air flow sensor on PCB...

[Jan Panteltje]

On a sunny day (Wed, 12 Apr 2023 08:10:10 -0700) it happened John Larkin
<jjlarkin@highlandtechnology.com> wrote in
<nuhd3ilksk1n2v8c1j6rchslpnlfvvis11@4ax.com>:

....

On Wed, 12 Apr 2023 05:44:24 GMT, Jan Panteltje <alien@comet.invalid
wrote:
As to the \'hot wire\' system versus my \'diode system\'.
I do heat the diodes up to above environment with quite bit of current
so it is basically the same idea.
The diodes _have_ to be hotter than the max environment temperature
and using 2 in series put it at about about 1.4 V or so midway the 3.3 V ADC range.
I just use 4 diode pairs for north south west east, but with one
hot diode (or 2 in series) you get the same system as the hot wire thing.
https://panteltje.nl/panteltje/pic/wind_pic_thermal_udp/

What is that ethernet thing? I need to add ethernet to a Raspberry Pi
Pico, which doesn\'t have it. The current thinking is to add a WizNet
chip and an RJ45 with PoE magnetics.

It is bascically the ENC28J60
In the link I use an ebay module like this one:
https://www.ebay.com/itm/314356981186
That module has a SPI interface.

But in this project I use the chip itself:
https://panteltje.nl/panteltje/pic/ethernet_color_pic/

I wrote an UDP stack for it, in Microchip 18F asm...
Used in several projects, not all on the website, here are some:
https://panteltje.nl/panteltje/pic/step_pic_udp/
https://panteltje.nl/panteltje/pic/usa_pic/
UDP is not fault tolerant as to ask for a re-transmit on error, but suffices for many things,
has certain advantages too...

You probably want a TCP/IP stack for the Raspberry pico...
https://www.raspberrypi.com/news/how-to-add-ethernet-to-raspberry-pi-pico/
is the first link that google found for \"adding ethernet to a raspberry pico\"
Similar to the module I used?

You could probably replace the diodes with that sort of wire,
any idea what type of wire is used in that thing you have?

No idea. A hot wire anemometer is appealing but I don\'t have a good,
production-quality, way to get the hot wire.

Suitable wire probably doesn\'t solder well either.

Back to broken? light-bulbs!
Those have a large resistance difference between cold and hot.

 

[Jan Panteltje]

On a sunny day (Wed, 12 Apr 2023 08:00:58 -0700) it happened John Larkin
<jjlarkin@highlandtechnology.com> wrote in
<55hd3il43d21m7vd5d17oftvt4a6f7ar7b@4ax.com>:

I was on the chair looking at clouds and saw a tiny white dot. It
slowly expanded into a 6-sided snowflake shape and then faded away.
Then some giant bird flew over, black with orange wing undersides.
This being the sierra crest, it didn\'t need to flap its wings.

I feed the crows here... little pieces of pizza, they like that.

There are new streams running across Donner Pass Road and it\'s an
obstacle course of potholes, and the melt has just begun.

There are trees down all over San Francisco. The wind and rain toppled
a lot of shallow-rooted non-native trees.

https://www.dropbox.com/sh/w95izryfx600edu/AACkEds0Way2x1wfnyNVGxV-a?dl=0

Better watch out when walking there...

Yes. People have been killed.

https://www.dropbox.com/s/0964vbmzf03g8xy/20230330_155807.jpg?raw=1




Yesterday I was watching one of those survival programs...
Guy climbs all the way up in a coconut tree,
then slipped, and fell while holding on to the tree,
whole body skin red and scratched open, he had to give up.
Must be better ways to get a coconut down...

We shot pecans out of trees with a .22 rifle.

Yes, but those guys in the series had at the most a bow and arrow...
Maybe a long bamboo with a loop of wire on the end to pull the coconuts down ...

Well the sun is back here for now!
Heavy lightning predicted for the night...

 

[John Larkin]

On Wed, 12 Apr 2023 16:51:46 GMT, Jan Panteltje <alien@comet.invalid>
wrote:

On a sunny day (Wed, 12 Apr 2023 08:00:58 -0700) it happened John Larkin
jjlarkin@highlandtechnology.com> wrote in
55hd3il43d21m7vd5d17oftvt4a6f7ar7b@4ax.com>:

I was on the chair looking at clouds and saw a tiny white dot. It
slowly expanded into a 6-sided snowflake shape and then faded away.
Then some giant bird flew over, black with orange wing undersides.
This being the sierra crest, it didn\'t need to flap its wings.

I feed the crows here... little pieces of pizza, they like that.

Extensive controlled experiment shows that birds prefer Fritos.

There are new streams running across Donner Pass Road and it\'s an
obstacle course of potholes, and the melt has just begun.

There are trees down all over San Francisco. The wind and rain toppled
a lot of shallow-rooted non-native trees.

https://www.dropbox.com/sh/w95izryfx600edu/AACkEds0Way2x1wfnyNVGxV-a?dl=0

Better watch out when walking there...

Yes. People have been killed.

https://www.dropbox.com/s/0964vbmzf03g8xy/20230330_155807.jpg?raw=1




Yesterday I was watching one of those survival programs...
Guy climbs all the way up in a coconut tree,
then slipped, and fell while holding on to the tree,
whole body skin red and scratched open, he had to give up.
Must be better ways to get a coconut down...

We shot pecans out of trees with a .22 rifle.

Yes, but those guys in the series had at the most a bow and arrow...
Maybe a long bamboo with a loop of wire on the end to pull the coconuts down ...

