Thread: Testo 405i Hot Wire Anemometer inappropriate for small (150mm) duct

So this would beg the question, are there any affordable hot wire anemometers where the sensors are close to the tip so this error doesn't occur?

Pete

I don't know. I'm going to stay with a pitot-static tube and manometer. The Testo 510i seems to be a good manometer. I wish I had just bought two of those.

The advantage to a pitot-static tube is that it requires no calibration. It's a primary standard. The manometer can also be calibrated to a water column which is, of course, also a primary standard.

The one issue I've had with the 510i is a bit of zero drift, as much as .02"wg. It's necessary to save data before re-zeroing because it's cleared with that operation.

I was wondering if you have any experience/knowledge of these kind of pitot static traverse probes? https://www.airmonitor.com/wp-conten...m-brochure.pdf Sold by VOLU-probe Pitot Airflow Measurement Traverse Probe - Air Monitor Corporation With something along those lines there would be no need to take the multiple readings and do the calculations. Probably wallet melting though. There are these but they have fewer ports for sampling. Pitot Tube | WorkACI

Pete

This is a good point worth raising but I have not seen this problem probably because the air temperature outside my ducts don't differ significantly from inside the duct ,and because I use a 240 mm diameter test duct.
I would never do a direct measurement in a 150 mm duct with any sensor as the air flow is much too turbulent.
So for testing air flow in 150 mm ducts I use a 240 mm test duct, for testing air flow in 100 mm ducts I use a 150 mm test duct ETC

However even when using smaller test ducts my Testo results still consistently agree (perhaps a bit lower ) with my Pitot Tube measurements, and with my other two hot wire anemometers which have their temp sensors much closer to the hot wire air sensor than the Testo. This suggests that air temps are not a factor for my measurements.

Also what air temperature were you performing the measurements at?
My tests usually take place with the test duct drawing air from outside my shed with my shed door wide open so the shed air is pretty well equilibrated temperature wise with the outside air. The air temps my testing normally takes place at 20ºC or above.

100mmflexytest.jpg

I’d be interested in see a diagram of how you positioned a pitot tube close to the walls of the pipe. I thought that pitot tubes where normally mounted centrally in the tube.
I use both a pitot tube/manometer combo and a Testo unit. Central and near central measurements show the Testo recorded a slightly higher flow than the Pitot (HW = 2952.75; PT = 2791.409)
However, you show close on a 38% variation between central and near wall readings for the HW (4750 vs 2900) but my tests show only about 2% variation? (The reading I have on hand are in CFM)
Maybe BobL’s reference to very similar inner and outer pipe air temps may explain this.
I also have a great deal of experience with Automotive MAF sensors, both Hot Wire, Hot film and Cold wire units.

Originally Posted by BobL

This is a good point worth raising but I have not seen this problem probably because the air temperature outside my ducts don't differ significantly from inside the duct ,and because I use a 240 mm diameter test duct....My tests usually take place with the test duct drawing air from outside my shed with my shed door wide open so the shed air is pretty well equilibrated temperature wise with the outside air. The air temps my testing normally takes place at 20ºC or above.

My testing is done entirely within my shop so the temperatures are nominally the same. The (admittedly small) differences are those caused by the differences in pressure and density inside as well as some thermal energy lost to the kinetic energy of the moving air. It's not something I take into account but if you follow AMCA 210 test methods I think you're supposed to.

The temperatures reported by the 405i for the data in the above plot ranged from 19.9°C to 20.8°C. That's inside the duct.

Originally Posted by BobL

I would never do a direct measurement in a 150 mm duct with any sensor as the air flow is much too turbulent.

Both the 405i and 510i will average as many velocity readings as you like. This helps a lot with the turbulence issue. I get very good repeatability. Is there something I'm missing?

Originally Posted by BobL

However even when using smaller test ducts my Testo results still consistently agree (perhaps a bit lower ) with my Pitot Tube measurements, and with my other two hot wire anemometers which have their temp sensors much closer to the hot wire air sensor than the Testo. This suggests that air temps are not a factor for my measurements.

Interesting. Can you share your secret sauce? I'd love to be able to use the 405i for velocity measurements and free up the 510i for pressure.

