Hank I want to know more about the Anemometer please

[beananimal]

nice job done 

I am sure, when it comes to the point for me to try something similar , it will be more a tinkering version.

[SusanJoM]

There is one serious flaw in my design. 
The tubing gets attached to that copper pipe (after the tape is removed), and it is thin and flexible to the point that it bends right over.  Something has to be rigged up to keep the tubing from crimping and blocking the air/sensor corridor.....

I still don't have the formulas right in Artisan, but I'm working on it...

[hankua]

Is the issue the flexable tubing or the copper tubing?

If it's the flexable tubing there's a semi-rigid or semi-flexible tubing that doesn't kink. Same product you would use running water lines for a filter or espresso machine.

[SusanJoM]

Thanks, Hank.  I have a wire suspended from another wire holding the silicone tubing from kinking :-))))

Can anyone tell me whether there is a diameter of the copper tube that goes to > silicone tube > Phidgets 1136 that would be required ?  Does it have to have a specific relationship to the diameter of the whole exhaust tube (2"); and/or is that dependent on the kPa levels involved?

Here's a rough sketch of what I'm trying to figure out.

The current copper tubing I have inserted is 1/8", and I'm not getting any readings, although I have ascertained that the 1136 is working and that Artisan is getting data from it, so I'm thinking it has to be in the hardware side of the equation.

[beananimal]

The current copper tubing I have inserted is 1/8", and I'm not getting any readings, although I have ascertained that the 1136 is working and that Artisan is getting data from it, so I'm thinking it has to be in the hardware side of the equation.

What does it mean "not getting any readings"?
a) no signal coming in?
b? you get just a signal line "constant 0 mbar" - no pressure change?

In case b): How deep is your 1/8" in the J-pipe.
I believe that it should be minimum down to the middle of the elbow - better deeper.

I just did some quick test with a flexibel hose through the top-hole of the T-pipe and the max. value I reached was 0.48 mbar. Not so much. Having the open end not really in the flow limited the value ~0.1 mbar

[SusanJoM]

In case b): How deep is your 1/8" in the J-pipe.
I believe that it should be minimum down to the middle of the elbow - better deeper.

Well, wouldn't you know that I just cut my only piece of 1/8" copper tubing so that it is even shorter.   I'll order a longer piece.

But check the picture.  There is no elbow after the tube.  It enters AT the elbow.

I don't speak 'mbar'.  What is that?  I'm using kPa.

[beananimal]

I made some quick measures  in the T-pipe with my new Extech HD755 0 - 0.5psi (35mbar - 3.5kPa).




The maximum value was around 0.48mbar (0.048kPa), with full speed, no chaff on the sieve, fixing the funnel by hands to reduce leakages.




The hoses were inserted as far as possible. If the tip is not passing the elbow, the values drop significantly I tried with one hose direction fan and with two hoses - one in each direction.



I started with some chaff from the last roast on the sieve
The max value then was only 0.22mbar. After removing the chaff it increased up to 0.44mbar. ( 0.044kPa)

My idea with the second hose was to compensate effect of the chaff, but this did not really work. Generally the values decreased with the second hose.
Any attempt to simulate beans in the drum did not cause any change.

50% power at my dimmer is around 75% of pressure (which is also not linear to airflow).

0.4mbar is not so much. 0.4cm water column. I expected more, but without background, just a feeling. 
( I hope my unit conversions are right )

[SusanJoM]

That is very helpful.
At some point in my fidgeting around (pun intended), it seemed to me that readings indicated flow from .1 to .6 kPa, so I think I am on the right track at least.

I am going to order a new longer piece of the 1/8"OD copper tubing to see if having it extend as far down towards the fan as possible helps.

So my next question is a mathematical one.
If you have .1, .2, .3, .4, .5, and .6 as readings, what formula would translate them to percentages, given that .1 really is 0% and .6 is 100% ?

[hankua]

I ended up with a 0-0.50 Dwyer Magnehelic gauge from EBay, and the smaller range 0-0.25 version would have been better or had more needle swing. That's in inchs of water column.

Did this with my 800n without a pitot tube, just measuring from an open port in the plenum air exhaust. If the internet conversions are correct .48 millibar worked out to .19" wc.

[beananimal]

The weak point of the soft flexible hose I used, is that it is laying aside the pipe wall. I hope to get higher values after fixing it in the center of the pipe as there should be more flow.

[beananimal]

Looking at my Huky I realized that the T-pipe is made of of two pieces, so that it is possible to install an additional piece of pipe in between. Maybe 10cm to 20cm.
Then I can drill many holes in it for several testing positions for the hose-tip without ruining my original Huky piece. Even a simple version of a venturi seems to be possible.