Well the sun is back here for now!
Heavy lightning predicted for the night...

I miss lightning. We rarely see it on the west coast. The power lines
don\'t have the high ground wire between poles here.

 

[Joe Gwinn]

On Tue, 11 Apr 2023 13:22:48 -0700, John Larkin
<jjlarkin@highlandtechnology.com> wrote:

On Tue, 11 Apr 2023 15:38:45 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Fri, 07 Apr 2023 10:14:50 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Fri, 07 Apr 2023 12:41:44 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Thu, 06 Apr 2023 16:12:31 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Thu, 06 Apr 2023 13:51:54 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Thu, 06 Apr 2023 09:39:19 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Thu, 06 Apr 2023 12:00:15 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Wed, 05 Apr 2023 21:35:47 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

Given a PC board in a crate, like PCIe or PXI or VME or something,
what would be a good way to check air flow across the board, to verify
that the box fans and filters are OK? Assume this board can\'t access
the fan tachs or anything like that.

I was thinking that I might stick a small thinfilm RTD in the air
stream and measure its temperature at two different voltages, to
estimate its self-heating, which would vary with air flow.

One can get 1000-ohm platinum RTDs for small dollars:

.<https://www.newark.com/te-connectivity-sensors/nb-ptco-012/rtd-sensor-thin-film-platinum/dp/03AC1640

If the issue is just to tell if there is a lot of airflow, the
temperature delta does not need to be large, so a simple bridge with
AC source and synchronous detector should work.

Cheaper still are RTDs made of nickel, versus platinum. Nickel is OK
if the temperature in still air isn\'t too high.

Joe Gwinn

We use leaded ceramic slab and 1206 platinum RTDs, and a cute SOT23
nickel RTD, but they would be down flat on the PC board. I think it
would be better to get the sensor up in the air stream, so the
preferred gadget might be a TO92 transistor.

A 2N4402 costs us 7 cents. An RTD wouldn\'t break the bank, but simple
and cheap are games we play.

Metallic RTDs are quite stable, and the PTFD form (in the datasheet
above) can stick up 10 mm, which is about the same as a TO-92. One
can compare a SMD version with a nearby PTFD version, canceling local
ambient temp variation.

That\'s a good price. We have an essentially identical Minco part in
stock that costs $4. But it would be flimsy to mount on a board,
sticking up into the air flow.

I would guess that a TO-92 and a PTFD RTD are equally fragile in
practice. Don\'t know about the Minco. Part number?

My inventory program shows

MINCO S262PF
MINCO S247PF12
MURATA TRFA102A
OMEGA F3141

for the ceramic slab 1K thin-film RTD.

The 100 ohm 1206 is

ENERCORP PCS 1.1503.1
VISHAY PTS1206M1B100RP100

I\'ll look these up.

One thing that strikes me is that some Minco units I looked at had
thin silver wire leads, whereas the TO-92 leads will be kovar, and
maybe the PTFD RTDs as well.

If I use a TO92 transistor as a self-heating thermometer to sense air
flow, I\'d prefer the leads to be bad thermal conductors. Kovar is a
terrible heat conductor, 17 w/mk.

Yes, that helps too.

And it needs to self-heat to measure
flow.

Isn\'t that true in general?

A 100-ohm RTD will be easier to heat from 5 volts than the 1000-ohm
RTD, but one tenth as sensitive to temperature change, so there is a
tradeoff to be made.

I\'d probably have 12 or even 24 volts available in the case I\'m
considering. That would toast a 1K RTD.

Yes. Probably best to feed the RTD using constant-current sources.


I guess I could do my sequenced heat and cool time constant
measurement on it.

Sort of like a ring-down measurement - fit the log of the offset
versus time to a line, which can be done pretty fast - fever
thermometers do this and the algorithm is simple. So it could work.

It could be very simple, just a couple of timed limit checks. But with
6 GHz of ARM available, we could get fancy.

I would not depend on any single data point, as this data is likely to
be noisy and impulse-ridden.

Some kind of moving-window median filter is needed at the least, to
prevent random false positives and false negatives.

I like exponential smoothing, out = out + k*(in-out)

which models a 1st order lowpass.

The problem is that a high-energy impulse takes a long time to die
out, so direct impulse handling is usually worthwhile. Low-pass
filtering of what survives median or middle half filters also work
well.

How fast does loss of airflow need to be detected to avert damage?

I\'d think that one test per minute would be enough. We could probably
manage two. With two fans, nothing will fry soon.

A minute. That ought to suffice.

What if the airflow is blocked by a cleverly-placed random piece of
trash, preventing the fans from doing anything useful?

That\'s the reason to measure air flow!

Well, we are using temperature difference between RTDs (or
transistors) as a proxy for air flow, itself a proxy for cooling
capability. So we need to worry about places the proxy may fail to
predict.

Probably need to also sense the base temperature of the SMD RTD
regardless of air flow, to trigger shutdown in time.

I\'d have a separate pcb temperature sensor, an LM35 maybe.

Also a good approach.

Joe Gwinn

 

[John Larkin]

On Fri, 14 Apr 2023 15:25:43 -0400, Joe Gwinn <joegwinn@comcast.net>
wrote:

On Tue, 11 Apr 2023 13:22:48 -0700, John Larkin
jjlarkin@highlandtechnology.com> wrote:

On Tue, 11 Apr 2023 15:38:45 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Fri, 07 Apr 2023 10:14:50 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Fri, 07 Apr 2023 12:41:44 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Thu, 06 Apr 2023 16:12:31 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Thu, 06 Apr 2023 13:51:54 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Thu, 06 Apr 2023 09:39:19 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Thu, 06 Apr 2023 12:00:15 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Wed, 05 Apr 2023 21:35:47 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

Given a PC board in a crate, like PCIe or PXI or VME or something,
what would be a good way to check air flow across the board, to verify
that the box fans and filters are OK? Assume this board can\'t access
the fan tachs or anything like that.

