
The first four columns are the settings of fan controller 1 through 4
respectively (values: B - bottom power, T - Top Power, H - Half Power, 2 -
2/3 Power). The fifth column is whether the floppy drive failed (values:
F - failed, W - worked)

B B T T F
T T B B W
B B H H F
2 2 H H W
B T H H W
T B H H W
T B B B F

Conclusion:
So clearly, there is a problem related to the Vantec Fan controller
controlling these fans but after that I don't know what to say. The tests
"B B T T F" and "T T B B W" indicate that the problem is context dependant
on which fans are being controlled. Specifically, turning the first two
fans, the ones close to the floppy drive, down will create a floppy failed
while turning down the other two fans may not. However "T B H H W" and "T B
B B F" indicate that the problem is NOT related to those two close intake
fans because they were set the same in both tests but setting the far two
fans low caused a failure.

Ummm.

Andrew Diamond wrote:
My thesis, which will sound nutty, is the following: The use of my Vantec
Fan Controller to turn down the speed of my two intake 80mm case fans causes
my floppy drive to fail. I will defend this below but now for the setup,
pathology, and history.
snip


I did this test multiple times. When I turned the two (front) fan
controllers all the way up the floppy always worked but when I turned it all
the way down it always failed. It seems to fail even if the fan controllers
half way but does work if its about 2/3 way.

Then I tried turning the other two fan controllers all the way down with the
intake fan controllers all the way up. This worked fine. But when I turned
one of the intake fan controllers all the way down the floppy failed. Here's
a table:
The first four columns are the settings of fan controller 1 through 4
respectively (values: B - bottom power, T - Top Power, H - Half Power, 2 -
2/3 Power). The fifth column is whether the floppy drive failed (values:
F - failed, W - worked)

B B T T F
T T B B W
B B H H F
2 2 H H W
B T H H W
T B H H W
T B B B F

Conclusion:
So clearly, there is a problem related to the Vantec Fan controller
controlling these fans but after that I don't know what to say. The tests
"B B T T F" and "T T B B W" indicate that the problem is context dependant
on which fans are being controlled. Specifically, turning the first two
fans, the ones close to the floppy drive, down will create a floppy failed
while turning down the other two fans may not. However "T B H H W" and "T B
B B F" indicate that the problem is NOT related to those two close intake
fans because they were set the same in both tests but setting the far two
fans low caused a failure.

Ummm.

I don't know what kind of circuit the Vantec controller is using but it sounds a
lot like EMI interference, either radiated or conducted, like from a PWM.

Conducted EMI would be through the power lines so try putting the Vantec on it's
own power connector, it only on that chain, and nearest the PSU. Put the floppy
on a completely different power chain.

Reason: Current pulses, like from a PWM, cause transient voltage drops across
power wires, especially long ones. If another device taps it's power from
between the pulsing device and the power supply then it's power rails are
bouncing around and that can get injected into it's circuitry as noise, causing
erratic operation. Of course, for a floppy, 'erratic' means bad data and bad
data means it isn't going to boot.

If it's radiated you could try shielding the floppy drive but the case mount
should be doing that already.

Thanks for the response,

As it so happens, by accident, the two devices, floppy drive and fan
controller, are on different power strands. In fact, there is nothing else
on either of the power strands.
The fan controller is connected pretty much at the end of its chain but then
given that nothing else is on its chain I'm not sure that makes a
difference.

In any event, the fact that the floppy drive failure didn't happen when the
everything was connected the same way but the floppy was outside of the case
would seem to rule out power noise in the cable (right?).

As far as the floppy shielding goes, I tried two different floppy drives.
One has the controller right on the underside, right above the electronics
of the fan controller which are completely unshielded. The other floppy
drive is completely encased in metal. Both failed the same way.

"David Maynard" wrote in message
...
Andrew Diamond wrote:
My thesis, which will sound nutty, is the following: The use of my
Vantec
Fan Controller to turn down the speed of my two intake 80mm case fans
causes
my floppy drive to fail. I will defend this below but now for the
setup,
pathology, and history.
snip


I did this test multiple times. When I turned the two (front) fan
controllers all the way up the floppy always worked but when I turned it
all
the way down it always failed. It seems to fail even if the fan
controllers
half way but does work if its about 2/3 way.