Originally Posted by dmorse

My testing is done entirely within my shop so the temperatures are nominally the same. The (admittedly small) differences are those caused by the differences in pressure and density inside as well as some thermal energy lost to the kinetic energy of the moving air. It's not something I take into account but if you follow AMCA 210 test methods I think you're supposed to. The temperatures reported by the 405i for the data in the above plot ranged from 19.9°C to 20.8°C. That's inside the duct.

Thanks for that info.

Both the 405i and 510i will average as many velocity readings as you like. This helps a lot with the turbulence issue. I get very good repeatability. Is there something I'm missing?

I agree the averaging capabilities of the Testo are very useful but I just realised that the max air speeds (3300 fpm) in your 150 mm duct is considerably lower than mine (>8000 fpm) which also explains your smaller differences between the minimum (or wall) air speed, and the higher speeds in the centre of the duct. Your differs by about 10% where as mine can be as much as 250%. At these speeds the turbulence in my 150 mm ducting is horrendous and becomes very difficult to deal with even if I was able to use the averaging capability of the Testo but is not possible because these speeds are too high for the Testo to measure. The turbulence issue and the absolute air speeds is why I use a larger diameter test duct which typically reduces the air speeds to about what you are measuring.

I haven't done anywhere near as many measurements with the Testo as I have with my other air flow meters but of the few I have done with the Testo I do not recall seeing an inverted speed profile such as you see. I will need to take a closer look when I get a chance. This could take some time as due to health problems and the hot weather I have not been able to get into my shed.

Interesting. Can you share your secret sauce? I'd love to be able to use the 405i for velocity measurements and free up the 510i for pressure.

The obvious one is to use a test duct but as your speeds are already ~3000 fpm there's less of an incentive to do this. BTW a test duct does not have to be round - eg it could be a long MDF box. eg A 300 x 300 mm cross section box would drop the nominal air speed compared to a 150 mm diameter duct by about 5X whereby the wall friction effect would be very small and turbulence minimised even further and you could just use centre point measurements especially for comparative speeds. The length of such a box should ideally be as long as possible (ideally >10X the diameter) but even a 2.4 m long box would probably suffice. You will need a decent tapered transition from square to round otherwise this will reflect some turbulence back into the test duct.

However, if you are trying to perform measurements inside existing ducting then this is not going to be possible. My problem with insitu measurements is not only the air speeds but the presence of many nearby junctions which injects even more turbulence into my system. One point I would really like to know the air speed is about 3m back from the entry into the DC but about meter before that point there are 3 Wyes, a 90º bend, and 4 gates. I have attempted to measure this point with a Pitot tube many times but it's been different every time I measure it

Dave it just might be in the first sticky. DRAFT: FAQ - Dust Extraction (Practical Aspects). Otherwise you'll have to search through Bob's posts. There are lots though.


Pete

Originally Posted by BobL

I agree the averaging capabilities of the Testo are very useful but I just realised that the max air speeds (3300 fpm) in your 150 mm duct is considerably lower than mine (>8000 fpm) which also explains your smaller differences between the minimum (or wall) air speed, and the higher speeds in the centre of the duct. Your differs by about 10% where as mine can be as much as 250%. At these speeds the turbulence in my 150 mm ducting is horrendous and becomes very difficult to deal with even if I was able to use the averaging capability of the Testo but is not possible because these speeds are too high for the Testo to measure. The turbulence issue and the absolute air speeds is why I use a larger diameter test duct which typically reduces the air speeds to about what you are measuring.

The data in the chart I posted was taken with a 75mm orifice on the test duct to give me a velocity that was both low enough for the 405i and high enough for good resolution on the 510i and pitot combination.

I've tested up to about 6000fpm with the pitot in that duct. My maximum variation from center to wall was 1.21:1 using Centroids of Equal Area traverse points. I didn't do a traverse with Log-Tchebycheff points at that flow rate but the ratio would almost certainly be a bit higher for that. I don't understand how you can get 3.5:1 at 8000fpm.

Originally Posted by BobL

The obvious one is to use a test duct but as your speeds are already ~3000 fpm there's less of an incentive to do this. BTW a test duct does not have to be round - eg it could be a long MDF box. eg A 300 x 300 mm cross section box would drop the nominal air speed compared to a 150 mm diameter duct by about 5X whereby the wall friction effect would be very small and turbulence minimised even further and you could just use centre point measurements especially for comparative speeds. The length of such a box should ideally be as long as possible (ideally >10X the diameter) but even a 2.4 m long box would probably suffice. You will need a decent tapered transition from square to round otherwise this will reflect some turbulence back into the test duct.