I was thinking that I might stick a small thinfilm RTD in the air
stream and measure its temperature at two different voltages, to
estimate its self-heating, which would vary with air flow.

One can get 1000-ohm platinum RTDs for small dollars:

.<https://www.newark.com/te-connectivity-sensors/nb-ptco-012/rtd-sensor-thin-film-platinum/dp/03AC1640

If the issue is just to tell if there is a lot of airflow, the
temperature delta does not need to be large, so a simple bridge with
AC source and synchronous detector should work.

Cheaper still are RTDs made of nickel, versus platinum. Nickel is OK
if the temperature in still air isn\'t too high.

Joe Gwinn

We use leaded ceramic slab and 1206 platinum RTDs, and a cute SOT23
nickel RTD, but they would be down flat on the PC board. I think it
would be better to get the sensor up in the air stream, so the
preferred gadget might be a TO92 transistor.

A 2N4402 costs us 7 cents. An RTD wouldn\'t break the bank, but simple
and cheap are games we play.

Metallic RTDs are quite stable, and the PTFD form (in the datasheet
above) can stick up 10 mm, which is about the same as a TO-92. One
can compare a SMD version with a nearby PTFD version, canceling local
ambient temp variation.

That\'s a good price. We have an essentially identical Minco part in
stock that costs $4. But it would be flimsy to mount on a board,
sticking up into the air flow.

I would guess that a TO-92 and a PTFD RTD are equally fragile in
practice. Don\'t know about the Minco. Part number?

My inventory program shows

MINCO S262PF
MINCO S247PF12
MURATA TRFA102A
OMEGA F3141

for the ceramic slab 1K thin-film RTD.

The 100 ohm 1206 is

ENERCORP PCS 1.1503.1
VISHAY PTS1206M1B100RP100

I\'ll look these up.

One thing that strikes me is that some Minco units I looked at had
thin silver wire leads, whereas the TO-92 leads will be kovar, and
maybe the PTFD RTDs as well.

If I use a TO92 transistor as a self-heating thermometer to sense air
flow, I\'d prefer the leads to be bad thermal conductors. Kovar is a
terrible heat conductor, 17 w/mk.

Yes, that helps too.


And it needs to self-heat to measure
flow.

Isn\'t that true in general?

A 100-ohm RTD will be easier to heat from 5 volts than the 1000-ohm
RTD, but one tenth as sensitive to temperature change, so there is a
tradeoff to be made.

I\'d probably have 12 or even 24 volts available in the case I\'m
considering. That would toast a 1K RTD.

Yes. Probably best to feed the RTD using constant-current sources.


I guess I could do my sequenced heat and cool time constant
measurement on it.

Sort of like a ring-down measurement - fit the log of the offset
versus time to a line, which can be done pretty fast - fever
thermometers do this and the algorithm is simple. So it could work.

It could be very simple, just a couple of timed limit checks. But with
6 GHz of ARM available, we could get fancy.

I would not depend on any single data point, as this data is likely to
be noisy and impulse-ridden.

Some kind of moving-window median filter is needed at the least, to
prevent random false positives and false negatives.

I like exponential smoothing, out = out + k*(in-out)

which models a 1st order lowpass.

The problem is that a high-energy impulse takes a long time to die
out, so direct impulse handling is usually worthwhile. Low-pass
filtering of what survives median or middle half filters also work
well.

An impulse is still real signal. I don\'t expect an air flow
measurement to be precise and it can be lowpass filtered hard; the
hazards to the system are long-term and thermal.

It\'s trivial to do a 2nd order lowpass in c or an FPGA, if you want
spikes to wash out sooner.

How fast does loss of airflow need to be detected to avert damage?

I\'d think that one test per minute would be enough. We could probably
manage two. With two fans, nothing will fry soon.

A minute. That ought to suffice.

What if the airflow is blocked by a cleverly-placed random piece of
trash, preventing the fans from doing anything useful?

That\'s the reason to measure air flow!

Well, we are using temperature difference between RTDs (or
transistors) as a proxy for air flow, itself a proxy for cooling
capability. So we need to worry about places the proxy may fail to
predict.

I can\'t predict what direction air may flow in a card cage, so I
prefer an omnidirectional sensor, namely a cylinder.

Probably need to also sense the base temperature of the SMD RTD
regardless of air flow, to trigger shutdown in time.

I\'d have a separate pcb temperature sensor, an LM35 maybe.

Also a good approach.

Joe Gwinn

This would just be a bullet feature to help sell the controller, so it
doesn\'t need to be super precise. Just an air flow warning. Maybe a
fan failed, maybe a cat is napping on the intake.

 

[M Nelson]

On Thursday, April 6, 2023 at 12:36:04 AM UTC-4, John Larkin wrote:

Given a PC board in a crate, like PCIe or PXI or VME or something,
what would be a good way to check air flow across the board, to verify
that the box fans and filters are OK? Assume this board can\'t access
the fan tachs or anything like that.

I was thinking that I might stick a small thinfilm RTD in the air
stream and measure its temperature at two different voltages, to
estimate its self-heating, which would vary with air flow.

The classic broken light bulb hot-wire anemometer is a nuisance.

Carbon comp resistor?