Then I tried turning the other two fan controllers all the way down with
the
intake fan controllers all the way up. This worked fine. But when I
turned
one of the intake fan controllers all the way down the floppy failed.
Here's
a table:
The first four columns are the settings of fan controller 1 through 4
respectively (values: B - bottom power, T - Top Power, H - Half Power,
2 -
2/3 Power). The fifth column is whether the floppy drive failed
(values:
F - failed, W - worked)

B B T T F
T T B B W
B B H H F
2 2 H H W
B T H H W
T B H H W
T B B B F

Conclusion:
So clearly, there is a problem related to the Vantec Fan controller
controlling these fans but after that I don't know what to say. The
tests
"B B T T F" and "T T B B W" indicate that the problem is context
dependant
on which fans are being controlled. Specifically, turning the first two
fans, the ones close to the floppy drive, down will create a floppy
failed
while turning down the other two fans may not. However "T B H H W" and
"T B
B B F" indicate that the problem is NOT related to those two close
intake
fans because they were set the same in both tests but setting the far
two
fans low caused a failure.

Ummm.

I don't know what kind of circuit the Vantec controller is using but it
sounds a
lot like EMI interference, either radiated or conducted, like from a PWM.

Conducted EMI would be through the power lines so try putting the Vantec
on it's
own power connector, it only on that chain, and nearest the PSU. Put the
floppy
on a completely different power chain.

Reason: Current pulses, like from a PWM, cause transient voltage drops
across
power wires, especially long ones. If another device taps it's power from
between the pulsing device and the power supply then it's power rails are
bouncing around and that can get injected into it's circuitry as noise,
causing
erratic operation. Of course, for a floppy, 'erratic' means bad data and
bad
data means it isn't going to boot.

If it's radiated you could try shielding the floppy drive but the case
mount
should be doing that already.

Andrew Diamond wrote:
Thanks for the response,

As it so happens, by accident, the two devices, floppy drive and fan
controller, are on different power strands. In fact, there is nothing else
on either of the power strands.

That should be the safest arrangement.

The fan controller is connected pretty much at the end of its chain but then
given that nothing else is on its chain I'm not sure that makes a
difference.

The possible difference would be radiated noise from the long power wires. Same
with the fan cables except the power lines have all of them combined.

In any event, the fact that the floppy drive failure didn't happen when the
everything was connected the same way but the floppy was outside of the case
would seem to rule out power noise in the cable (right?).

It would seem to, unless there's some kind of ground loop created when the
floppy case is connected to the system case.


As far as the floppy shielding goes, I tried two different floppy drives.
One has the controller right on the underside, right above the electronics
of the fan controller which are completely unshielded. The other floppy
drive is completely encased in metal. Both failed the same way.

Well, just being in metal isn't necessarily enough. The question is whether it's
grounded or not.

New test - It seems to be the EMI from the Vantec controller itself rather
then the fans, per se, or the power lines. In my latest experiment I moved
the entire Vantec controller unit outside the case without detaching any of
the cables (while the machine was still on. Cheap thrills). I turned the
two intake fan controllers all the way down to ensure failure. However, now
the floppy copy (via Nero as described previously) test worked fine. I did
this twice. Then I gently slit the unit back into the case and ran the test
again and, as expected, it failed.

Note that the Vantec unit's electronics are unshielded and some of them
come within about 1/4" of the bottom of the floppy drive.

I'm really happy I didn't put this immediately under my HD. They,
fortunately, are in removable 5.25" bays located a few inches above my
controller.
"David Maynard" wrote in message
...
Andrew Diamond wrote:
Thanks for the response,

As it so happens, by accident, the two devices, floppy drive and fan
controller, are on different power strands. In fact, there is nothing
else
on either of the power strands.

That should be the safest arrangement.

The fan controller is connected pretty much at the end of its chain but
then
given that nothing else is on its chain I'm not sure that makes a
difference.

The possible difference would be radiated noise from the long power wires.
Same
with the fan cables except the power lines have all of them combined.

In any event, the fact that the floppy drive failure didn't happen when
the
everything was connected the same way but the floppy was outside of the
case
would seem to rule out power noise in the cable (right?).

