Skip navigation.
Home
The QNX Community Portal

View topic - QNX Public Newsgroups

QNX Public Newsgroups

Read-only archive of qnx.cafe (generail discussion forum) at inn.qnx.com

Re: More Tri-M, or the problems of getting ruggedized crunch

Postby Colin Burgess » Tue Sep 27, 2005 1:19 pm

Does the vehicle have an air conditioner? ;v)

Seriously, is there some way you could jury-rig a source of cool air to vent into
the boxes?

John Nagle wrote:
Evan Hillas wrote:

John Nagle wrote:

It's not the power supply. Tri-M is trying to figure out
what's failing.

The problem seems to be airflow related; the packaging
and fan positioning really don't match the motherboard
well.


It's airflow. We blew another NOVA-8890 during field testing.
Temperature 42C at inlet, 68C at CPU heat sink, measured
with an infrared temperature sensor immediately after
removing the cover. The computer's spec said 70C, but
that turns out to be the board, not the systems as cased
with fans. It's in a little box about 1U high, but
the airflow wasn't well worked out. The exhaust fan
for the CPU blows at a steel cover about 5mm from
the fan.
So we're off to the DARPA Grand Challenge with one
computer of three. One is on the vehicle, one is being
shipped back from Tri-M, and one is being shipped to
Tri-M for repair. The one on the vehicle is running
with its top cover off. What's a good Pentium IV or faster x86
machine with
solidly reliable industrial temperature and vibration ranges?
When we picked these things two years ago there were few
options, but there must have been progress by now? What
are people using for CPU-intensive embedded work in tough
environments?

In retrospect, we should have used a stack of Panasonic
Toughbooks.
John Nagle
Team Overbot


--
cburgess@qnx.com
Colin Burgess
 

Re: More Tri-M, or the problems of getting ruggedized crunch

Postby Igor Kovalenko » Tue Sep 27, 2005 5:40 pm

And a power seat with lumbar support will help tremendously.

"Colin Burgess" <cburgess@qnx.com> wrote in message
news:dhbgrf$m9o$1@inn.qnx.com...
Does the vehicle have an air conditioner? ;v)

Seriously, is there some way you could jury-rig a source of cool air to
vent into
the boxes?

John Nagle wrote:
Evan Hillas wrote:

John Nagle wrote:

It's not the power supply. Tri-M is trying to figure out
what's failing.

The problem seems to be airflow related; the packaging
and fan positioning really don't match the motherboard
well.


It's airflow. We blew another NOVA-8890 during field testing.
Temperature 42C at inlet, 68C at CPU heat sink, measured
with an infrared temperature sensor immediately after
removing the cover. The computer's spec said 70C, but
that turns out to be the board, not the systems as cased
with fans. It's in a little box about 1U high, but
the airflow wasn't well worked out. The exhaust fan
for the CPU blows at a steel cover about 5mm from
the fan.
So we're off to the DARPA Grand Challenge with one
computer of three. One is on the vehicle, one is being
shipped back from Tri-M, and one is being shipped to
Tri-M for repair. The one on the vehicle is running
with its top cover off. What's a good Pentium IV or faster x86
machine with
solidly reliable industrial temperature and vibration ranges?
When we picked these things two years ago there were few
options, but there must have been progress by now? What
are people using for CPU-intensive embedded work in tough
environments?

In retrospect, we should have used a stack of Panasonic
Toughbooks. John Nagle
Team Overbot


--
cburgess@qnx.com
Igor Kovalenko
 

Re: More Tri-M, or the problems of getting ruggedized crunch

Postby emuis » Thu Sep 29, 2005 8:01 pm

Best of luck John, hope all goes well! We'll be watching.

E.


John Nagle wrote:
Evan Hillas wrote:

John Nagle wrote:

It's not the power supply. Tri-M is trying to figure out
what's failing.

The problem seems to be airflow related; the packaging
and fan positioning really don't match the motherboard
well.


It's airflow. We blew another NOVA-8890 during field testing.
Temperature 42C at inlet, 68C at CPU heat sink, measured
with an infrared temperature sensor immediately after
removing the cover. The computer's spec said 70C, but
that turns out to be the board, not the systems as cased
with fans. It's in a little box about 1U high, but
the airflow wasn't well worked out. The exhaust fan
for the CPU blows at a steel cover about 5mm from
the fan.
So we're off to the DARPA Grand Challenge with one
computer of three. One is on the vehicle, one is being
shipped back from Tri-M, and one is being shipped to
Tri-M for repair. The one on the vehicle is running
with its top cover off. What's a good Pentium IV or faster x86
machine with
solidly reliable industrial temperature and vibration ranges?
When we picked these things two years ago there were few
options, but there must have been progress by now? What
are people using for CPU-intensive embedded work in tough
environments?

In retrospect, we should have used a stack of Panasonic
Toughbooks.
John Nagle
Team Overbot
emuis
New Member
 
Posts: 7
Joined: Wed Aug 31, 2005 1:28 pm

Re: More Mindready, and Tri-M

Postby Warren Deitch » Thu Sep 29, 2005 10:14 pm

John,
I have been watching with keen interest your efforts (and
sympathy for the problems).

I do not know how feasable this is -- put a peltier device in
front of the air inlet to lower the air temperature.

Warren



John Nagle wrote:
Evan Hillas wrote:

I presume this: http://www.tri-m.com/products/icp/nova8890.html

It sounds like you are using a supplied chassis w/power supply. It
might not fit mechanically but try plugging in a standard PC power
supply the next time one of the Novas stops working.


Evan


It's not the power supply. Tri-M is trying to figure out
what's failing.

The problem seems to be airflow related; the packaging
and fan positioning really don't match the motherboard
well.

John Nagle
Warren Deitch
 

Previous

Return to qnx.cafe

Who is online

Users browsing this forum: No registered users and 1 guest

cron