This can\'t be a new problem. Any other suggestions?

 

[M Nelson]

If you really want something that is not analogous to a hot wire, then, depending on the expected air speed, and what sort of analog input capability you have, an ultrasound transducer and receiver can work. My experience is with ceramics for that purpose. But I suspect piezo film would work too. The trick is to avoid multi-pathing - short pulse and look at first signal only. And of course, and maybe even simpler, you might try using the distortion of a single piece of piezo film directly.

 

[John Larkin]

On Fri, 14 Apr 2023 15:31:50 -0700 (PDT), M Nelson
<drmcnelson@gmail.com> wrote:

>If you really want something that is not analogous to a hot wire, then, depending on the expected air speed, and what sort of analog input capability you have, an ultrasound transducer and receiver can work. My experience is with ceramics for that purpose. But I suspect piezo film would work too. The trick is to avoid multi-pathing - short pulse and look at first signal only. And of course, and maybe even simpler, you might try using the distortion of a single piece of piezo film directly.

Ultrasound measurement will be directional. And big. And expensive.

 

[John Larkin]

On Wed, 12 Apr 2023 05:53:54 GMT, Jan Panteltje <alien@comet.invalid>
wrote:

On a sunny day (Tue, 11 Apr 2023 09:47:54 -0700) it happened John Larkin
jjlarkin@highlandtechnology.com> wrote in
oa3b3i97igik4l7l1k7lm74fk7cuelsc4h@4ax.com>:

On Tue, 11 Apr 2023 07:56:13 -0700, John Larkin
jjlarkin@highlandtechnology.com> wrote:

On Mon, 10 Apr 2023 06:12:11 GMT, Jan Panteltje <alien@comet.invalid
wrote:

On a sunny day (Sun, 09 Apr 2023 09:32:44 -0700) it happened John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote in
9mp53ipjlshtur3rin9p0k10tnmv16i04f@4ax.com>:

It\'s trivial if it helps up get a multi-million-dollar product right,
or saves even a day of engineering time.

We compared it against a couple of the propeller-type air speed meters
and it agreed nicely.

The other gadget, much cheaper, is to go to Chinatown and buy a bunch
of incense sticks and use the smoke to visualize air flow patterns.

OK, that is the one!
Marihuana?
Weed?

\'High\' speed fan?

You may have a market there!

But I better stay with my 10 dollar propellor thing ;-)
first flowers in the garden are already blooming,
and insects come with it..
Last year I had grapes from the garden, not bad at all!



We had a cold, very wet winter, and then it got warm fast. The plants
are exploding, so much green it hurts our eyes. The roses and the
lemon tree are sagging with fruit and flowers. I expect a good
blackberry year in our canyon.

Lots of trees fell in San Franciso, from the water and wind.

I have some awesome snow pictures too. I\'ll post them soon. It\'s warm
up here in the sierras now and the runoff is already causing troubles.

https://www.dropbox.com/sh/p6854tn4p5u1ih8/AABu27KEF4wBVeVYFFxtr8BUa?dl=0

Wow! That is very beautiful!


There are new streams running across Donner Pass Road and it\'s an
obstacle course of potholes, and the melt has just begun.

This is our driveway.

https://www.dropbox.com/s/gkt76y8prm6vu69/20230410_150109.jpg?raw=1

This morning the trickles of water were streams, eroding the dirt from
under our driveway asphalt. Envision giant expensive potholes.
Multiply that by maybe a million other locations nearby.

 

[Jan Panteltje]

On a sunny day (Sun, 16 Apr 2023 20:24:09 -0700) it happened John Larkin
<jlarkin@highlandSNIPMEtechnology.com> wrote in
<8lep3i9us0ac8nvbco0c2s0qgj7ia8dj0g@4ax.com>:

This is our driveway.

https://www.dropbox.com/s/gkt76y8prm6vu69/20230410_150109.jpg?raw=1

This morning the trickles of water were streams, eroding the dirt from
under our driveway asphalt. Envision giant expensive potholes.
Multiply that by maybe a million other locations nearby.

Yes, that road needs some work!
Rain has stopped here for now, wind mainly from east..
When it turns west next week (from the sea) it will be wet again...

BTW ultrasonic air speed measurement is not expensive,
the distance measurement modules that you can modify for that purpose are less than 2 dollars on ebay.
Directional yes. but fans usually only blow one way.
https://panteltje.nl/pub/wind_speed_by_differential_2_ebay_distance_meters_IMG_4891.JPG
The ebay module requires a pulse, and returns a pulse a bit later depending on the distance.
For fun I XORed the pulse form one looking one way with one looking the other way

The output pulses then cancel if there is no air flow,
you could then integrate (low pass) the XOR output pulses and even drive an analog meter.
You do not need 2 units if you can do the timing in FPGA or software.
So the only modification is to remove say the TX module from the module PCB and place it so it sees the RX module directly.

 

[M Nelson]

On Sunday, April 16, 2023 at 11:10:16 PM UTC-4, John Larkin wrote:

On Fri, 14 Apr 2023 15:31:50 -0700 (PDT), M Nelson
drmcn...@gmail.com> wrote:

Ultrasound measurement will be directional. And big. And expensive.

Use a circular mode, or, two arms

About 1 or 2 cm at f>=30khz.

Traditional, simple ultrasonic transducers are available under $4 on Mouser..

Or, you might find the following datasheet interesting, though a little more expensive at $24. Circuits are included in the datasheet.
https://www.mouser.com/datasheet/2/418/80kHz_Ultrasound_Transducer-775561.pdf

You might try just measuring phase. But then, spi digitizers are cheap too.