It would seem to, unless there's some kind of ground loop created when the
floppy case is connected to the system case.


As far as the floppy shielding goes, I tried two different floppy
drives.
One has the controller right on the underside, right above the
electronics
of the fan controller which are completely unshielded. The other floppy
drive is completely encased in metal. Both failed the same way.

Well, just being in metal isn't necessarily enough. The question is
whether it's
grounded or not.

Andrew Diamond wrote:
New test - It seems to be the EMI from the Vantec controller itself rather
then the fans, per se, or the power lines. In my latest experiment I moved
the entire Vantec controller unit outside the case without detaching any of
the cables (while the machine was still on. Cheap thrills). I turned the
two intake fan controllers all the way down to ensure failure. However, now
the floppy copy (via Nero as described previously) test worked fine. I did
this twice. Then I gently slit the unit back into the case and ran the test
again and, as expected, it failed.

Note that the Vantec unit's electronics are unshielded and some of them
come within about 1/4" of the bottom of the floppy drive.

I'm really happy I didn't put this immediately under my HD. They,
fortunately, are in removable 5.25" bays located a few inches above my
controller.

Hmm. Seems like they didn't pay enough attention to the design if it's pouring
out that much EMI.

"Andrew Diamond" wrote in message ...

Vantec Silver Nexus NXP-205 Rheobus (with 4 fan controllers)

the two devices, floppy drive and fan controller, are on different
power strands.

I tried two different floppy drives. One has the controller right
on the underside, right above the electronics of the fan controller
which are completely unshielded. The other floppy drive is
completely encased in metal. Both failed the same way.

Try running the fan controller form a separate power supply (be sure
it and the computer's supply are both grounded at the wall outlet or
the computer could be damaged), and if the problem disappears, then it
was caused by the controller radiating inteference through its power
lines, not through the air. And even though your floppy drive and
controller are on physically separate lines, they go to the same
places inside the power supply and are so short that there's
practically no separation.

Each fan controller channel probably uses a pulse width modulator, and
its 12V pulses are getting into the power lines because the controller
has inadequate filtering against this. A good filter would consist of
a choke (wire wrapped around a ferrous core) in series with the +12V
connection and a low-value capacitor (roughly 0.1 uF) between +12V and
ground, before and after this choke, and you can buy such a filter
already made, but to maintain the 18W/channel rating of the
controller, you'll need something rated for more than 6 amps. A car
audio power filter (Radio Shack, car stereo dealers) will work, but so
will a 120VAC line filter (the voltage doesn't have to match, as long
as it's high enough). See www.cor.com or www.delta.com.tw for
information about line filters. Since there aren't any electrocution
issues here, consider buying a filter from an electronics surplus
parts dealer for $2. You'll need to add suitable power connectors,
which electronics dealers carry, or you can simply splice into a disk
drive extension power cable (probably a Y adapter).

You should complain to Vantec for releasing such a badly designed
product to the public, but most likely they won't have anyone whose
technical qualifications are adequate to understand the problem. It's
always a poor idea to buy silly accessories because they tend to be
made by silly companies, and I can't think of anything less useless
than a fan controller where more effort was put into its cosmetics
than its function. If you want to quiet down your fans, check any of
the numerous web sites dedicated to cooling or quiet PC operation,
such as www.silentpcreview.com , and build one of their very simple
controllers. Even a single 50 ohm thermistor, wired in series with
the fan's 12V line, is usually adequate, and instead of a pulse width
modulator, use a linear regulator on a heatsink for each fan.

Thanks for the lengthy and detailed response.
As I posted last night, the fact that the floppy drive works when the
controller is slid, while still attached, out of the case would seem to
indicate that its EMI generated by the fan controller through the air
(probably).

As far as "always a poor idea to buy silly accessories". Ahmen.
I just wasn't expecting the Spanish Inquisition. If perhaps the box had
said - "may screw up your system and drive you bananas" I wouldn't have
bought it.

Seriously, this is my work computer and all I wanted to do was to make it
quieter and this seemed (hah) like an easy solution. In classic Murphy's
law style this has not proven to be the case. Playing around with mutliple
PSUs at some risk to investigate this trinket is more than I signed on for
or am willing to risk.