This is my test duct:





Does this forum allow linking to posts on other woodworking forums? If so, I'll post a link to more information about my setup and the project goals.

Originally Posted by BobL

However, if you are trying to perform measurements inside existing ducting then this is not going to be possible. My problem with insitu measurements is not only the air speeds but the presence of many nearby junctions which injects even more turbulence into my system. One point I would really like to know the air speed is about 3m back from the entry into the DC but about meter before that point there are 3 Wyes, a 90º bend, and 4 gates. I have attempted to measure this point with a Pitot tube many times but it's been different every time I measure it

In situ measurements can be frustrating. One thing I'm doing is keeping track of motor current (using a meter that averages samples taken at 0.5s intervals). After adjusting for air density and supply voltage I have good correlation with my flow measurements. I can use that to make in situ measurements that would otherwise be impossible to do with any kind of accuracy.

Originally Posted by QC Inspector

I was wondering if you have any experience/knowledge of these kind of pitot static traverse probes? https://www.airmonitor.com/wp-conten...m-brochure.pdf Sold by VOLU-probe Pitot Airflow Measurement Traverse Probe - Air Monitor Corporation With something along those lines there would be no need to take the multiple readings and do the calculations. Probably wallet melting though. There are these but they have fewer ports for sampling. Pitot Tube | WorkACI

Pete

Sorry, no, I have no experience with those. I designed one for 3D printing a while back but I couldn't get printed test parts airtight so I never actually built it.

One thing that bothers my math morality is that they average the pressure instead of the velocity. That gives you the RMS value over the traverse instead of the mean. Not a big deal but feels like a stone in my shoe.

Here's a somewhat interesting paper,

Interesting paper. The size of the copper tube they used would be a little big for our purposes but a smaller diameter version is worth a try someday. Bookmarked it.

The blue device midway on your duct. Is it to hold the pitot tube in position and alignment for a reading? Buy it with your pitot tube or design and print it yourself? Any chance of a close up?

Pete

"Does this forum allow linking to posts on other woodworking forums? If so, I'll post a link to more information about my setup and the project goals.

Per the Forum Terms and conditions of Use. https://www.woodworkforums.com/f90/woodwork-forums-terms-conditions-tou-198339
Section 15 says. * Members may include in any post, any website address other than one in which they have some commercial interest so long a it is pertinent to the thread or post.
So you are allowed to link to another site. Hopefully one that allows non members access to read the content and view the pictures without having to join or pay. Generally it is better to have the information posted here in case the linked site closes or starts to charge to see the content the thread here stays intact.

Pete

Originally Posted by QC Inspector

The blue device midway on your duct. Is it to hold the pitot tube in position and alignment for a reading? Buy it with your pitot tube or design and print it yourself? Any chance of a close up?

Pete

Thanks for asking. I was hoping someone would.




I 3D printed it. It has positive stops for the traverse points for repeatable measurements. The riser with the indexing notches can be replaced for alternate traverse patterns.




I also made a slider sized for the Hot Wire Anemometer.

Originally Posted by QC Inspector

"Does this forum allow linking to posts on other woodworking forums? If so, I'll post a link to more information about my setup and the project goals.

Per the Forum Terms and conditions of Use. https://www.woodworkforums.com/f90/woodwork-forums-terms-conditions-tou-198339
Section 15 says. * Members may include in any post, any website address other than one in which they have some commercial interest so long a it is pertinent to the thread or post.
So you are allowed to link to another site. Hopefully one that allows non members access to read the content and view the pictures without having to join or pay. Generally it is better to have the information posted here in case the linked site closes or starts to charge to see the content the thread here stays intact.

Pete

Thank you for that information. I guess I could have looked that up myself (head slap). Sorry.

I agree that keeping the post self contained is very desirable but this is an ongoing blog that may get fairly long so for now I think a link may be appropriate. I'm working on the next post now. It's all about the start current. Bores even me so it's going slowly.

Here's the link.