Remember to avoid multipathing or driving resonances, i.e. keep the pulse short. You can model the transducer and driver in spice, and play with the values for the passives until you get a short pulse, that essentially means low Q. With the RLC driver in the data sheet you can get it down to a one or two cylce pulse.

Sorry, it is a little bit of a project rather than just plug it in and use it.

The other option I mentioned was just letting the air pressure deform a small pieze of piezo film. I have not tried that for your application, but I do know from experience that they are sensitive to deformation.

 

[Joe Gwinn]

The issue is that On Fri, 14 Apr 2023 15:12:19 -0700, John Larkin
<jjlarkin@highlandtechnology.com> wrote:

On Fri, 14 Apr 2023 15:25:43 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Tue, 11 Apr 2023 13:22:48 -0700, John Larkin
jjlarkin@highlandtechnology.com> wrote:

On Tue, 11 Apr 2023 15:38:45 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Fri, 07 Apr 2023 10:14:50 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Fri, 07 Apr 2023 12:41:44 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Thu, 06 Apr 2023 16:12:31 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Thu, 06 Apr 2023 13:51:54 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Thu, 06 Apr 2023 09:39:19 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Thu, 06 Apr 2023 12:00:15 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Wed, 05 Apr 2023 21:35:47 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

Given a PC board in a crate, like PCIe or PXI or VME or something,
what would be a good way to check air flow across the board, to verify
that the box fans and filters are OK? Assume this board can\'t access
the fan tachs or anything like that.

I was thinking that I might stick a small thinfilm RTD in the air
stream and measure its temperature at two different voltages, to
estimate its self-heating, which would vary with air flow.

One can get 1000-ohm platinum RTDs for small dollars:

.<https://www.newark.com/te-connectivity-sensors/nb-ptco-012/rtd-sensor-thin-film-platinum/dp/03AC1640

If the issue is just to tell if there is a lot of airflow, the
temperature delta does not need to be large, so a simple bridge with
AC source and synchronous detector should work.

Cheaper still are RTDs made of nickel, versus platinum. Nickel is OK
if the temperature in still air isn\'t too high.

Joe Gwinn

We use leaded ceramic slab and 1206 platinum RTDs, and a cute SOT23
nickel RTD, but they would be down flat on the PC board. I think it
would be better to get the sensor up in the air stream, so the
preferred gadget might be a TO92 transistor.

A 2N4402 costs us 7 cents. An RTD wouldn\'t break the bank, but simple
and cheap are games we play.

Metallic RTDs are quite stable, and the PTFD form (in the datasheet
above) can stick up 10 mm, which is about the same as a TO-92. One
can compare a SMD version with a nearby PTFD version, canceling local
ambient temp variation.

That\'s a good price. We have an essentially identical Minco part in
stock that costs $4. But it would be flimsy to mount on a board,
sticking up into the air flow.

I would guess that a TO-92 and a PTFD RTD are equally fragile in
practice. Don\'t know about the Minco. Part number?

My inventory program shows

MINCO S262PF
MINCO S247PF12
MURATA TRFA102A
OMEGA F3141

for the ceramic slab 1K thin-film RTD.

The 100 ohm 1206 is

ENERCORP PCS 1.1503.1
VISHAY PTS1206M1B100RP100

I\'ll look these up.

One thing that strikes me is that some Minco units I looked at had
thin silver wire leads, whereas the TO-92 leads will be kovar, and
maybe the PTFD RTDs as well.

If I use a TO92 transistor as a self-heating thermometer to sense air
flow, I\'d prefer the leads to be bad thermal conductors. Kovar is a
terrible heat conductor, 17 w/mk.

Yes, that helps too.


And it needs to self-heat to measure
flow.

Isn\'t that true in general?

A 100-ohm RTD will be easier to heat from 5 volts than the 1000-ohm
RTD, but one tenth as sensitive to temperature change, so there is a
tradeoff to be made.

I\'d probably have 12 or even 24 volts available in the case I\'m
considering. That would toast a 1K RTD.

Yes. Probably best to feed the RTD using constant-current sources.


I guess I could do my sequenced heat and cool time constant
measurement on it.

Sort of like a ring-down measurement - fit the log of the offset
versus time to a line, which can be done pretty fast - fever
thermometers do this and the algorithm is simple. So it could work.

It could be very simple, just a couple of timed limit checks. But with
6 GHz of ARM available, we could get fancy.

I would not depend on any single data point, as this data is likely to
be noisy and impulse-ridden.

Some kind of moving-window median filter is needed at the least, to
prevent random false positives and false negatives.

I like exponential smoothing, out = out + k*(in-out)

which models a 1st order lowpass.

The problem is that a high-energy impulse takes a long time to die
out, so direct impulse handling is usually worthwhile. Low-pass
filtering of what survives median or middle half filters also work
well.


An impulse is still real signal. I don\'t expect an air flow
measurement to be precise and it can be lowpass filtered hard; the
hazards to the system are long-term and thermal.

The issue is that not all impulses are equal. Gaussian noise does
have peaks, but it does average out to reasonably well behaved values.

But impulse noise (as from a pulse jammer or a sparking contact
somewhere) does not follow Gaussian statistics, and can inject far
more than the average power level. Mathematically, this is often
modeled as the sum of two independent Gaussian noise signals, one of a
far larger standard deviation than the other. This kind of signal
causes long-tail effects in linear filters of all kinds, and it\'s
pretty much universal to provide a non-linear pre-filter to clamp or
blank those impulses before the are permitted to enter the linear
filter stage.