I will be replacing the loud front intake 80mm case fans with a couple of
21db Vantec Stealth fans and that will obviate the need for the controller
on those two fans. The other fans can remain turned up pretty high which
seems to mitigate any EMI effect sufficiently.

There is a slightly infuriating irony/injustice that I will be "solving"
this Vantech fan controller problem by purchasing 2 case fans from the same
company. I would feel better the new fans came from another company but
they're already on order. Oh well.

Anyway, thanks for the detailed response. It almost makes me feel guilty
for not trying all the suggestions.

"do_not_spam_me" wrote in message
om...
"Andrew Diamond" wrote in message
...

Vantec Silver Nexus NXP-205 Rheobus (with 4 fan controllers)

the two devices, floppy drive and fan controller, are on different
power strands.

I tried two different floppy drives. One has the controller right
on the underside, right above the electronics of the fan controller
which are completely unshielded. The other floppy drive is
completely encased in metal. Both failed the same way.

Try running the fan controller form a separate power supply (be sure
it and the computer's supply are both grounded at the wall outlet or
the computer could be damaged), and if the problem disappears, then it
was caused by the controller radiating inteference through its power
lines, not through the air. And even though your floppy drive and
controller are on physically separate lines, they go to the same
places inside the power supply and are so short that there's
practically no separation.

Each fan controller channel probably uses a pulse width modulator, and
its 12V pulses are getting into the power lines because the controller
has inadequate filtering against this. A good filter would consist of
a choke (wire wrapped around a ferrous core) in series with the +12V
connection and a low-value capacitor (roughly 0.1 uF) between +12V and
ground, before and after this choke, and you can buy such a filter
already made, but to maintain the 18W/channel rating of the
controller, you'll need something rated for more than 6 amps. A car
audio power filter (Radio Shack, car stereo dealers) will work, but so
will a 120VAC line filter (the voltage doesn't have to match, as long
as it's high enough). See www.cor.com or www.delta.com.tw for
information about line filters. Since there aren't any electrocution
issues here, consider buying a filter from an electronics surplus
parts dealer for $2. You'll need to add suitable power connectors,
which electronics dealers carry, or you can simply splice into a disk
drive extension power cable (probably a Y adapter).

You should complain to Vantec for releasing such a badly designed
product to the public, but most likely they won't have anyone whose
technical qualifications are adequate to understand the problem. It's
always a poor idea to buy silly accessories because they tend to be
made by silly companies, and I can't think of anything less useless
than a fan controller where more effort was put into its cosmetics
than its function. If you want to quiet down your fans, check any of
the numerous web sites dedicated to cooling or quiet PC operation,
such as www.silentpcreview.com , and build one of their very simple
controllers. Even a single 50 ohm thermistor, wired in series with
the fan's 12V line, is usually adequate, and instead of a pulse width
modulator, use a linear regulator on a heatsink for each fan.

Response from Vantec (originally they couldn't reproduce the problem.)
Andrew,

Correction, we let the test run for a while longer and the disk did become
corrupt. This is a very startling issue that has NEVER come up before.
Which is amazing since the NXP-205 is not new to the market. I will run
more tests tomorrow and hopefully come up with something more conclusive.
We have also contacted our engineers to see if they have a response. I also
requested that readings be made on emissions. I will keep you informed.

Thank you for choosing Vantec.

"Andrew Diamond" wrote in message
. ..
My thesis, which will sound nutty, is the following: The use of my Vantec
Fan Controller to turn down the speed of my two intake 80mm case fans
causes
my floppy drive to fail. I will defend this below but now for the
setup,
pathology, and history.

Setup:
I recently installed a Vantec Silver Nexus NXP-205 Rheobus (with 4 fan
controllers) in my Kingwin K11 case. This case comes with 4 80mm fans two
of which are intake fans that sit at the bottom front of the machine. Not
much above these fans is the 3.5 inch drive external bays. The bottom bay
contains my Vantec fan controller and the one immediately above that has
my
floppy drive. These two intake fans, that are close to the floppy drive,
are attached to the first two of the 4 fan speed controllers. The other
two
controllers are attached to the other two distant case fans but apparently
have no effect adverse on my floppy drive (regardless of how I set their
speed controllers).