Then I found that the plexiglas pipe I used for the exhaust elbow of the transparent model has the needed dimension. So if I am lucky it might be possible to have transparent test setup to find the best probe setup.

[beananimal]

I continued "my little PHD"    and made more systematic tests to understand the different version of pressure measuring at the Huky and to get higher results as the signal level was not so much higher than the nominal potential error of the hardware. This is around 0.011 kPA for the Extech and is somehow similar for the Phidget 1036 version.

This here is quit different compared to the gas pressure task. There the pressure is around 30mBar // 3kPa. For fan pressure it much less, about a factor of 20 to 50.

For a single measurement point - either under-pressure or overpressure I ended up around 0.43 mbar // 0.043kPa // 43Pa.

This can be doubled by a combination - adding up both to a differential pressure.

The result at each point can be doubled, by using simplified "aerodynamics parts" in the style of a venturi or pitot pipe. Here this acts mainly by local reduction of effective cross section of the pipe. As the Huky fan has lots of power, a minimal reduction in flow seems to be easily acceptable. Don't worry about geometry details of my parts. I didn't either.

Just remember, as found the last days, all values may divide by half, as soon as there is some chaff on the sieve of the fan as it is often at the end of the roast!
Considering this it makes double sense to give more power to the fan at the end of the roast.

All values on the Extech are at full power of the fan and in mbar.
I left mbar as in kPa I would have lost one digit on the screen
( e.g.: 0,43mbar would be shown as 0.04kPa ) = 4.3mm water column, or ~0.17" wc


Pictures (part 1)
0) My first try to measure air pressure in the exhaust in water column, about one year ago - without any visible results
0)


1 + 2) I made a plexiglass pipe. Inner diameter 34mm, which just fit as a coupling.
And bent a short piece of a 3mm brass with pliers. (Aluminum pipe broke always)
1)

2)


3) With the opening of the 3mm pipe towards Huky: 0.20mbar // 0.02kPA
3)


4) Opening to the fan: 0.43mbar // 0.043kPA
4)


5) Then I made a maximum test and pluged the pipe with some foil, to see the under-pressure the Huky + funnel + leakages can produce: 0.63mbar // 0.063kPa. Not so much more.
5)


6) Then I cut a standard rubber gasket from a sewage pipe to length to set it inside the pipe to reduce the cross-section. A simple form of venturi. It is very relevant where the opening of the pipe is related to the venturi. It is some mm behind the front of the gasket.
6)
 

7)
The under-pressure increased to 0.88mbar // 0.088kPa.
In real life the part must be made from metal due to the temperature.
Aluminum or stainless steel.
7)

[SusanJoM]

So, Peter, you and I seem to be approaching this air pressure project from different angles.
Here's where I have gotten to. 

What does it mean "not getting any readings"?
....no signal coming in?

It turns out that 'not getting any readings' had nothing to do with the hardware and everything to do with the software.

Here's what I can now see in Artisan when I make adjustments to the fan speed

[beananimal]

So, Peter, you and I seem to be approaching this air pressure project from different angles. 

Yes, that is because within engineering / mechanics I know quite well what I am doing and it is fun for me.

Regarding programming anything in Artinsan, I am still waiting for my first success and therefore shy to start it. And I have to start from the very beginners level.

That is why I (also) bought the Extech, so that I can directly see the results, but have also the option to implement it into Artisan.
I am happy to see results from your meassurements and that the pressure level seems to be high enough for processing.

[beananimal]

During brainstorming with a friend I came aware that alternatively // additionally to the under-pressure the pressure behind the fan could be used.

Similar to venturi, I made sort of a pitot pipe adapter to increase the measured values in the outlet of the bowl underneath the fan.

 


With a combination of venturi and pitot adapter, and measuring the differential pressure between both it gave 1.71mbar // 0.71kPa. ( about 2 x 2 x 0.43 )
The maximum value I ever got in this game - with the maximum of effort.   
9)
 


To check the setup for somebody who doesn't, or doesn't want to do all this,
I then checked version with reduced effort.


This is both locations, but without "venturi" or "pitot" installed: 0.83mbar // 0.083kPa
10)
 


As the insert in the fan-bowl is quiet easy, and the insert in the hot Huky-exhaust might be complicated, the next test was without "venturi". 1.31mbar // 0.131kPa
11)
 


The easiest version - just to work on the outlet of the fan-bowl,
including the "pitot" shows 0.84mbar // 0.084kPa.
12)
 
13)
 


No need to touch the exhaust pipe at all
and it seems that the results are as reasonable as the originally intended version.

This solution is quiet different from what I expected at the beginning.

It seems that I am approaching to the point to try to implement the Extech into Artisan to record the pressure --- and later then how to control the fan.