It\'s trivial to do a 2nd order lowpass in c or an FPGA, if you want
spikes to wash out sooner.

Yes, one can do all this in C or a FPGA, but one still needs the
non-linear impulse filter before the linear smoothing filter.

How fast does loss of airflow need to be detected to avert damage?

I\'d think that one test per minute would be enough. We could probably
manage two. With two fans, nothing will fry soon.

A minute. That ought to suffice.

What if the airflow is blocked by a cleverly-placed random piece of
trash, preventing the fans from doing anything useful?

That\'s the reason to measure air flow!

Well, we are using temperature difference between RTDs (or
transistors) as a proxy for air flow, itself a proxy for cooling
capability. So we need to worry about places the proxy may fail to
predict.



I can\'t predict what direction air may flow in a card cage, so I
prefer an omnidirectional sensor, namely a cylinder.

Probably need to also sense the base temperature of the SMD RTD
regardless of air flow, to trigger shutdown in time.

I\'d have a separate pcb temperature sensor, an LM35 maybe.

Also a good approach.

Joe Gwinn

This would just be a bullet feature to help sell the controller, so it
doesn\'t need to be super precise. Just an air flow warning. Maybe a
fan failed, maybe a cat is napping on the intake.

Yes. One thing that was learned in the GPS receiver realm was that
outdoor antennas for precision receivers need to be impossible for a
seagull to perch upon.

And then there is the Starlink Effect on cats:

..<https://www.smithsonianmag.com/smart-news/outdoor-cats-are-using-500-starlink-satellite-dishes-as-self-heating-beds-180979401/>

Joe Gwinn

 

[M Nelson]

Perhaps another, even simpler solution

How about measuring power (or voltage and current) into the fan? That should surely tell you whether the fan is working and whether it is pushing air.

 

[John Larkin]

On Mon, 17 Apr 2023 15:26:36 -0400, Joe Gwinn <joegwinn@comcast.net>
wrote:

The issue is that On Fri, 14 Apr 2023 15:12:19 -0700, John Larkin
jjlarkin@highlandtechnology.com> wrote:

On Fri, 14 Apr 2023 15:25:43 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Tue, 11 Apr 2023 13:22:48 -0700, John Larkin
jjlarkin@highlandtechnology.com> wrote:

On Tue, 11 Apr 2023 15:38:45 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Fri, 07 Apr 2023 10:14:50 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Fri, 07 Apr 2023 12:41:44 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Thu, 06 Apr 2023 16:12:31 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Thu, 06 Apr 2023 13:51:54 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Thu, 06 Apr 2023 09:39:19 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Thu, 06 Apr 2023 12:00:15 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Wed, 05 Apr 2023 21:35:47 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

Given a PC board in a crate, like PCIe or PXI or VME or something,
what would be a good way to check air flow across the board, to verify
that the box fans and filters are OK? Assume this board can\'t access
the fan tachs or anything like that.

I was thinking that I might stick a small thinfilm RTD in the air
stream and measure its temperature at two different voltages, to
estimate its self-heating, which would vary with air flow.

One can get 1000-ohm platinum RTDs for small dollars:

.<https://www.newark.com/te-connectivity-sensors/nb-ptco-012/rtd-sensor-thin-film-platinum/dp/03AC1640

If the issue is just to tell if there is a lot of airflow, the
temperature delta does not need to be large, so a simple bridge with
AC source and synchronous detector should work.

Cheaper still are RTDs made of nickel, versus platinum. Nickel is OK
if the temperature in still air isn\'t too high.

Joe Gwinn

We use leaded ceramic slab and 1206 platinum RTDs, and a cute SOT23
nickel RTD, but they would be down flat on the PC board. I think it
would be better to get the sensor up in the air stream, so the
preferred gadget might be a TO92 transistor.

A 2N4402 costs us 7 cents. An RTD wouldn\'t break the bank, but simple
and cheap are games we play.

Metallic RTDs are quite stable, and the PTFD form (in the datasheet
above) can stick up 10 mm, which is about the same as a TO-92. One
can compare a SMD version with a nearby PTFD version, canceling local
ambient temp variation.

That\'s a good price. We have an essentially identical Minco part in
stock that costs $4. But it would be flimsy to mount on a board,
sticking up into the air flow.

I would guess that a TO-92 and a PTFD RTD are equally fragile in
practice. Don\'t know about the Minco. Part number?

My inventory program shows

MINCO S262PF
MINCO S247PF12
MURATA TRFA102A
OMEGA F3141

for the ceramic slab 1K thin-film RTD.

The 100 ohm 1206 is

ENERCORP PCS 1.1503.1
VISHAY PTS1206M1B100RP100

I\'ll look these up.

One thing that strikes me is that some Minco units I looked at had
thin silver wire leads, whereas the TO-92 leads will be kovar, and
maybe the PTFD RTDs as well.

If I use a TO92 transistor as a self-heating thermometer to sense air
flow, I\'d prefer the leads to be bad thermal conductors. Kovar is a
terrible heat conductor, 17 w/mk.

Yes, that helps too.


And it needs to self-heat to measure
flow.

Isn\'t that true in general?

A 100-ohm RTD will be easier to heat from 5 volts than the 1000-ohm
RTD, but one tenth as sensitive to temperature change, so there is a
tradeoff to be made.

I\'d probably have 12 or even 24 volts available in the case I\'m
considering. That would toast a 1K RTD.

Yes. Probably best to feed the RTD using constant-current sources.


I guess I could do my sequenced heat and cool time constant
measurement on it.

