Cool additions to this thread for sure, you've elevated it from a new sku tweet to a proper (and timely) lecture on the subject.
I have used some Freescale sensors before, and I think they were in the middle of these ranges (7 kpa?) I also have some unamplified sensors ($1) I have used - but for airspeed other than flight.
No question that the Freescale 3.92 kPa stands alone as an ideal spec.
That said, I'd guess that I could get more than useable data from the $3 sensor above, but more importantly, there is this classification question of Absolute, Differential, and Gauge, and whereas Chris has suggested the $3 sensor is of the wrong type (irrespective of the range).
As for $18 difference, that is the effect of the cheaper part on retail boards at 3x markup. (Trivial for a $1000 board, but downright huge for a $24 autopilot.)
@Bgatti "This is the AttoPilot choice - at 3.92 kPA it has ~10 times less range."
Yes - the "AttoPilot" choice is a choice that perfectly matches dynamic range of sensor with dynamic range of what is being sensed. Any higher sensor range is unusable, and this a waste of resolution overhead. I don't make decisions lightly when designing things. I am trying to add hard-won insight here to save other people time.
Look up "airspeed" on Wikipedia, and crank through the equations. Anything over 4 kPa is unusable unless you want to fly > 300 km/h (about 200mph).
Here is why I wrote that: 0-4kPa allows 0-300 km/h airspeed. Atto (not "Otto") V2/V3 control board uses a 16bit ADC with some smoothing and filtering algorithms. The resolution with a 4 kPa sensor and this ADC gives decent airspeed resolution (the sensor pictured on the top right in bGatti's post of the Atto board), but I wouldn't want to be hamstrung with a sensor with 10X too wide range and loss of approx 10x resolution. This would affect more strongly the low airspeed data. airspeed below about 30 km/h is already "iffy" enough due to the very low # of Pa ram air pressure generated at these low speed. Maybe a UAV travelling > 100km/h wouldn't mind the loss of resolution, but for most of us fly liteweight UAVs at airspeeds between 50 and 100 km/h.
I mean, this new sensor looks really great and I like how the port mounts through a hole in the PCB - nice and secure. But any ram air pressure over 4 kPa equates to airspeeds well above our range (> 300 km/h) and thus this sensor's dynamic range simply isn't well matched to the range of conditions for the purpose people are discussing here. It would be REALLY cool if this sensor came in a 0-4 kPa version, but I didn't see that on the supplier website.
This is the AttoPilot choice - at 3.92 kPA it has ~10 times less range.
This leaves open the question of useful resolution etc as range does not equal sensativity or sig/noise ratio, it merely indicates the point of saturation. Still a $6 part difference is an $18 difference at retail...
@Dean, I checked. It looks like 37 kPa is roughly 5 psi. It's very rare to find a pressure sensor with higher sensitivity - in my experience. I think, it's analog out, so even if you were right, this would provide nearly an integer-level resolution per mph.
Range is 0 - 37 kPa. This is approx 10x too large for normal airspeeds (0 to 300 km/h). I write this comment because the wide range probably means too low of a practical resolution. My Adobe reader is hosed, so I'm not able to open the spec sheet to verify right now.
Only Absolute can only be used for altitude, Guage and Differential are much the same, but Differential has an extra port for the back side of the chamber. If you use and need to use the 2cd port, then you need a differential sensor chamber; however, most applications don't use or need the 2cd port.
New Part on Mouser.
Comments
@Dean,
Cool additions to this thread for sure, you've elevated it from a new sku tweet to a proper (and timely) lecture on the subject.
I have used some Freescale sensors before, and I think they were in the middle of these ranges (7 kpa?) I also have some unamplified sensors ($1) I have used - but for airspeed other than flight.
No question that the Freescale 3.92 kPa stands alone as an ideal spec.
That said, I'd guess that I could get more than useable data from the $3 sensor above, but more importantly, there is this classification question of Absolute, Differential, and Gauge, and whereas Chris has suggested the $3 sensor is of the wrong type (irrespective of the range).
As for $18 difference, that is the effect of the cheaper part on retail boards at 3x markup. (Trivial for a $1000 board, but downright huge for a $24 autopilot.)
http://www.sparkfun.com/search/results?term=ardupilot&what=prod...
@Bgatti "This is the AttoPilot choice - at 3.92 kPA it has ~10 times less range."
Yes - the "AttoPilot" choice is a choice that perfectly matches dynamic range of sensor with dynamic range of what is being sensed. Any higher sensor range is unusable, and this a waste of resolution overhead. I don't make decisions lightly when designing things. I am trying to add hard-won insight here to save other people time.
Look up "airspeed" on Wikipedia, and crank through the equations. Anything over 4 kPa is unusable unless you want to fly > 300 km/h (about 200mph).
And the AttoPilot sensor is only $10, not $18.
The Atto picture posted above shows:
1) top left is altitude sensor (absolute, only 1 port)
2) top right is differential (one usable port shown, the other 'vented' port on bottom of package between component and the PCB surface.
Here is why I wrote that: 0-4kPa allows 0-300 km/h airspeed. Atto (not "Otto") V2/V3 control board uses a 16bit ADC with some smoothing and filtering algorithms. The resolution with a 4 kPa sensor and this ADC gives decent airspeed resolution (the sensor pictured on the top right in bGatti's post of the Atto board), but I wouldn't want to be hamstrung with a sensor with 10X too wide range and loss of approx 10x resolution. This would affect more strongly the low airspeed data. airspeed below about 30 km/h is already "iffy" enough due to the very low # of Pa ram air pressure generated at these low speed. Maybe a UAV travelling > 100km/h wouldn't mind the loss of resolution, but for most of us fly liteweight UAVs at airspeeds between 50 and 100 km/h.
I mean, this new sensor looks really great and I like how the port mounts through a hole in the PCB - nice and secure. But any ram air pressure over 4 kPa equates to airspeeds well above our range (> 300 km/h) and thus this sensor's dynamic range simply isn't well matched to the range of conditions for the purpose people are discussing here. It would be REALLY cool if this sensor came in a 0-4 kPa version, but I didn't see that on the supplier website.
http://www.newark.com/freescale-semiconductor/mpxv5004dp/ic-pressur...
This is the AttoPilot choice - at 3.92 kPA it has ~10 times less range.
This leaves open the question of useful resolution etc as range does not equal sensativity or sig/noise ratio, it merely indicates the point of saturation. Still a $6 part difference is an $18 difference at retail...
@Dean, I checked. It looks like 37 kPa is roughly 5 psi. It's very rare to find a pressure sensor with higher sensitivity - in my experience. I think, it's analog out, so even if you were right, this would provide nearly an integer-level resolution per mph.
As one can see, the OttoPilot uses single-port pressure sensors (one absolute, one gauge).
Range is 0 - 37 kPa. This is approx 10x too large for normal airspeeds (0 to 300 km/h). I write this comment because the wide range probably means too low of a practical resolution. My Adobe reader is hosed, so I'm not able to open the spec sheet to verify right now.
http://www.components.omron.com/components/web/webfiles.nsf$FILES/f...
There are 3 Type of pressure sensors:
1. Absolute
2. Gauge
3. Differential
Only Absolute can only be used for altitude, Guage and Differential are much the same, but Differential has an extra port for the back side of the chamber. If you use and need to use the 2cd port, then you need a differential sensor chamber; however, most applications don't use or need the 2cd port.
-
1
-
2
-
3
of 3 Next