A couple very interesting sensors
13 posts
• Page 1 of 1
by 86rocco » Fri Jan 13, 2006 1:23 pm
I had a little too much time on my hands this morning and went surfing for pressure sensors which might be useful for my flowbench and I came across the and sensors. They're amplified differential pressure sensor in the 0-7" and 0-14" ranges respectively so they in the perfect range to replace the Dwyer Durablock 246 manometers commonly used on Superflow benches. BUT, what I find so interesting about these particular sensors is that their output range is not linear but rather output is proportional to the square root of the differential pressure and that means if you use one of these on a orifice type flow bench, the output voltage is linear w.r.t. flow or more precisely 37.5mV= 1% flow (when 100%=14"dp). It's like these sensors where custom made for flowbenchin', the only downside is they're kind of pricely about $85.00, I'm going to make a few more enquiries see if I can do better than that.
- 86rocco
- Posts: 319
- Joined: Fri Mar 12, 2004 9:11 am
by 86rocco » Fri Jan 13, 2006 4:23 pm
I've already got a data acquisition unit for my computer so I'd use that but I was thinking it might also be nice to have a dedicated display on the flowbench too. One of would be very easy to adapt for that purpose.
It's really cheap ($20.00) and it's got the display, a microcontroller, A/D circuitry, user interface and computer interface all built in already, you just need to solder on a few wires, connectors and program it
It's really cheap ($20.00) and it's got the display, a microcontroller, A/D circuitry, user interface and computer interface all built in already, you just need to solder on a few wires, connectors and program it
- 86rocco
- Posts: 319
- Joined: Fri Mar 12, 2004 9:11 am
by 86rocco » Fri Jan 13, 2006 7:04 pm
Actually the output voltage is offset by a quarter volt so it's 0.25-4.0 V.
I'd use this particular sensor if I wanted to display flow percentage or cfm. If I what to display was the actual differential pressure, I'd use one of the linear output sensors instead.
I'd use this particular sensor if I wanted to display flow percentage or cfm. If I what to display was the actual differential pressure, I'd use one of the linear output sensors instead.
- 86rocco
- Posts: 319
- Joined: Fri Mar 12, 2004 9:11 am
by Tony » Fri Jan 13, 2006 7:24 pm
This looks like a rather interesting way to measure pressure, it is fairly unique in how it works, and it very cleverly overcomes some of the problems, and introduces a few more. But be aware it is not perfect, there are some rather nasty hidden traps for the unwary with this !
The first limitation is that the output goes from 0.25v to 4.0v so it will need to be scaled and level shifted before it can be used. The external circuit they suggest will do that, but it adds another source of potential error.
Another difficulty I see is the speed of response, it is fairly fast at 40ms. That means that any turbulent or fluttering air pressures will probably drive this sensor totally nuts. It needs a steady constant differential pressure to work from without "acoustic noise". Trying to use small bore tube or other restriction to eliminate any fast pressure fluctuations will not work because that would introduce an undesired additional pressure drop into the system. This may or may not be a problem with your particular flow bench, but it is certainly something to think about.
A further complication is that THIS IS NOT A TRUE DIFFERENTIAL PRESSURE MEASUREMENT. They give a chart to correct for altitude. That is most unfortunate, because our test pressures of up to 28" means that this sensor will be working at pressures far below atmospheric in a flow bench.
In fact 28" of water is 69.7431 millibars, and that is 17,680 metres (58,006 feet) of altitude.
From their altitude chart you can figure out how much error that will introduce. So be aware that this sensor while interesting, is not without some rather serious limitations for our application.
Sorry to rain on the parade, but as an old and cynical electronic engineer, I have had a fair bit to do with electronic instrumentation, and the devil is in the details.
The first limitation is that the output goes from 0.25v to 4.0v so it will need to be scaled and level shifted before it can be used. The external circuit they suggest will do that, but it adds another source of potential error.
Another difficulty I see is the speed of response, it is fairly fast at 40ms. That means that any turbulent or fluttering air pressures will probably drive this sensor totally nuts. It needs a steady constant differential pressure to work from without "acoustic noise". Trying to use small bore tube or other restriction to eliminate any fast pressure fluctuations will not work because that would introduce an undesired additional pressure drop into the system. This may or may not be a problem with your particular flow bench, but it is certainly something to think about.
A further complication is that THIS IS NOT A TRUE DIFFERENTIAL PRESSURE MEASUREMENT. They give a chart to correct for altitude. That is most unfortunate, because our test pressures of up to 28" means that this sensor will be working at pressures far below atmospheric in a flow bench.
In fact 28" of water is 69.7431 millibars, and that is 17,680 metres (58,006 feet) of altitude.
From their altitude chart you can figure out how much error that will introduce. So be aware that this sensor while interesting, is not without some rather serious limitations for our application.
Sorry to rain on the parade, but as an old and cynical electronic engineer, I have had a fair bit to do with electronic instrumentation, and the devil is in the details.
Also known as the infamous "Warpspeed" on some other Forums.
- Tony
- Posts: 824
- Joined: Sat Dec 03, 2005 12:34 pm
- Location: Melbourne, Australia
by Tony » Sat Jan 14, 2006 5:34 pm
86rocco, you are absolutely right about the pressure, sorry about that. I woke up suddenly at 2AM with a flash of inspiration and realised my silly mistake. I converted 28 inches of water pressure direct to millibars of pressure, and then used a conversion program I found with Google to convert millibars pressure to altitude. That would have been correct, except we are not dealing with 28 inches of water ABSOLUTE, but 28 inches of water test pressure below atmospheric. A huge difference.