Sort of like a ring-down measurement - fit the log of the offset
versus time to a line, which can be done pretty fast - fever
thermometers do this and the algorithm is simple. So it could work.

It could be very simple, just a couple of timed limit checks. But with
6 GHz of ARM available, we could get fancy.

I would not depend on any single data point, as this data is likely to
be noisy and impulse-ridden.

Some kind of moving-window median filter is needed at the least, to
prevent random false positives and false negatives.

I like exponential smoothing, out = out + k*(in-out)

which models a 1st order lowpass.

The problem is that a high-energy impulse takes a long time to die
out, so direct impulse handling is usually worthwhile. Low-pass
filtering of what survives median or middle half filters also work
well.


An impulse is still real signal. I don\'t expect an air flow
measurement to be precise and it can be lowpass filtered hard; the
hazards to the system are long-term and thermal.

The issue is that not all impulses are equal. Gaussian noise does
have peaks, but it does average out to reasonably well behaved values.

I\'ll be measuring Vbe of a diode-connected transistor to measure
junction temperature. I don\'t expect a lot of impulse noise, and not
much gaussian noise. Figure a 700 mV level that changes a couple of mV
per degree C. Lots of signal compared, say, to a thermocouple.

But impulse noise (as from a pulse jammer or a sparking contact
somewhere) does not follow Gaussian statistics, and can inject far
more than the average power level. Mathematically, this is often
modeled as the sum of two independent Gaussian noise signals, one of a
far larger standard deviation than the other. This kind of signal
causes long-tail effects in linear filters of all kinds, and it\'s
pretty much universal to provide a non-linear pre-filter to clamp or
blank those impulses before the are permitted to enter the linear
filter stage.


It\'s trivial to do a 2nd order lowpass in c or an FPGA, if you want
spikes to wash out sooner.

Yes, one can do all this in C or a FPGA, but one still needs the
non-linear impulse filter before the linear smoothing filter.

Probably not.

 

[Joe Gwinn]

On Mon, 17 Apr 2023 12:40:53 -0700, John Larkin
<jlarkin@highlandSNIPMEtechnology.com> wrote:

On Mon, 17 Apr 2023 15:26:36 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

The issue is that On Fri, 14 Apr 2023 15:12:19 -0700, John Larkin
jjlarkin@highlandtechnology.com> wrote:

On Fri, 14 Apr 2023 15:25:43 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Tue, 11 Apr 2023 13:22:48 -0700, John Larkin
jjlarkin@highlandtechnology.com> wrote:

On Tue, 11 Apr 2023 15:38:45 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Fri, 07 Apr 2023 10:14:50 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Fri, 07 Apr 2023 12:41:44 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Thu, 06 Apr 2023 16:12:31 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Thu, 06 Apr 2023 13:51:54 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Thu, 06 Apr 2023 09:39:19 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

On Thu, 06 Apr 2023 12:00:15 -0400, Joe Gwinn <joegwinn@comcast.net
wrote:

On Wed, 05 Apr 2023 21:35:47 -0700, John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote:

Given a PC board in a crate, like PCIe or PXI or VME or something,
what would be a good way to check air flow across the board, to verify
that the box fans and filters are OK? Assume this board can\'t access
the fan tachs or anything like that.

I was thinking that I might stick a small thinfilm RTD in the air
stream and measure its temperature at two different voltages, to
estimate its self-heating, which would vary with air flow.

One can get 1000-ohm platinum RTDs for small dollars:

.<https://www.newark.com/te-connectivity-sensors/nb-ptco-012/rtd-sensor-thin-film-platinum/dp/03AC1640

If the issue is just to tell if there is a lot of airflow, the
temperature delta does not need to be large, so a simple bridge with
AC source and synchronous detector should work.

Cheaper still are RTDs made of nickel, versus platinum. Nickel is OK
if the temperature in still air isn\'t too high.

Joe Gwinn

We use leaded ceramic slab and 1206 platinum RTDs, and a cute SOT23
nickel RTD, but they would be down flat on the PC board. I think it
would be better to get the sensor up in the air stream, so the
preferred gadget might be a TO92 transistor.

A 2N4402 costs us 7 cents. An RTD wouldn\'t break the bank, but simple
and cheap are games we play.

Metallic RTDs are quite stable, and the PTFD form (in the datasheet
above) can stick up 10 mm, which is about the same as a TO-92. One
can compare a SMD version with a nearby PTFD version, canceling local
ambient temp variation.

That\'s a good price. We have an essentially identical Minco part in
stock that costs $4. But it would be flimsy to mount on a board,
sticking up into the air flow.

I would guess that a TO-92 and a PTFD RTD are equally fragile in
practice. Don\'t know about the Minco. Part number?

My inventory program shows

MINCO S262PF
MINCO S247PF12
MURATA TRFA102A
OMEGA F3141

for the ceramic slab 1K thin-film RTD.

The 100 ohm 1206 is

ENERCORP PCS 1.1503.1
VISHAY PTS1206M1B100RP100

I\'ll look these up.

One thing that strikes me is that some Minco units I looked at had
thin silver wire leads, whereas the TO-92 leads will be kovar, and
maybe the PTFD RTDs as well.

If I use a TO92 transistor as a self-heating thermometer to sense air
flow, I\'d prefer the leads to be bad thermal conductors. Kovar is a
terrible heat conductor, 17 w/mk.

Yes, that helps too.


And it needs to self-heat to measure
flow.

Isn\'t that true in general?