History/pathology:
The two intake fans are the loudest. I experimented with turning their
speed down by setting their fan controllers to the lowest setting and
noticing that the system temperature was not adversely effected (even
though
it was clear from the noise attenuation that the fans turned much much
slower). I don't use my floppy drive much but eventually, over a few
weeks,
when I did I noticed it behaved erratically. I did not think for a moment
that these two things were related. Specifically, some boot floppies (3
different ones) wouldn't boot almost all of the time. I thought the
floppies were bad but they booted in other machines.

I decided to quickly bypass this problem by making a boot CD from the
floppy, one that same computer because it had the burner, using Nero CD
burner, but when I tried that it couldn't read the files from the floppy
or
of that floppy's copy. I tried this multiple times.

To me, from my experience, this indicated that the floppy drive was
probably
bad (versus the motherboard or the cable).

However, when I swapped the floppy with that from another computer the
supposedly bad floppy drive worked great.

So, I figured it must be the floppy cable but when I tried the original
floppy cable with the original floppy drive on the second computer it
still
worked great!

Maybe a bad MB, hard to believe, so now I was confused so I reattached
that
cable and floppy to my original machine but I decided, really out of
laziness, to leave the floppy drive outside the case on a mat. This time
everything worked. I booted it to the floppy no less than 10 times with
no
problem (taking this floppy out and putting it back in - rebooting by
power
off, reset and CTRL-ALT-DEL).

I booted up into Win2k and using Nero I was now able to make my boot CD
from
this floppy. I did this multiple times.

I then repeated this whole process a few times removing the floppy from
the
drive bay (but still having it attached) and then putting it back in. I
even removed it from the case out the front still leaving the cables
attached so I wouldn't have to remove and reattach the floppy cables as
that
could theoretically be another source of error. Every time when the
floppy
drive was in the case it wouldn't work but when it was outside the case it
would.

Couldn't believe it!

Finally, with the floppy drive installed in the drive bay, I turned up the
fan controllers of the two intake fans all the way and NOW I was able to
use
Nero to successfully create the boot CD from the boot floppy.

I did this test multiple times. When I turned the two (front) fan
controllers all the way up the floppy always worked but when I turned it
all
the way down it always failed. It seems to fail even if the fan
controllers
half way but does work if its about 2/3 way.

Then I tried turning the other two fan controllers all the way down with
the
intake fan controllers all the way up. This worked fine. But when I
turned
one of the intake fan controllers all the way down the floppy failed.
Here's
a table:
The first four columns are the settings of fan controller 1 through 4
respectively (values: B - bottom power, T - Top Power, H - Half Power, 2 -
2/3 Power). The fifth column is whether the floppy drive failed (values:
F - failed, W - worked)

B B T T F
T T B B W
B B H H F
2 2 H H W
B T H H W
T B H H W
T B B B F

Conclusion:
So clearly, there is a problem related to the Vantec Fan controller
controlling these fans but after that I don't know what to say. The
tests
"B B T T F" and "T T B B W" indicate that the problem is context dependant
on which fans are being controlled. Specifically, turning the first two
fans, the ones close to the floppy drive, down will create a floppy failed
while turning down the other two fans may not. However "T B H H W" and "T
B
B B F" indicate that the problem is NOT related to those two close intake
fans because they were set the same in both tests but setting the far two
fans low caused a failure.

Ummm.

Andrew Diamond wrote:
Response from Vantec (originally they couldn't reproduce the problem.)
Andrew,

Correction, we let the test run for a while longer and the disk did become
corrupt. This is a very startling issue that has NEVER come up before.
Which is amazing since the NXP-205 is not new to the market. I will run
more tests tomorrow and hopefully come up with something more conclusive.
We have also contacted our engineers to see if they have a response. I also
requested that readings be made on emissions. I will keep you informed.

Thank you for choosing Vantec.

I'm impressed that they not only tried hard enough to repeat it but told you
about it too. It will be interesting to see what they come up with as the culprit.

Btw, if I were there, considering they claim it's a 'new' problem on a long
running product, first thing I'd do would be to check ECNs for recent changes
and the buyers to see if they've approved alternate suppliers. I've seen even
'equivalent' components from different manufacturers turn a working device into
a non functional one (because the original design engineer was a little *too*
'creative' with his implementation).