How to determine sensor response time is a rather complex and interesting question, there are several quite different ways to go about it, each of which would give a quite different result. Here is a link to a rather imaginitive way to transient test low pressure transducers, the "balloon test" . You blow up a balloon and connect it to the transducer so it reads some fixed positive static pressure. Then pop the balloon !
But even that is open to some wide interpretation. The response time could be specified as the time to reach 50% of final pressure, or the time to read to within 10% of the final pressure, or the time required to read to within 1% of final steady pressure. It can be specified any way you want to specify it, depending on the application.
Another quite different way would be to measure the frequency response, and quote the half power, or half amplitude frequency. That is more common in direct acoustic applications.
I was once employed by Westinghouse Brake and Signal in the product development section, and one of their products was a computerized safety system to monitor the brake air line pressure in a long train. The pressure transducer acted like a microphone, and the compressed air system like the old bridge to engine room speaking tube in a ship. Every valve and solenoid, and every stroke of the air compressor produced pressure fluctuations that made a complete nonsense of any steady pressure measurement. Pipe resonances created echoes, the whole thing was acoustically alive. Trying to rapidly read real time pressure samples with a microprocessor just resulted in a series of wildly scattered points around some sort of varying average.
If the system is completely linear, you can just average the output electrically or in software, but this particular transducer we are discussing here, has an inverse square characteristic. Any averaging will have to be done acoustically before the transducer. I suspect that will not be easy, especially at low frequencies.
Even airconditioning plants can produce significant low frequency rumble that is below the lowest frequencies we can hear (<20Hz). I suspect an air flow bench would produce quite a bit of roar and rumble due to flow instabilities. Some cylinder heads even whistle loudly !
By all means try it, but I suspect that the last digits of a digital display may fluctuate randomly so that a precise pressure reading down to the desired accuracy may not be easily achievable.
The water in water manometer has considerable mass, and the air in the connecting tubes is springy, so the whole thing mechanically averages any rapidly fluctuating pressure, so the problem does not occur.
I am sure it can be done, but my gut instinct tells me it is not going to be as easy as it initially appears.
How to determine sensor response time is a rather complex and interesting question, there are several quite different ways to go about it, each of which would give a quite different result. Here is a link to a rather imaginitive way to transient test low pressure transducers, the "balloon test" . You blow up a balloon and connect it to the transducer so it reads some fixed positive static pressure. Then pop the balloon !
But even that is open to some wide interpretation. The response time could be specified as the time to reach 50% of final pressure, or the time to read to within 10% of the final pressure, or the time required to read to within 1% of final steady pressure. It can be specified any way you want to specify it, depending on the application.
Another quite different way would be to measure the frequency response, and quote the half power, or half amplitude frequency. That is more common in direct acoustic applications.
I was once employed by Westinghouse Brake and Signal in the product development section, and one of their products was a computerized safety system to monitor the brake air line pressure in a long train. The pressure transducer acted like a microphone, and the compressed air system like the old bridge to engine room speaking tube in a ship. Every valve and solenoid, and every stroke of the air compressor produced pressure fluctuations that made a complete nonsense of any steady pressure measurement. Pipe resonances created echoes, the whole thing was acoustically alive. Trying to rapidly read real time pressure samples with a microprocessor just resulted in a series of wildly scattered points around some sort of varying average.
If the system is completely linear, you can just average the output electrically or in software, but this particular transducer we are discussing here, has an inverse square characteristic. Any averaging will have to be done acoustically before the transducer. I suspect that will not be easy, especially at low frequencies.
Even airconditioning plants can produce significant low frequency rumble that is below the lowest frequencies we can hear (<20Hz). I suspect an air flow bench would produce quite a bit of roar and rumble due to flow instabilities. Some cylinder heads even whistle loudly !
By all means try it, but I suspect that the last digits of a digital display may fluctuate randomly so that a precise pressure reading down to the desired accuracy may not be easily achievable.
The water in water manometer has considerable mass, and the air in the connecting tubes is springy, so the whole thing mechanically averages any rapidly fluctuating pressure, so the problem does not occur.
I am sure it can be done, but my gut instinct tells me it is not going to be as easy as it initially appears.
Also known as the infamous "Warpspeed" on some other Forums.
- Tony
- Posts: 824
- Joined: Sat Dec 03, 2005 12:34 pm
- Location: Melbourne, Australia
by 86rocco » Sat Jan 14, 2006 6:54 pm
Tony,
It is my intention to keep the water manometer in parallel with whatever sensor I eventually use, I just just like the idea of having some sort of a visual cross reference. How do you think the sensor would react favourably to that dampening effect of the water?
It is my intention to keep the water manometer in parallel with whatever sensor I eventually use, I just just like the idea of having some sort of a visual cross reference. How do you think the sensor would react favourably to that dampening effect of the water?
- 86rocco
- Posts: 319
- Joined: Fri Mar 12, 2004 9:11 am
13 posts
• Page 1 of 1
Who is online
Users browsing this forum: No registered users and 8 guests