A 100-ohm RTD will be easier to heat from 5 volts than the 1000-ohm
RTD, but one tenth as sensitive to temperature change, so there is a
tradeoff to be made.

I\'d probably have 12 or even 24 volts available in the case I\'m
considering. That would toast a 1K RTD.

Yes. Probably best to feed the RTD using constant-current sources.


I guess I could do my sequenced heat and cool time constant
measurement on it.

Sort of like a ring-down measurement - fit the log of the offset
versus time to a line, which can be done pretty fast - fever
thermometers do this and the algorithm is simple. So it could work.

It could be very simple, just a couple of timed limit checks. But with
6 GHz of ARM available, we could get fancy.

I would not depend on any single data point, as this data is likely to
be noisy and impulse-ridden.

Some kind of moving-window median filter is needed at the least, to
prevent random false positives and false negatives.

I like exponential smoothing, out = out + k*(in-out)

which models a 1st order lowpass.

The problem is that a high-energy impulse takes a long time to die
out, so direct impulse handling is usually worthwhile. Low-pass
filtering of what survives median or middle half filters also work
well.


An impulse is still real signal. I don\'t expect an air flow
measurement to be precise and it can be lowpass filtered hard; the
hazards to the system are long-term and thermal.

The issue is that not all impulses are equal. Gaussian noise does
have peaks, but it does average out to reasonably well behaved values.

I\'ll be measuring Vbe of a diode-connected transistor to measure
junction temperature. I don\'t expect a lot of impulse noise, and not
much gaussian noise. Figure a 700 mV level that changes a couple of mV
per degree C. Lots of signal compared, say, to a thermocouple.


But impulse noise (as from a pulse jammer or a sparking contact
somewhere) does not follow Gaussian statistics, and can inject far
more than the average power level. Mathematically, this is often
modeled as the sum of two independent Gaussian noise signals, one of a
far larger standard deviation than the other. This kind of signal
causes long-tail effects in linear filters of all kinds, and it\'s
pretty much universal to provide a non-linear pre-filter to clamp or
blank those impulses before the are permitted to enter the linear
filter stage.


It\'s trivial to do a 2nd order lowpass in c or an FPGA, if you want
spikes to wash out sooner.

Yes, one can do all this in C or a FPGA, but one still needs the
non-linear impulse filter before the linear smoothing filter.

Probably not.

Given the low impedance of a forward-biased PN junction, it may be
herd for external EMI to get in, which would be nice. We\'ll soon
know.

But in the big radar systems of my experience, we always needed a
non-linear impulse blanker or interpolator. This was generally proven
in the lab when they were trading various sign el-processing
approaches off. For low sidelobes (in time), this mattered a lot.

Fault Detection Fault Isolation has a parallel problem, that many
detected faults are in fact transient false alarms, so one always has
some kind of M-of-N (often 3-of-5) filter to ignore blips.

Joe Gwinn

 

[Jasen Betts]

On 2023-04-17, M Nelson <drmcnelson@gmail.com> wrote:

Perhaps another, even simpler solution

How about measuring power (or voltage and current) into the fan? That should surely tell you whether the fan is working and whether it is pushing air.

increased bearing friction will look like moving air even when it\'s not.



--
Jasen.
🇺🇦 Слава Україні

 

[John Larkin]

On Mon, 17 Apr 2023 12:40:18 -0700 (PDT), M Nelson
<drmcnelson@gmail.com> wrote:

Perhaps another, even simpler solution

How about measuring power (or voltage and current) into the fan? That should surely tell you whether the fan is working and whether it is pushing air.

As noted, the flow sensor needs to be on a controller board that plugs
into a generic VME crate. That board has no access to the fans.

 

[M Nelson]

> increased bearing friction will look like moving air even when it\'s not.

That seems unlikely - that the load vs speed for bearing friction would really look the same as that for air.

It sort of doesn\'t make sense also, if they would be even similar magnitude, it means the fan is not very efficient. And even then, as long as the friction is not very much larger, you still have a curve for load vs speed that includes a contribution from pushing the air.

And if the bearings are failing, the curve will likely be changing - upward..

So overall, it seems like it could be a pretty nifty diagnostic. The concern for bearing friction doesn\'t seem like a good reason not to try it, quite the opposite.

 

[M Nelson]

As noted, the flow sensor needs to be on a controller board that plugs
into a generic VME crate. That board has no access to the fans.

Fine, then the choices would seem to be (a) thermal conduction of moving air, (b) propagation of sound, or (c) direct pressure on a deformable object.. All three are cheap. The first and third are probably the most reliable, the third might be the more accurate with changes in temperature, humidity, and air pressure.

Another perhaps more exotic idea that comes to mind is deflection of electrons. But I think the first three might be more practical.

 

[M Nelson]

On Tuesday, April 18, 2023 at 10:05:35 AM UTC-4, M Nelson wrote:

As noted, the flow sensor needs to be on a controller board that plugs
into a generic VME crate. That board has no access to the fans.
Fine, then the choices would seem to be (a) thermal conduction of moving air, (b) propagation of sound, or (c) direct pressure on a deformable object. All three are cheap. The first and third are probably the most reliable, the third might be the more accurate with changes in temperature, humidity, and air pressure.

Another perhaps more exotic idea that comes to mind is deflection of electrons. But I think the first three might be more practical.

One more almost out of the box idea. Many MCUs and CPUs nowadays have temperature sensors. They are in effect thermal conduction type anemometers.. Find a cheap one, and running a simple loop, clock it just fast enough to produce a small increment in temperature over the ambient when the fan is off. You have to admit, in some ways its a pretty relevant measurement for the purpose.